diff --git "a/checkpoint-2823828/trainer_state.json" "b/checkpoint-2823828/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2823828/trainer_state.json"
@@ -0,0 +1,33411 @@
+{
+  "best_metric": 3.971450090408325,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full/lstm/4/checkpoints/checkpoint-2823828",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 2823828,
+  "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
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999905330514432e-05,
+      "loss": 4.1163,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99906673576338e-05,
+      "loss": 4.1024,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998228141012328e-05,
+      "loss": 4.0934,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9973895462612756e-05,
+      "loss": 4.0939,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9965509515102236e-05,
+      "loss": 4.0864,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995712356759172e-05,
+      "loss": 4.0884,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99487376200812e-05,
+      "loss": 4.0884,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994035167257068e-05,
+      "loss": 4.0872,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993196572506016e-05,
+      "loss": 4.0876,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992357977754964e-05,
+      "loss": 4.1002,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991519383003912e-05,
+      "loss": 4.0953,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99068078825286e-05,
+      "loss": 4.0935,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989842193501808e-05,
+      "loss": 4.1057,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989003598750756e-05,
+      "loss": 4.088,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988165003999704e-05,
+      "loss": 4.1032,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987326409248652e-05,
+      "loss": 4.0796,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9864878144975996e-05,
+      "loss": 4.084,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9856492197465476e-05,
+      "loss": 4.095,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848106249954956e-05,
+      "loss": 4.0934,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9839720302444436e-05,
+      "loss": 4.1034,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983136711254139e-05,
+      "loss": 4.092,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982298116503087e-05,
+      "loss": 4.0956,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981459521752035e-05,
+      "loss": 4.0976,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980620927000982e-05,
+      "loss": 4.0891,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97978233224993e-05,
+      "loss": 4.0959,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978943737498878e-05,
+      "loss": 4.0852,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978105142747826e-05,
+      "loss": 4.0966,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977266547996774e-05,
+      "loss": 4.0855,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976429591126095e-05,
+      "loss": 4.1022,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975590996375043e-05,
+      "loss": 4.0802,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974752401623991e-05,
+      "loss": 4.0821,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973913806872939e-05,
+      "loss": 4.0949,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9730768500022605e-05,
+      "loss": 4.0925,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9722382552512085e-05,
+      "loss": 4.0982,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9713996605001565e-05,
+      "loss": 4.088,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9705610657491045e-05,
+      "loss": 4.0891,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9697241088784254e-05,
+      "loss": 4.0903,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9688855141273734e-05,
+      "loss": 4.0981,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9680469193763214e-05,
+      "loss": 4.0965,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9672083246252694e-05,
+      "loss": 4.0819,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9663697298742174e-05,
+      "loss": 4.0793,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9655311351231654e-05,
+      "loss": 4.0773,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9646925403721134e-05,
+      "loss": 4.0958,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9638539456210614e-05,
+      "loss": 4.092,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963016988750382e-05,
+      "loss": 4.0793,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962181669760077e-05,
+      "loss": 4.1014,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.961343075009025e-05,
+      "loss": 4.0825,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960504480257973e-05,
+      "loss": 4.0971,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959665885506921e-05,
+      "loss": 4.0819,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958827290755869e-05,
+      "loss": 4.0756,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957988696004817e-05,
+      "loss": 4.081,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957150101253765e-05,
+      "loss": 4.0803,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956311506502713e-05,
+      "loss": 4.0836,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955472911751661e-05,
+      "loss": 4.0829,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954634317000609e-05,
+      "loss": 4.084,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953795722249557e-05,
+      "loss": 4.0681,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952957127498505e-05,
+      "loss": 4.0798,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9521201706278263e-05,
+      "loss": 4.0724,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951281575876774e-05,
+      "loss": 4.0845,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950442981125722e-05,
+      "loss": 4.0745,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94960438637467e-05,
+      "loss": 4.0828,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948767429503991e-05,
+      "loss": 4.0787,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947928834752939e-05,
+      "loss": 4.0689,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947090240001887e-05,
+      "loss": 4.0871,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946251645250835e-05,
+      "loss": 4.0733,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945414688380156e-05,
+      "loss": 4.0869,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944576093629104e-05,
+      "loss": 4.0743,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943737498878052e-05,
+      "loss": 4.0751,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942900542007373e-05,
+      "loss": 4.0757,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942061947256322e-05,
+      "loss": 4.076,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94122335250527e-05,
+      "loss": 4.082,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940384757754218e-05,
+      "loss": 4.0788,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939546163003166e-05,
+      "loss": 4.0814,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9387092061324866e-05,
+      "loss": 4.0992,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9378706113814346e-05,
+      "loss": 4.0901,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9370320166303826e-05,
+      "loss": 4.0799,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9361934218793306e-05,
+      "loss": 4.0835,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935354827128278e-05,
+      "loss": 4.0653,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934516232377226e-05,
+      "loss": 4.0771,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933677637626174e-05,
+      "loss": 4.0896,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932839042875122e-05,
+      "loss": 4.0905,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9320020860044435e-05,
+      "loss": 4.0792,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9311634912533915e-05,
+      "loss": 4.0665,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9303248965023395e-05,
+      "loss": 4.086,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9294863017512875e-05,
+      "loss": 4.0606,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9286493448806084e-05,
+      "loss": 4.0774,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9278107501295564e-05,
+      "loss": 4.0777,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9269721553785044e-05,
+      "loss": 4.0769,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9261335606274524e-05,
+      "loss": 4.0869,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925296603756773e-05,
+      "loss": 4.0854,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924458009005721e-05,
+      "loss": 4.0706,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923619414254669e-05,
+      "loss": 4.0774,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922780819503617e-05,
+      "loss": 4.0777,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921943862632939e-05,
+      "loss": 4.0706,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921105267881887e-05,
+      "loss": 4.0675,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920266673130835e-05,
+      "loss": 4.08,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919428078379783e-05,
+      "loss": 4.0825,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918591121509104e-05,
+      "loss": 4.0934,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917752526758052e-05,
+      "loss": 4.0732,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916913932007e-05,
+      "loss": 4.0814,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916075337255948e-05,
+      "loss": 4.0679,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915236742504896e-05,
+      "loss": 4.0794,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914398147753844e-05,
+      "loss": 4.0848,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9135611908831646e-05,
+      "loss": 4.0804,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9127225961321126e-05,
+      "loss": 4.0839,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9118840013810606e-05,
+      "loss": 4.0598,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9110454066300086e-05,
+      "loss": 4.0637,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91020844975933e-05,
+      "loss": 4.0662,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909369855008278e-05,
+      "loss": 4.0753,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908531260257226e-05,
+      "loss": 4.078,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907692665506174e-05,
+      "loss": 4.0813,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906854070755122e-05,
+      "loss": 4.051,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90601547600407e-05,
+      "loss": 4.0771,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905176881253018e-05,
+      "loss": 4.0833,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904338286501966e-05,
+      "loss": 4.0773,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903499691750914e-05,
+      "loss": 4.079,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902662734880235e-05,
+      "loss": 4.076,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901824140129183e-05,
+      "loss": 4.0631,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900985545378131e-05,
+      "loss": 4.0802,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900146950627079e-05,
+      "loss": 4.0815,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899308355876027e-05,
+      "loss": 4.0664,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898469761124975e-05,
+      "loss": 4.0661,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897631166373923e-05,
+      "loss": 4.0669,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896792571622871e-05,
+      "loss": 4.0709,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895955614752192e-05,
+      "loss": 4.0711,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8951186578815135e-05,
+      "loss": 4.078,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8942800631304615e-05,
+      "loss": 4.0721,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893441468379409e-05,
+      "loss": 4.0614,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892602873628357e-05,
+      "loss": 4.0855,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891764278877305e-05,
+      "loss": 4.0622,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890925684126253e-05,
+      "loss": 4.0784,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890087089375201e-05,
+      "loss": 4.085,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8892484946241495e-05,
+      "loss": 4.0776,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8884098998730975e-05,
+      "loss": 4.0825,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8875729430024184e-05,
+      "loss": 4.0726,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8867343482513664e-05,
+      "loss": 4.0607,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8858957535003144e-05,
+      "loss": 4.0682,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8850571587492624e-05,
+      "loss": 4.0696,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884220201878583e-05,
+      "loss": 4.0823,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883381607127531e-05,
+      "loss": 4.077,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882543012376479e-05,
+      "loss": 4.073,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881704417625427e-05,
+      "loss": 4.0734,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880867460754748e-05,
+      "loss": 4.0696,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880028866003696e-05,
+      "loss": 4.0588,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879190271252645e-05,
+      "loss": 4.0717,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878351676501593e-05,
+      "loss": 4.0793,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877514719630914e-05,
+      "loss": 4.0832,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876676124879862e-05,
+      "loss": 4.063,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.87583753012881e-05,
+      "loss": 4.0666,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.088281154632568,
+      "eval_runtime": 540.0083,
+      "eval_samples_per_second": 706.639,
+      "eval_steps_per_second": 22.083,
+      "step": 686871
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874998935377758e-05,
+      "loss": 4.0795,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874161978507079e-05,
+      "loss": 4.0761,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.873323383756027e-05,
+      "loss": 4.0689,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.872484789004975e-05,
+      "loss": 4.0642,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871646194253923e-05,
+      "loss": 4.0612,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870807599502871e-05,
+      "loss": 4.0592,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8699706426321916e-05,
+      "loss": 4.0589,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86913204788114e-05,
+      "loss": 4.0618,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.868293453130088e-05,
+      "loss": 4.0618,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.867454858379036e-05,
+      "loss": 4.0733,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.866616263627984e-05,
+      "loss": 4.0679,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865777668876932e-05,
+      "loss": 4.068,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86493907412588e-05,
+      "loss": 4.0774,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8641004793748275e-05,
+      "loss": 4.0621,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.863263522504149e-05,
+      "loss": 4.0727,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862424927753097e-05,
+      "loss": 4.0538,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8615863330020444e-05,
+      "loss": 4.0586,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8607477382509924e-05,
+      "loss": 4.0715,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8599091434999404e-05,
+      "loss": 4.0653,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8590705487488884e-05,
+      "loss": 4.0799,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85823359187821e-05,
+      "loss": 4.0627,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857394997127158e-05,
+      "loss": 4.0779,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.856556402376106e-05,
+      "loss": 4.0634,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855717807625054e-05,
+      "loss": 4.0635,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854879212874002e-05,
+      "loss": 4.0679,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85404061812295e-05,
+      "loss": 4.065,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853203661252271e-05,
+      "loss": 4.069,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852365066501219e-05,
+      "loss": 4.0585,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851526471750167e-05,
+      "loss": 4.0749,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.850687876999115e-05,
+      "loss": 4.0579,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849849282248063e-05,
+      "loss": 4.049,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849010687497011e-05,
+      "loss": 4.0715,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848172092745959e-05,
+      "loss": 4.0651,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847333497994907e-05,
+      "loss": 4.0706,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846494903243855e-05,
+      "loss": 4.0657,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8456563084928035e-05,
+      "loss": 4.0601,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8448193516221245e-05,
+      "loss": 4.0672,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8439823947514454e-05,
+      "loss": 4.0695,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8431438000003934e-05,
+      "loss": 4.0745,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8423052052493414e-05,
+      "loss": 4.0513,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8414666104982894e-05,
+      "loss": 4.0568,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8406280157472373e-05,
+      "loss": 4.0536,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8397894209961853e-05,
+      "loss": 4.0627,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838950826245133e-05,
+      "loss": 4.0683,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8381122314940807e-05,
+      "loss": 4.0557,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8372736367430287e-05,
+      "loss": 4.0739,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836435041991977e-05,
+      "loss": 4.0558,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835598085121299e-05,
+      "loss": 4.0728,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83476112825062e-05,
+      "loss": 4.0575,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833922533499568e-05,
+      "loss": 4.0503,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833083938748516e-05,
+      "loss": 4.0584,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832245343997464e-05,
+      "loss": 4.0486,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831406749246411e-05,
+      "loss": 4.0643,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830568154495359e-05,
+      "loss": 4.0549,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.829729559744307e-05,
+      "loss": 4.0619,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828890964993255e-05,
+      "loss": 4.0416,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828054008122576e-05,
+      "loss": 4.0575,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.827215413371524e-05,
+      "loss": 4.0436,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826376818620473e-05,
+      "loss": 4.062,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.825538223869421e-05,
+      "loss": 4.0496,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8247012669987416e-05,
+      "loss": 4.0557,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8238626722476896e-05,
+      "loss": 4.0515,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8230240774966376e-05,
+      "loss": 4.0444,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8221854827455856e-05,
+      "loss": 4.0603,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8213485258749065e-05,
+      "loss": 4.0503,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8205099311238545e-05,
+      "loss": 4.0635,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8196713363728025e-05,
+      "loss": 4.0445,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8188327416217505e-05,
+      "loss": 4.0525,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8179957847510714e-05,
+      "loss": 4.0478,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817158827880393e-05,
+      "loss": 4.053,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816320233129341e-05,
+      "loss": 4.0526,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815481638378289e-05,
+      "loss": 4.0548,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814643043627237e-05,
+      "loss": 4.0566,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813804448876185e-05,
+      "loss": 4.0754,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812967492005506e-05,
+      "loss": 4.0625,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812128897254454e-05,
+      "loss": 4.058,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811290302503402e-05,
+      "loss": 4.0569,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810453345632723e-05,
+      "loss": 4.044,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.809614750881671e-05,
+      "loss": 4.0489,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808776156130619e-05,
+      "loss": 4.0631,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807937561379567e-05,
+      "loss": 4.0683,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807098966628515e-05,
+      "loss": 4.0585,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8062603718774634e-05,
+      "loss": 4.04,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8054217771264114e-05,
+      "loss": 4.0564,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8045831823753594e-05,
+      "loss": 4.0407,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8037445876243074e-05,
+      "loss": 4.0509,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8029059928732554e-05,
+      "loss": 4.0533,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8020673981222034e-05,
+      "loss": 4.0545,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8012288033711514e-05,
+      "loss": 4.0622,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.800391846500472e-05,
+      "loss": 4.0576,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79955325174942e-05,
+      "loss": 4.0464,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798714656998368e-05,
+      "loss": 4.0512,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797876062247316e-05,
+      "loss": 4.0549,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797037467496264e-05,
+      "loss": 4.0444,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796200510625585e-05,
+      "loss": 4.0456,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.795361915874533e-05,
+      "loss": 4.054,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.794523321123481e-05,
+      "loss": 4.0582,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79368472637243e-05,
+      "loss": 4.0667,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792846131621377e-05,
+      "loss": 4.0483,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792007536870325e-05,
+      "loss": 4.0602,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791168942119273e-05,
+      "loss": 4.0459,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.790331985248594e-05,
+      "loss": 4.0572,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789493390497542e-05,
+      "loss": 4.0549,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78865479574649e-05,
+      "loss": 4.0563,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787816200995438e-05,
+      "loss": 4.061,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786977606244386e-05,
+      "loss": 4.0363,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786140649373707e-05,
+      "loss": 4.0404,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785302054622655e-05,
+      "loss": 4.0435,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7844634598716036e-05,
+      "loss": 4.047,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7836248651205516e-05,
+      "loss": 4.0537,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7827862703694996e-05,
+      "loss": 4.0555,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7819476756184476e-05,
+      "loss": 4.0323,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7811090808673956e-05,
+      "loss": 4.0514,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7802704861163436e-05,
+      "loss": 4.0592,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7794335292456645e-05,
+      "loss": 4.0527,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7785949344946125e-05,
+      "loss": 4.0561,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7777563397435605e-05,
+      "loss": 4.0502,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7769177449925085e-05,
+      "loss": 4.0424,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7760791502414565e-05,
+      "loss": 4.0515,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7752405554904045e-05,
+      "loss": 4.0581,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7744019607393525e-05,
+      "loss": 4.0474,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7735633659883005e-05,
+      "loss": 4.0376,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772726409117622e-05,
+      "loss": 4.0438,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77188781436657e-05,
+      "loss": 4.0464,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771049219615518e-05,
+      "loss": 4.0512,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770210624864466e-05,
+      "loss": 4.0487,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769373667993787e-05,
+      "loss": 4.0542,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768536711123108e-05,
+      "loss": 4.0379,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767698116372056e-05,
+      "loss": 4.06,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766859521621004e-05,
+      "loss": 4.0433,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766020926869952e-05,
+      "loss": 4.0522,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765183969999273e-05,
+      "loss": 4.0595,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764345375248221e-05,
+      "loss": 4.0533,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763506780497169e-05,
+      "loss": 4.0602,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7626681857461175e-05,
+      "loss": 4.0513,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7618295909950654e-05,
+      "loss": 4.037,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7609926341243864e-05,
+      "loss": 4.0437,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7601540393733344e-05,
+      "loss": 4.0408,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7593154446222823e-05,
+      "loss": 4.0585,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7584768498712303e-05,
+      "loss": 4.0555,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.757638255120178e-05,
+      "loss": 4.052,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7567996603691257e-05,
+      "loss": 4.0527,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7559610656180737e-05,
+      "loss": 4.0448,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7551224708670216e-05,
+      "loss": 4.0352,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7542855139963426e-05,
+      "loss": 4.0505,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753448557125664e-05,
+      "loss": 4.0582,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752609962374613e-05,
+      "loss": 4.0587,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75177136762356e-05,
+      "loss": 4.0378,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.750932772872508e-05,
+      "loss": 4.0436,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0722222328186035,
+      "eval_runtime": 535.0337,
+      "eval_samples_per_second": 713.209,
+      "eval_steps_per_second": 22.288,
+      "step": 763190
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750094178121456e-05,
+      "loss": 4.0804,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749255583370404e-05,
+      "loss": 4.0463,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748416988619352e-05,
+      "loss": 4.0507,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7475783938683e-05,
+      "loss": 4.0396,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746739799117248e-05,
+      "loss": 4.045,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745901204366196e-05,
+      "loss": 4.0355,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745062609615144e-05,
+      "loss": 4.0387,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744224014864092e-05,
+      "loss": 4.0356,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.74338542011304e-05,
+      "loss": 4.0406,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.742546825361988e-05,
+      "loss": 4.0496,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.741708230610936e-05,
+      "loss": 4.0485,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740869635859884e-05,
+      "loss": 4.046,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740031041108832e-05,
+      "loss": 4.0527,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73919244635778e-05,
+      "loss": 4.0415,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738353851606728e-05,
+      "loss": 4.0522,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737515256855676e-05,
+      "loss": 4.0309,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736678299984997e-05,
+      "loss": 4.0327,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735839705233945e-05,
+      "loss": 4.0502,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735001110482893e-05,
+      "loss": 4.0421,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734162515731841e-05,
+      "loss": 4.0561,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733323920980789e-05,
+      "loss": 4.0369,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732485326229737e-05,
+      "loss": 4.0577,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731646731478685e-05,
+      "loss": 4.0434,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730808136727633e-05,
+      "loss": 4.0446,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729969541976581e-05,
+      "loss": 4.0476,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7291325851059025e-05,
+      "loss": 4.0425,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7282939903548505e-05,
+      "loss": 4.0475,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7274553956037985e-05,
+      "loss": 4.0349,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7266168008527465e-05,
+      "loss": 4.0514,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7257798439820674e-05,
+      "loss": 4.0405,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7249412492310154e-05,
+      "loss": 4.027,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7241026544799634e-05,
+      "loss": 4.0439,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7232640597289114e-05,
+      "loss": 4.045,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722427102858232e-05,
+      "loss": 4.0488,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.72158850810718e-05,
+      "loss": 4.0402,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720749913356128e-05,
+      "loss": 4.0426,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71991295648545e-05,
+      "loss": 4.0454,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719074361734398e-05,
+      "loss": 4.0467,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.718235766983346e-05,
+      "loss": 4.0522,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717397172232294e-05,
+      "loss": 4.0291,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.716558577481242e-05,
+      "loss": 4.0342,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71571998273019e-05,
+      "loss": 4.031,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714881387979138e-05,
+      "loss": 4.0429,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714042793228085e-05,
+      "loss": 4.0448,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.713204198477033e-05,
+      "loss": 4.0323,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.712367241606355e-05,
+      "loss": 4.0529,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711528646855303e-05,
+      "loss": 4.036,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71069005210425e-05,
+      "loss": 4.0501,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709851457353198e-05,
+      "loss": 4.0402,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709012862602147e-05,
+      "loss": 4.0232,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.708174267851095e-05,
+      "loss": 4.0369,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7073373109804157e-05,
+      "loss": 4.0276,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7064987162293637e-05,
+      "loss": 4.0409,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7056601214783116e-05,
+      "loss": 4.0336,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7048215267272596e-05,
+      "loss": 4.0362,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7039845698565806e-05,
+      "loss": 4.018,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7031459751055285e-05,
+      "loss": 4.0367,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7023073803544765e-05,
+      "loss": 4.0234,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7014687856034245e-05,
+      "loss": 4.0417,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7006301908523725e-05,
+      "loss": 4.0305,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6997915961013205e-05,
+      "loss": 4.0329,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6989530013502685e-05,
+      "loss": 4.0318,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6981144065992165e-05,
+      "loss": 4.0195,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697277449728538e-05,
+      "loss": 4.039,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696440492857859e-05,
+      "loss": 4.0292,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695601898106807e-05,
+      "loss": 4.0432,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694763303355755e-05,
+      "loss": 4.0247,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693924708604703e-05,
+      "loss": 4.0284,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693086113853651e-05,
+      "loss": 4.0283,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.692247519102599e-05,
+      "loss": 4.0314,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691408924351547e-05,
+      "loss": 4.0296,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.690570329600495e-05,
+      "loss": 4.0368,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689733372729816e-05,
+      "loss": 4.0317,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688894777978764e-05,
+      "loss": 4.0537,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688056183227712e-05,
+      "loss": 4.0431,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6872192263570335e-05,
+      "loss": 4.0347,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6863806316059815e-05,
+      "loss": 4.0377,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6855420368549295e-05,
+      "loss": 4.022,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6847034421038775e-05,
+      "loss": 4.0323,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6838648473528255e-05,
+      "loss": 4.0349,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6830278904821464e-05,
+      "loss": 4.0468,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6821892957310944e-05,
+      "loss": 4.04,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6813507009800424e-05,
+      "loss": 4.0213,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6805121062289904e-05,
+      "loss": 4.0341,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6796735114779383e-05,
+      "loss": 4.0209,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6788349167268863e-05,
+      "loss": 4.0304,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677996321975834e-05,
+      "loss": 4.03,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677159365105155e-05,
+      "loss": 4.0353,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676320770354104e-05,
+      "loss": 4.0419,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675482175603051e-05,
+      "loss": 4.0365,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.674643580851999e-05,
+      "loss": 4.0231,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673804986100947e-05,
+      "loss": 4.0325,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672966391349895e-05,
+      "loss": 4.0339,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672127796598843e-05,
+      "loss": 4.0224,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.671289201847791e-05,
+      "loss": 4.0237,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.670452244977112e-05,
+      "loss": 4.033,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66961365022606e-05,
+      "loss": 4.0404,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668775055475008e-05,
+      "loss": 4.0431,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667936460723956e-05,
+      "loss": 4.0307,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667099503853278e-05,
+      "loss": 4.0387,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.666260909102226e-05,
+      "loss": 4.0266,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.665422314351174e-05,
+      "loss": 4.0357,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664583719600122e-05,
+      "loss": 4.0311,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66374512484907e-05,
+      "loss": 4.0409,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662906530098018e-05,
+      "loss": 4.0349,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662067935346966e-05,
+      "loss": 4.0138,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.661229340595914e-05,
+      "loss": 4.0218,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6603923837252346e-05,
+      "loss": 4.0213,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6595537889741826e-05,
+      "loss": 4.026,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6587151942231306e-05,
+      "loss": 4.0306,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6578765994720786e-05,
+      "loss": 4.0374,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6570396426013995e-05,
+      "loss": 4.0127,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6562010478503475e-05,
+      "loss": 4.0259,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655362453099296e-05,
+      "loss": 4.0384,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654523858348244e-05,
+      "loss": 4.0317,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.653685263597192e-05,
+      "loss": 4.0362,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65284666884614e-05,
+      "loss": 4.0323,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6520080740950875e-05,
+      "loss": 4.0151,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651171117224409e-05,
+      "loss": 4.0306,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650332522473357e-05,
+      "loss": 4.0397,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649493927722305e-05,
+      "loss": 4.0251,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648656970851626e-05,
+      "loss": 4.0166,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647818376100574e-05,
+      "loss": 4.0245,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646979781349522e-05,
+      "loss": 4.0267,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64614118659847e-05,
+      "loss": 4.0316,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.645302591847418e-05,
+      "loss": 4.0264,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644463997096366e-05,
+      "loss": 4.0357,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643625402345314e-05,
+      "loss": 4.0138,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642786807594262e-05,
+      "loss": 4.0388,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641949850723583e-05,
+      "loss": 4.0241,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641111255972531e-05,
+      "loss": 4.033,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.640272661221479e-05,
+      "loss": 4.0457,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.639434066470427e-05,
+      "loss": 4.0278,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.638597109599748e-05,
+      "loss": 4.0357,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637760152729069e-05,
+      "loss": 4.034,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6369215579780166e-05,
+      "loss": 4.0182,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636082963226965e-05,
+      "loss": 4.0215,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635244368475913e-05,
+      "loss": 4.026,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634405773724861e-05,
+      "loss": 4.0335,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633567178973809e-05,
+      "loss": 4.035,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632728584222757e-05,
+      "loss": 4.0331,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631889989471705e-05,
+      "loss": 4.0294,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631053032601026e-05,
+      "loss": 4.025,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630214437849974e-05,
+      "loss": 4.0147,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629377480979295e-05,
+      "loss": 4.0284,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628538886228243e-05,
+      "loss": 4.0332,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627700291477191e-05,
+      "loss": 4.0391,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626861696726139e-05,
+      "loss": 4.0199,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626023101975087e-05,
+      "loss": 4.0239,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0580315589904785,
+      "eval_runtime": 542.5463,
+      "eval_samples_per_second": 703.334,
+      "eval_steps_per_second": 21.98,
+      "step": 839509
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6251861451044087e-05,
+      "loss": 4.0335,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6243475503533566e-05,
+      "loss": 4.0312,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6235089556023046e-05,
+      "loss": 4.0248,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6226703608512526e-05,
+      "loss": 4.0204,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6218317661002006e-05,
+      "loss": 4.0265,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6209931713491486e-05,
+      "loss": 4.0148,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6201545765980966e-05,
+      "loss": 4.0147,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6193176197274175e-05,
+      "loss": 4.0162,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6184790249763655e-05,
+      "loss": 4.0249,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6176404302253135e-05,
+      "loss": 4.0265,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6168034733546344e-05,
+      "loss": 4.0241,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6159648786035824e-05,
+      "loss": 4.0259,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6151262838525304e-05,
+      "loss": 4.0315,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6142876891014784e-05,
+      "loss": 4.0216,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.613449094350427e-05,
+      "loss": 4.032,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612610499599375e-05,
+      "loss": 4.0137,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611771904848323e-05,
+      "loss": 4.0108,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610933310097271e-05,
+      "loss": 4.0309,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6100947153462184e-05,
+      "loss": 4.0248,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6092561205951664e-05,
+      "loss": 4.0319,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608419163724488e-05,
+      "loss": 4.0209,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607580568973436e-05,
+      "loss": 4.0359,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606741974222383e-05,
+      "loss": 4.0223,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605903379471331e-05,
+      "loss": 4.0249,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605064784720279e-05,
+      "loss": 4.0242,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.604226189969227e-05,
+      "loss": 4.0251,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603387595218175e-05,
+      "loss": 4.0244,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602549000467124e-05,
+      "loss": 4.0154,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601712043596445e-05,
+      "loss": 4.0321,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600875086725766e-05,
+      "loss": 4.0199,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600036491974714e-05,
+      "loss": 4.0114,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599197897223662e-05,
+      "loss": 4.0221,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59835930247261e-05,
+      "loss": 4.0228,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597520707721558e-05,
+      "loss": 4.0321,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596682112970506e-05,
+      "loss": 4.0179,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595843518219454e-05,
+      "loss": 4.023,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5950065613487747e-05,
+      "loss": 4.027,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5941679665977227e-05,
+      "loss": 4.0252,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5933293718466706e-05,
+      "loss": 4.0351,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.592490777095619e-05,
+      "loss": 4.0115,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591652182344567e-05,
+      "loss": 4.016,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590813587593515e-05,
+      "loss": 4.0125,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589974992842463e-05,
+      "loss": 4.0185,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589136398091411e-05,
+      "loss": 4.0267,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588299441220732e-05,
+      "loss": 4.014,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587462484350053e-05,
+      "loss": 4.0309,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586623889599001e-05,
+      "loss": 4.0173,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585785294847949e-05,
+      "loss": 4.0326,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584946700096897e-05,
+      "loss": 4.0223,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584108105345845e-05,
+      "loss": 4.0026,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583269510594793e-05,
+      "loss": 4.0157,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582432553724114e-05,
+      "loss": 4.012,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581593958973063e-05,
+      "loss": 4.0213,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580755364222011e-05,
+      "loss": 4.0137,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579916769470959e-05,
+      "loss": 4.018,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579078174719907e-05,
+      "loss": 4.0019,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.578239579968855e-05,
+      "loss": 4.0165,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577400985217802e-05,
+      "loss": 4.0075,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57656239046675e-05,
+      "loss": 4.0224,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5757254335960716e-05,
+      "loss": 4.0027,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574886838845019e-05,
+      "loss": 4.0184,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574048244093967e-05,
+      "loss": 4.0123,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5732112872232885e-05,
+      "loss": 4.0047,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5723726924722365e-05,
+      "loss": 4.0174,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5715340977211845e-05,
+      "loss": 4.0123,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5706955029701325e-05,
+      "loss": 4.0226,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5698569082190804e-05,
+      "loss": 4.004,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5690183134680284e-05,
+      "loss": 4.0135,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5681797187169764e-05,
+      "loss": 4.0053,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5673411239659244e-05,
+      "loss": 4.0129,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5665041670952453e-05,
+      "loss": 4.0096,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565667210224566e-05,
+      "loss": 4.0203,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564828615473514e-05,
+      "loss": 4.0113,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563990020722462e-05,
+      "loss": 4.0345,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56315142597141e-05,
+      "loss": 4.0268,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562312831220358e-05,
+      "loss": 4.0139,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561474236469306e-05,
+      "loss": 4.016,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.560635641718255e-05,
+      "loss": 4.0078,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559798684847576e-05,
+      "loss": 4.0102,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558960090096524e-05,
+      "loss": 4.018,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558121495345472e-05,
+      "loss": 4.0233,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55728290059442e-05,
+      "loss": 4.0264,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556444305843368e-05,
+      "loss": 4.003,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555605711092316e-05,
+      "loss": 4.0117,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554767116341264e-05,
+      "loss": 3.9998,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553928521590212e-05,
+      "loss": 4.0178,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553091564719533e-05,
+      "loss": 4.0103,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.552252969968481e-05,
+      "loss": 4.0167,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551414375217429e-05,
+      "loss": 4.0225,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55057741834675e-05,
+      "loss": 4.0212,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549738823595698e-05,
+      "loss": 4.0038,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548900228844646e-05,
+      "loss": 4.0098,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548061634093594e-05,
+      "loss": 4.0174,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.547223039342542e-05,
+      "loss": 4.0046,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54638444459149e-05,
+      "loss": 4.0032,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.545545849840438e-05,
+      "loss": 4.016,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5447072550893856e-05,
+      "loss": 4.0212,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5438686603383336e-05,
+      "loss": 4.027,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543031703467655e-05,
+      "loss": 4.0166,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5421931087166025e-05,
+      "loss": 4.0176,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5413545139655505e-05,
+      "loss": 4.0074,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5405159192144985e-05,
+      "loss": 4.0179,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539677324463447e-05,
+      "loss": 4.0133,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538840367592768e-05,
+      "loss": 4.0241,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5380034107220896e-05,
+      "loss": 4.0158,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5371648159710376e-05,
+      "loss": 4.003,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5363262212199856e-05,
+      "loss": 4.003,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535487626468933e-05,
+      "loss": 4.0004,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534649031717881e-05,
+      "loss": 4.0125,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533810436966829e-05,
+      "loss": 4.0174,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532971842215777e-05,
+      "loss": 4.0161,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532134885345098e-05,
+      "loss": 3.9986,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531296290594046e-05,
+      "loss": 4.0046,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.530457695842994e-05,
+      "loss": 4.018,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5296191010919425e-05,
+      "loss": 4.0181,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5287821442212634e-05,
+      "loss": 4.0201,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5279435494702114e-05,
+      "loss": 4.0119,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5271049547191594e-05,
+      "loss": 4.0015,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5262663599681074e-05,
+      "loss": 4.0126,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5254277652170554e-05,
+      "loss": 4.0198,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524590808346376e-05,
+      "loss": 4.0097,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523752213595324e-05,
+      "loss": 3.9991,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522913618844272e-05,
+      "loss": 4.0049,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52207502409322e-05,
+      "loss": 4.0089,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521238067222541e-05,
+      "loss": 4.0147,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520399472471489e-05,
+      "loss": 4.0088,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519560877720437e-05,
+      "loss": 4.0194,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518722282969386e-05,
+      "loss": 3.9984,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517883688218334e-05,
+      "loss": 4.0198,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517045093467282e-05,
+      "loss": 4.005,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51620649871623e-05,
+      "loss": 4.0166,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515367903965178e-05,
+      "loss": 4.0258,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.514529309214126e-05,
+      "loss": 4.0163,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.513690714463074e-05,
+      "loss": 4.014,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512852119712021e-05,
+      "loss": 4.0192,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512013524960969e-05,
+      "loss": 3.9998,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.511176568090291e-05,
+      "loss": 4.0069,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.510337973339239e-05,
+      "loss": 4.005,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.509499378588186e-05,
+      "loss": 4.0204,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5086624217175076e-05,
+      "loss": 4.0171,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5078238269664556e-05,
+      "loss": 4.0191,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506985232215404e-05,
+      "loss": 4.0121,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5061466374643516e-05,
+      "loss": 4.0055,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5053080427132996e-05,
+      "loss": 3.9981,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5044694479622476e-05,
+      "loss": 4.015,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5036308532111956e-05,
+      "loss": 4.0167,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5027922584601436e-05,
+      "loss": 4.0238,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5019553015894645e-05,
+      "loss": 4.0024,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5011167068384125e-05,
+      "loss": 4.0089,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.047278881072998,
+      "eval_runtime": 541.5004,
+      "eval_samples_per_second": 704.692,
+      "eval_steps_per_second": 22.022,
+      "step": 915828
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5002781120873605e-05,
+      "loss": 4.0035,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4994395173363085e-05,
+      "loss": 4.028,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4986009225852565e-05,
+      "loss": 4.0006,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4977623278342045e-05,
+      "loss": 3.9909,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4969237330831525e-05,
+      "loss": 4.0146,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496085138332101e-05,
+      "loss": 4.0095,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495246543581049e-05,
+      "loss": 4.0101,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494407948829997e-05,
+      "loss": 4.0024,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493569354078945e-05,
+      "loss": 4.0069,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492730759327893e-05,
+      "loss": 3.9897,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4918921645768405e-05,
+      "loss": 4.0237,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4910535698257885e-05,
+      "loss": 4.0024,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4902149750747364e-05,
+      "loss": 4.014,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4893763803236844e-05,
+      "loss": 4.0133,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4885377855726324e-05,
+      "loss": 4.0057,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4876991908215804e-05,
+      "loss": 4.0004,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4868605960705284e-05,
+      "loss": 4.0116,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4860220013194764e-05,
+      "loss": 4.0084,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4851834065684244e-05,
+      "loss": 4.004,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4843448118173724e-05,
+      "loss": 4.0198,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483507854946694e-05,
+      "loss": 4.0281,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482669260195642e-05,
+      "loss": 4.0099,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481832303324963e-05,
+      "loss": 3.999,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480993708573911e-05,
+      "loss": 4.011,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480155113822859e-05,
+      "loss": 4.0044,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.47931815695218e-05,
+      "loss": 4.0008,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478479562201128e-05,
+      "loss": 4.0032,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477640967450076e-05,
+      "loss": 4.003,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476802372699024e-05,
+      "loss": 4.0063,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.475963777947972e-05,
+      "loss": 4.0128,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.47512518319692e-05,
+      "loss": 4.0126,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4742882263262414e-05,
+      "loss": 4.0061,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4734496315751894e-05,
+      "loss": 4.0082,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4726110368241374e-05,
+      "loss": 4.0155,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4717724420730854e-05,
+      "loss": 3.9995,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4709338473220334e-05,
+      "loss": 4.0043,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4700952525709814e-05,
+      "loss": 4.0027,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469256657819929e-05,
+      "loss": 4.0184,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46841970094925e-05,
+      "loss": 4.0195,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467581106198198e-05,
+      "loss": 4.0171,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466742511447146e-05,
+      "loss": 4.0125,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4659039166960936e-05,
+      "loss": 4.0101,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4650653219450416e-05,
+      "loss": 4.0021,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4642267271939896e-05,
+      "loss": 3.9966,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463388132442938e-05,
+      "loss": 3.9943,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462549537691886e-05,
+      "loss": 3.9979,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461712580821207e-05,
+      "loss": 4.0004,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460873986070155e-05,
+      "loss": 4.0034,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460035391319103e-05,
+      "loss": 3.9986,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459196796568051e-05,
+      "loss": 4.008,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.458358201816999e-05,
+      "loss": 3.9934,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45752124494632e-05,
+      "loss": 4.0022,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.456682650195268e-05,
+      "loss": 4.0053,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455845693324589e-05,
+      "loss": 3.9816,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455007098573537e-05,
+      "loss": 3.9906,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.454168503822485e-05,
+      "loss": 3.9992,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4533299090714336e-05,
+      "loss": 4.0122,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4524913143203816e-05,
+      "loss": 3.9875,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4516527195693296e-05,
+      "loss": 4.0019,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4508141248182776e-05,
+      "loss": 4.0033,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4499755300672256e-05,
+      "loss": 3.9951,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4491369353161736e-05,
+      "loss": 3.9905,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4482983405651216e-05,
+      "loss": 3.9886,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4474613836944425e-05,
+      "loss": 3.9903,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4466227889433905e-05,
+      "loss": 4.0083,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4457841941923385e-05,
+      "loss": 4.0077,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4449455994412865e-05,
+      "loss": 3.9987,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4441070046902345e-05,
+      "loss": 4.0015,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4432684099391825e-05,
+      "loss": 4.0078,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4424298151881305e-05,
+      "loss": 3.9873,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.441592858317452e-05,
+      "loss": 4.0135,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4407542635664e-05,
+      "loss": 3.999,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439915668815348e-05,
+      "loss": 3.9957,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4390770740642954e-05,
+      "loss": 3.9946,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4382384793132433e-05,
+      "loss": 4.0142,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4373998845621913e-05,
+      "loss": 3.994,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436561289811139e-05,
+      "loss": 4.0104,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435722695060087e-05,
+      "loss": 4.002,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434884100309035e-05,
+      "loss": 3.999,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434047143438356e-05,
+      "loss": 3.9974,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433210186567677e-05,
+      "loss": 4.0016,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432371591816626e-05,
+      "loss": 3.9947,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.431532997065574e-05,
+      "loss": 3.9968,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.430694402314522e-05,
+      "loss": 3.9978,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.42985580756347e-05,
+      "loss": 3.9828,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429017212812418e-05,
+      "loss": 4.0056,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.428180255941739e-05,
+      "loss": 3.986,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427341661190687e-05,
+      "loss": 3.9976,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.426503066439635e-05,
+      "loss": 3.9874,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425664471688583e-05,
+      "loss": 4.0096,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424825876937531e-05,
+      "loss": 4.0037,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423987282186479e-05,
+      "loss": 4.0067,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423148687435427e-05,
+      "loss": 4.0089,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.422310092684375e-05,
+      "loss": 3.9861,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421471497933323e-05,
+      "loss": 4.0015,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4206345410626436e-05,
+      "loss": 3.9983,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419797584191965e-05,
+      "loss": 3.9941,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418958989440913e-05,
+      "loss": 3.9881,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418120394689861e-05,
+      "loss": 3.9931,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417281799938809e-05,
+      "loss": 4.0089,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.416443205187757e-05,
+      "loss": 3.9925,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.415604610436705e-05,
+      "loss": 4.0082,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414766015685653e-05,
+      "loss": 4.0019,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413927420934601e-05,
+      "loss": 3.996,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413090464063922e-05,
+      "loss": 3.9833,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41225186931287e-05,
+      "loss": 3.9974,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.411414912442191e-05,
+      "loss": 3.9981,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.410576317691139e-05,
+      "loss": 4.0046,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4097377229400876e-05,
+      "loss": 3.9896,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4088991281890356e-05,
+      "loss": 3.989,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4080605334379836e-05,
+      "loss": 3.9957,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407221938686931e-05,
+      "loss": 4.0016,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406383343935879e-05,
+      "loss": 3.9878,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4055463870652005e-05,
+      "loss": 3.9943,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4047077923141485e-05,
+      "loss": 3.9931,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403869197563096e-05,
+      "loss": 3.9874,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403030602812044e-05,
+      "loss": 3.9922,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402192008060992e-05,
+      "loss": 3.9973,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4013550511903134e-05,
+      "loss": 3.9969,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4005164564392614e-05,
+      "loss": 3.9929,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3996778616882094e-05,
+      "loss": 4.0073,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3988392669371574e-05,
+      "loss": 4.0124,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398002310066478e-05,
+      "loss": 3.9927,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397163715315426e-05,
+      "loss": 4.0018,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.396325120564374e-05,
+      "loss": 3.9931,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.395486525813322e-05,
+      "loss": 3.996,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.394649568942643e-05,
+      "loss": 4.001,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393810974191591e-05,
+      "loss": 3.9925,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392972379440539e-05,
+      "loss": 3.9981,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392133784689487e-05,
+      "loss": 3.9912,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.391296827818808e-05,
+      "loss": 4.0089,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.390458233067757e-05,
+      "loss": 3.9907,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.389619638316705e-05,
+      "loss": 4.0106,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.388781043565653e-05,
+      "loss": 3.9997,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387944086694974e-05,
+      "loss": 4.0008,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387105491943922e-05,
+      "loss": 3.9976,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38626689719287e-05,
+      "loss": 3.9968,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.385428302441818e-05,
+      "loss": 4.0062,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3845897076907657e-05,
+      "loss": 3.9933,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3837511129397137e-05,
+      "loss": 4.0012,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3829141560690346e-05,
+      "loss": 3.9992,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3820755613179826e-05,
+      "loss": 3.9937,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3812369665669306e-05,
+      "loss": 3.9906,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3803983718158785e-05,
+      "loss": 4.0007,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3795597770648265e-05,
+      "loss": 3.996,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378722820194148e-05,
+      "loss": 3.9998,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377884225443096e-05,
+      "loss": 3.998,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377047268572417e-05,
+      "loss": 3.9946,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.376208673821365e-05,
+      "loss": 3.9888,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.037090301513672,
+      "eval_runtime": 295.0004,
+      "eval_samples_per_second": 1293.527,
+      "eval_steps_per_second": 40.424,
+      "step": 992148
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.375370079070313e-05,
+      "loss": 3.995,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.374531484319261e-05,
+      "loss": 4.0131,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.373692889568209e-05,
+      "loss": 3.9885,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372854294817157e-05,
+      "loss": 3.9757,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372015700066105e-05,
+      "loss": 3.9997,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.371177105315053e-05,
+      "loss": 3.992,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.370338510564001e-05,
+      "loss": 3.9963,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.369501553693322e-05,
+      "loss": 3.986,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.36866295894227e-05,
+      "loss": 3.9899,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3678243641912186e-05,
+      "loss": 3.975,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3669857694401666e-05,
+      "loss": 4.0063,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3661471746891146e-05,
+      "loss": 3.9862,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3653102178184355e-05,
+      "loss": 3.9971,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3644716230673835e-05,
+      "loss": 4.0013,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3636330283163315e-05,
+      "loss": 3.9903,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3627944335652795e-05,
+      "loss": 3.9822,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361955838814227e-05,
+      "loss": 4.002,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361117244063175e-05,
+      "loss": 3.9926,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.360278649312123e-05,
+      "loss": 3.9878,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3594416924414444e-05,
+      "loss": 4.0056,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.358604735570765e-05,
+      "loss": 4.0141,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357766140819714e-05,
+      "loss": 3.9966,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356927546068662e-05,
+      "loss": 3.9843,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356088951317609e-05,
+      "loss": 3.9945,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.355250356566557e-05,
+      "loss": 3.9884,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.354413399695879e-05,
+      "loss": 3.9875,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.353574804944827e-05,
+      "loss": 3.9914,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.352736210193774e-05,
+      "loss": 3.9863,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351897615442722e-05,
+      "loss": 3.9896,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35105902069167e-05,
+      "loss": 3.995,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.350222063820991e-05,
+      "loss": 3.9981,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.349383469069939e-05,
+      "loss": 3.9922,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3485465121992606e-05,
+      "loss": 3.9914,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.347707917448209e-05,
+      "loss": 4.0021,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3468693226971566e-05,
+      "loss": 3.9849,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3460307279461046e-05,
+      "loss": 3.9899,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3451921331950526e-05,
+      "loss": 3.9867,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3443535384440006e-05,
+      "loss": 4.0008,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3435149436929486e-05,
+      "loss": 4.002,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3426763489418966e-05,
+      "loss": 4.0065,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3418377541908446e-05,
+      "loss": 3.9982,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3410007973201655e-05,
+      "loss": 3.9971,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3401622025691135e-05,
+      "loss": 3.986,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3393236078180615e-05,
+      "loss": 3.9832,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3384850130670095e-05,
+      "loss": 3.9806,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3376464183159575e-05,
+      "loss": 3.9784,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336807823564906e-05,
+      "loss": 3.983,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335969228813854e-05,
+      "loss": 3.9885,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335130634062802e-05,
+      "loss": 3.9872,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334295315072496e-05,
+      "loss": 3.9911,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.333456720321444e-05,
+      "loss": 3.9752,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.332618125570392e-05,
+      "loss": 3.9875,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.33177953081934e-05,
+      "loss": 3.9889,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.330940936068288e-05,
+      "loss": 3.9711,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.330102341317236e-05,
+      "loss": 3.9775,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.329263746566184e-05,
+      "loss": 3.986,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.328425151815132e-05,
+      "loss": 3.9972,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32758655706408e-05,
+      "loss": 3.9699,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3267496001934015e-05,
+      "loss": 3.9921,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3259126433227224e-05,
+      "loss": 3.9925,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3250740485716704e-05,
+      "loss": 3.9789,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3242354538206184e-05,
+      "loss": 3.9756,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3233968590695664e-05,
+      "loss": 3.9729,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3225582643185144e-05,
+      "loss": 3.976,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3217196695674624e-05,
+      "loss": 3.9953,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3208810748164104e-05,
+      "loss": 3.9944,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320042480065358e-05,
+      "loss": 3.9817,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319203885314306e-05,
+      "loss": 3.9868,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318366928443627e-05,
+      "loss": 3.9919,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317528333692575e-05,
+      "loss": 3.9793,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316689738941523e-05,
+      "loss": 3.9946,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315851144190471e-05,
+      "loss": 3.9835,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315014187319793e-05,
+      "loss": 3.9852,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.31417559256874e-05,
+      "loss": 3.9797,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.313336997817688e-05,
+      "loss": 3.9957,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.312498403066636e-05,
+      "loss": 3.9841,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.311661446195958e-05,
+      "loss": 3.9968,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310822851444905e-05,
+      "loss": 3.9907,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309984256693853e-05,
+      "loss": 3.9878,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309147299823175e-05,
+      "loss": 3.9823,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.308308705072122e-05,
+      "loss": 3.9902,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.307470110321071e-05,
+      "loss": 3.9806,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.306631515570019e-05,
+      "loss": 3.9799,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.305792920818967e-05,
+      "loss": 3.9825,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304954326067915e-05,
+      "loss": 3.9723,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304115731316863e-05,
+      "loss": 3.99,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3032771365658107e-05,
+      "loss": 3.973,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3024401796951316e-05,
+      "loss": 3.9855,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3016015849440796e-05,
+      "loss": 3.9727,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3007646280734005e-05,
+      "loss": 3.9922,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2999260333223485e-05,
+      "loss": 3.9912,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2990874385712965e-05,
+      "loss": 3.9927,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2982488438202445e-05,
+      "loss": 3.99,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2974102490691925e-05,
+      "loss": 3.9761,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.296573292198514e-05,
+      "loss": 3.9906,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295734697447462e-05,
+      "loss": 3.983,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29489610269641e-05,
+      "loss": 3.9831,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294057507945358e-05,
+      "loss": 3.9753,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.293218913194306e-05,
+      "loss": 3.9788,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.292380318443254e-05,
+      "loss": 3.9921,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.291541723692202e-05,
+      "loss": 3.9861,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.290704766821523e-05,
+      "loss": 3.9943,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289866172070471e-05,
+      "loss": 3.9838,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289027577319419e-05,
+      "loss": 3.9834,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.288188982568367e-05,
+      "loss": 3.9742,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.287350387817315e-05,
+      "loss": 3.978,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.286511793066263e-05,
+      "loss": 3.9887,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.285674836195584e-05,
+      "loss": 3.9909,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2848362414445325e-05,
+      "loss": 3.9762,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2839976466934805e-05,
+      "loss": 3.9722,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2831590519424285e-05,
+      "loss": 3.9847,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2823204571913765e-05,
+      "loss": 3.9864,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.281481862440324e-05,
+      "loss": 3.9715,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2806449055696454e-05,
+      "loss": 3.984,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2798063108185934e-05,
+      "loss": 3.9753,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278967716067541e-05,
+      "loss": 3.9773,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278129121316489e-05,
+      "loss": 3.9736,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.277290526565437e-05,
+      "loss": 3.9856,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.276451931814385e-05,
+      "loss": 3.9838,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.275613337063333e-05,
+      "loss": 3.9805,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.274774742312281e-05,
+      "loss": 3.9905,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273937785441602e-05,
+      "loss": 3.998,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273100828570924e-05,
+      "loss": 3.9821,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.272262233819871e-05,
+      "loss": 3.983,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.271423639068819e-05,
+      "loss": 3.9828,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.270585044317767e-05,
+      "loss": 3.9828,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269746449566715e-05,
+      "loss": 3.9819,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268907854815663e-05,
+      "loss": 3.9781,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268069260064611e-05,
+      "loss": 3.9838,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.267232303193932e-05,
+      "loss": 3.9771,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26639370844288e-05,
+      "loss": 3.9967,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.265555113691828e-05,
+      "loss": 3.9774,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264716518940776e-05,
+      "loss": 3.9958,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263877924189725e-05,
+      "loss": 3.9847,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2630409673190456e-05,
+      "loss": 3.9868,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2622023725679936e-05,
+      "loss": 3.9798,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2613637778169416e-05,
+      "loss": 3.9891,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2605251830658896e-05,
+      "loss": 3.9926,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2596882261952105e-05,
+      "loss": 3.9819,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2588496314441585e-05,
+      "loss": 3.9864,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2580110366931065e-05,
+      "loss": 3.988,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2571740798224274e-05,
+      "loss": 3.9786,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2563354850713754e-05,
+      "loss": 3.9801,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2554968903203234e-05,
+      "loss": 3.988,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2546582955692714e-05,
+      "loss": 3.9832,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25381970081822e-05,
+      "loss": 3.9834,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252982743947541e-05,
+      "loss": 3.9844,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252145787076862e-05,
+      "loss": 3.9801,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25130719232581e-05,
+      "loss": 3.9758,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.02940559387207,
+      "eval_runtime": 295.3105,
+      "eval_samples_per_second": 1292.169,
+      "eval_steps_per_second": 40.381,
+      "step": 1068468
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.250468597574758e-05,
+      "loss": 3.9679,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.249630002823706e-05,
+      "loss": 3.9704,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.248791408072654e-05,
+      "loss": 3.9779,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247952813321602e-05,
+      "loss": 3.9915,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.24711421857055e-05,
+      "loss": 3.9844,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.246275623819498e-05,
+      "loss": 3.9809,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.245437029068446e-05,
+      "loss": 3.9792,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.244598434317394e-05,
+      "loss": 3.982,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.243759839566342e-05,
+      "loss": 3.9715,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.24292124481529e-05,
+      "loss": 3.979,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242082650064238e-05,
+      "loss": 3.9813,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.241244055313186e-05,
+      "loss": 3.9945,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2404070984425074e-05,
+      "loss": 3.9921,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.239568503691455e-05,
+      "loss": 3.9592,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238729908940403e-05,
+      "loss": 3.9819,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.237891314189351e-05,
+      "loss": 3.9679,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.237052719438299e-05,
+      "loss": 3.9826,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.236214124687247e-05,
+      "loss": 3.9778,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.235375529936195e-05,
+      "loss": 3.984,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.234536935185143e-05,
+      "loss": 3.9724,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.233698340434091e-05,
+      "loss": 3.9961,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232859745683039e-05,
+      "loss": 3.9772,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232021150931987e-05,
+      "loss": 3.9832,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.231182556180935e-05,
+      "loss": 3.9785,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.230345599310256e-05,
+      "loss": 3.9908,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.229507004559204e-05,
+      "loss": 3.9777,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.228668409808152e-05,
+      "loss": 3.9705,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2278298150571e-05,
+      "loss": 3.9753,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.226992858186421e-05,
+      "loss": 3.9811,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.226154263435369e-05,
+      "loss": 3.9724,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.225315668684317e-05,
+      "loss": 3.9764,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.224477073933265e-05,
+      "loss": 3.9777,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223638479182213e-05,
+      "loss": 3.9876,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2227998844311605e-05,
+      "loss": 3.982,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2219612896801085e-05,
+      "loss": 3.9846,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.22112433280943e-05,
+      "loss": 3.9834,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220285738058379e-05,
+      "loss": 3.9864,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.219447143307326e-05,
+      "loss": 3.964,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.218608548556274e-05,
+      "loss": 3.973,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.217769953805222e-05,
+      "loss": 3.9764,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.21693135905417e-05,
+      "loss": 3.9638,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216092764303118e-05,
+      "loss": 3.9776,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.215254169552066e-05,
+      "loss": 3.9772,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.214417212681387e-05,
+      "loss": 3.9716,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.213578617930335e-05,
+      "loss": 3.983,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212740023179283e-05,
+      "loss": 3.9796,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211901428428231e-05,
+      "loss": 3.9838,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2110644715575525e-05,
+      "loss": 3.9788,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2102258768065005e-05,
+      "loss": 3.9591,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2093872820554485e-05,
+      "loss": 3.9755,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2085503251847694e-05,
+      "loss": 3.9765,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2077117304337174e-05,
+      "loss": 3.9817,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2068731356826654e-05,
+      "loss": 3.9699,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2060345409316134e-05,
+      "loss": 3.9703,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2051959461805614e-05,
+      "loss": 3.9671,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2043573514295094e-05,
+      "loss": 3.9687,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2035187566784574e-05,
+      "loss": 3.9769,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2026801619274054e-05,
+      "loss": 3.9571,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2018415671763534e-05,
+      "loss": 3.9836,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201004610305674e-05,
+      "loss": 3.982,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.200166015554622e-05,
+      "loss": 3.9541,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.19932742080357e-05,
+      "loss": 3.9865,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.198488826052519e-05,
+      "loss": 3.9597,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.19765186918184e-05,
+      "loss": 3.9692,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.196813274430788e-05,
+      "loss": 3.9657,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.195974679679736e-05,
+      "loss": 3.9761,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.195136084928684e-05,
+      "loss": 3.9636,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.194297490177632e-05,
+      "loss": 3.981,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.193458895426579e-05,
+      "loss": 3.9544,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.192620300675527e-05,
+      "loss": 3.9648,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191781705924475e-05,
+      "loss": 3.9807,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190943111173423e-05,
+      "loss": 3.9718,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190106154302744e-05,
+      "loss": 3.9801,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.189267559551693e-05,
+      "loss": 3.9778,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.188428964800641e-05,
+      "loss": 3.9877,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.187590370049589e-05,
+      "loss": 3.9659,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1867534131789096e-05,
+      "loss": 3.9856,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1859148184278576e-05,
+      "loss": 3.9668,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1850762236768056e-05,
+      "loss": 3.9695,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1842376289257536e-05,
+      "loss": 3.9713,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1833990341747016e-05,
+      "loss": 3.9809,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1825620773040225e-05,
+      "loss": 3.9754,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1817234825529705e-05,
+      "loss": 3.9873,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1808848878019185e-05,
+      "loss": 3.9636,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1800462930508665e-05,
+      "loss": 3.9572,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.179209336180188e-05,
+      "loss": 3.9678,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.178370741429136e-05,
+      "loss": 3.973,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.177533784558457e-05,
+      "loss": 3.9707,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.176695189807405e-05,
+      "loss": 3.9707,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.175856595056353e-05,
+      "loss": 3.9717,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.175018000305301e-05,
+      "loss": 3.9706,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.174179405554249e-05,
+      "loss": 3.9707,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.173340810803197e-05,
+      "loss": 3.9661,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.172502216052145e-05,
+      "loss": 3.9818,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.171663621301093e-05,
+      "loss": 3.9842,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170826664430414e-05,
+      "loss": 3.9773,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169988069679362e-05,
+      "loss": 3.9622,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1691511128086835e-05,
+      "loss": 3.9687,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1683125180576315e-05,
+      "loss": 3.9663,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1674739233065795e-05,
+      "loss": 3.9787,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1666353285555275e-05,
+      "loss": 3.9785,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1657967338044755e-05,
+      "loss": 3.9629,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1649581390534234e-05,
+      "loss": 3.9727,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1641195443023714e-05,
+      "loss": 3.9732,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1632809495513194e-05,
+      "loss": 3.9731,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1624439926806403e-05,
+      "loss": 3.9675,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1616053979295883e-05,
+      "loss": 3.9685,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1607668031785363e-05,
+      "loss": 3.9593,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159928208427484e-05,
+      "loss": 3.9716,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159089613676432e-05,
+      "loss": 3.9625,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.15825101892538e-05,
+      "loss": 3.9675,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157414062054702e-05,
+      "loss": 3.9663,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.15657546730365e-05,
+      "loss": 3.9704,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.155736872552598e-05,
+      "loss": 3.9633,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154898277801545e-05,
+      "loss": 3.9769,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154061320930867e-05,
+      "loss": 3.9794,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.153224364060188e-05,
+      "loss": 3.9801,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.152385769309136e-05,
+      "loss": 3.9782,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.151547174558084e-05,
+      "loss": 3.9674,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1507102176874046e-05,
+      "loss": 3.9705,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1498716229363526e-05,
+      "loss": 3.9722,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1490330281853006e-05,
+      "loss": 3.9553,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1481944334342486e-05,
+      "loss": 3.9721,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.147355838683197e-05,
+      "loss": 3.9638,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.146517243932145e-05,
+      "loss": 3.9783,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1456786491810926e-05,
+      "loss": 3.9667,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1448400544300406e-05,
+      "loss": 3.9747,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1440014596789886e-05,
+      "loss": 3.9628,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1431628649279366e-05,
+      "loss": 3.965,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1423242701768846e-05,
+      "loss": 3.9762,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1414856754258326e-05,
+      "loss": 3.9732,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1406487185551535e-05,
+      "loss": 3.9807,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1398101238041015e-05,
+      "loss": 3.9659,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1389715290530495e-05,
+      "loss": 3.9793,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1381329343019975e-05,
+      "loss": 3.9721,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.137295977431319e-05,
+      "loss": 3.9732,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.136457382680267e-05,
+      "loss": 3.9782,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.135618787929215e-05,
+      "loss": 3.9694,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.134781831058536e-05,
+      "loss": 3.9697,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133943236307484e-05,
+      "loss": 3.978,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133104641556432e-05,
+      "loss": 3.9633,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.132267684685753e-05,
+      "loss": 3.9655,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.131429089934701e-05,
+      "loss": 3.9739,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.130590495183649e-05,
+      "loss": 3.9713,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.129751900432597e-05,
+      "loss": 3.9672,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128913305681545e-05,
+      "loss": 3.9718,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128074710930493e-05,
+      "loss": 3.9666,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.127236116179441e-05,
+      "loss": 3.9867,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1263991593087624e-05,
+      "loss": 3.9699,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.02175760269165,
+      "eval_runtime": 298.1156,
+      "eval_samples_per_second": 1280.01,
+      "eval_steps_per_second": 40.001,
+      "step": 1144788
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1255605645577104e-05,
+      "loss": 3.9626,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1247219698066584e-05,
+      "loss": 3.9656,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1238833750556064e-05,
+      "loss": 3.9597,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1230447803045544e-05,
+      "loss": 3.9801,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1222061855535024e-05,
+      "loss": 3.971,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1213675908024504e-05,
+      "loss": 3.972,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1205289960513984e-05,
+      "loss": 3.9665,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1196904013003464e-05,
+      "loss": 3.9675,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118851806549294e-05,
+      "loss": 3.9582,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118013211798242e-05,
+      "loss": 3.9666,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.11717461704719e-05,
+      "loss": 3.9672,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.116337660176511e-05,
+      "loss": 3.9837,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.115500703305833e-05,
+      "loss": 3.9774,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.114662108554781e-05,
+      "loss": 3.9493,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.113823513803729e-05,
+      "loss": 3.9665,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.112984919052676e-05,
+      "loss": 3.959,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.112146324301624e-05,
+      "loss": 3.9685,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.111307729550572e-05,
+      "loss": 3.966,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.11046913479952e-05,
+      "loss": 3.969,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.109630540048468e-05,
+      "loss": 3.9617,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.108791945297416e-05,
+      "loss": 3.9808,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107954988426737e-05,
+      "loss": 3.9702,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107116393675685e-05,
+      "loss": 3.968,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.106277798924633e-05,
+      "loss": 3.9659,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.105439204173581e-05,
+      "loss": 3.9784,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1046022473029026e-05,
+      "loss": 3.9648,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1037636525518506e-05,
+      "loss": 3.9568,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1029250578007986e-05,
+      "loss": 3.9674,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1020864630497466e-05,
+      "loss": 3.963,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1012478682986946e-05,
+      "loss": 3.9594,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1004125493083884e-05,
+      "loss": 3.9636,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0995739545573364e-05,
+      "loss": 3.9617,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0987353598062844e-05,
+      "loss": 3.9734,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0978967650552324e-05,
+      "loss": 3.9719,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0970581703041804e-05,
+      "loss": 3.9704,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.096221213433502e-05,
+      "loss": 3.9708,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09538261868245e-05,
+      "loss": 3.9713,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.094544023931398e-05,
+      "loss": 3.9552,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.093705429180346e-05,
+      "loss": 3.9626,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092868472309667e-05,
+      "loss": 3.9605,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092029877558615e-05,
+      "loss": 3.9555,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.091191282807563e-05,
+      "loss": 3.9635,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.090352688056511e-05,
+      "loss": 3.9659,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.089514093305459e-05,
+      "loss": 3.9579,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.088675498554407e-05,
+      "loss": 3.9714,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087836903803355e-05,
+      "loss": 3.9681,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.086998309052303e-05,
+      "loss": 3.9712,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.086161352181624e-05,
+      "loss": 3.9686,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.085322757430572e-05,
+      "loss": 3.9483,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.08448416267952e-05,
+      "loss": 3.9625,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0836455679284685e-05,
+      "loss": 3.9695,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0828069731774164e-05,
+      "loss": 3.9674,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0819683784263644e-05,
+      "loss": 3.9593,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0811297836753124e-05,
+      "loss": 3.9593,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.08029118892426e-05,
+      "loss": 3.9518,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.079452594173208e-05,
+      "loss": 3.9533,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0786156373025293e-05,
+      "loss": 3.9639,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.077777042551477e-05,
+      "loss": 3.9512,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0769384478004247e-05,
+      "loss": 3.9664,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0760998530493726e-05,
+      "loss": 3.9677,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.075262896178694e-05,
+      "loss": 3.9504,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.074424301427642e-05,
+      "loss": 3.9717,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.07358570667659e-05,
+      "loss": 3.9492,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.072747111925538e-05,
+      "loss": 3.9556,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071908517174486e-05,
+      "loss": 3.9539,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071073198184181e-05,
+      "loss": 3.9666,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.070234603433129e-05,
+      "loss": 3.9508,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.069396008682077e-05,
+      "loss": 3.966,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.068557413931025e-05,
+      "loss": 3.9422,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.067718819179972e-05,
+      "loss": 3.9536,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.06688022442892e-05,
+      "loss": 3.9686,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.066041629677868e-05,
+      "loss": 3.9613,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.065203034926816e-05,
+      "loss": 3.9633,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0643677159365105e-05,
+      "loss": 3.9681,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.063529121185459e-05,
+      "loss": 3.9754,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.062690526434407e-05,
+      "loss": 3.955,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0618519316833545e-05,
+      "loss": 3.9736,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0610133369323025e-05,
+      "loss": 3.955,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0601747421812505e-05,
+      "loss": 3.957,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0593361474301985e-05,
+      "loss": 3.9581,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0584975526791465e-05,
+      "loss": 3.9677,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0576605958084674e-05,
+      "loss": 3.9659,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0568220010574154e-05,
+      "loss": 3.9735,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0559834063063634e-05,
+      "loss": 3.9558,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055146449435684e-05,
+      "loss": 3.9461,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.054309492565006e-05,
+      "loss": 3.9521,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0534708978139546e-05,
+      "loss": 3.9639,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.052632303062902e-05,
+      "loss": 3.9585,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.05179370831185e-05,
+      "loss": 3.9577,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.050955113560798e-05,
+      "loss": 3.965,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.050116518809746e-05,
+      "loss": 3.955,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.049277924058694e-05,
+      "loss": 3.9588,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.048439329307642e-05,
+      "loss": 3.9561,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.04760073455659e-05,
+      "loss": 3.9678,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.046762139805538e-05,
+      "loss": 3.9729,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.045923545054486e-05,
+      "loss": 3.965,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0450849503034338e-05,
+      "loss": 3.9484,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.044247993432755e-05,
+      "loss": 3.9597,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.043409398681703e-05,
+      "loss": 3.9555,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.042570803930651e-05,
+      "loss": 3.9649,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.041732209179599e-05,
+      "loss": 3.9684,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.040893614428547e-05,
+      "loss": 3.9522,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.040055019677495e-05,
+      "loss": 3.9645,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0392180628068163e-05,
+      "loss": 3.9614,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0383794680557643e-05,
+      "loss": 3.9594,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0375408733047123e-05,
+      "loss": 3.9567,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0367022785536603e-05,
+      "loss": 3.9545,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0358653216829812e-05,
+      "loss": 3.9508,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0350283648123024e-05,
+      "loss": 3.9579,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0341897700612504e-05,
+      "loss": 3.9509,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0333511753101984e-05,
+      "loss": 3.9532,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0325125805591464e-05,
+      "loss": 3.9526,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0316739858080944e-05,
+      "loss": 3.9634,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0308370289374153e-05,
+      "loss": 3.9522,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0299984341863637e-05,
+      "loss": 3.9584,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0291598394353117e-05,
+      "loss": 3.9712,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0283212446842597e-05,
+      "loss": 3.9683,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0274826499332077e-05,
+      "loss": 3.9671,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.026644055182155e-05,
+      "loss": 3.9546,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.025805460431103e-05,
+      "loss": 3.9604,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0249668656800513e-05,
+      "loss": 3.9572,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.024129908809373e-05,
+      "loss": 3.9483,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0232913140583202e-05,
+      "loss": 3.9602,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0224527193072682e-05,
+      "loss": 3.9546,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0216141245562162e-05,
+      "loss": 3.9631,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0207771676855378e-05,
+      "loss": 3.9563,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0199385729344855e-05,
+      "loss": 3.9606,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0190999781834334e-05,
+      "loss": 3.9533,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0182613834323814e-05,
+      "loss": 3.9544,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0174227886813294e-05,
+      "loss": 3.9618,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0165858318106503e-05,
+      "loss": 3.9593,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0157472370595983e-05,
+      "loss": 3.967,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0149086423085467e-05,
+      "loss": 3.9585,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0140716854378676e-05,
+      "loss": 3.9659,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0132330906868156e-05,
+      "loss": 3.9634,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0123944959357636e-05,
+      "loss": 3.9638,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0115559011847116e-05,
+      "loss": 3.9629,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0107173064336596e-05,
+      "loss": 3.9601,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0098787116826076e-05,
+      "loss": 3.9584,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.009040116931556e-05,
+      "loss": 3.9662,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0082031600608768e-05,
+      "loss": 3.9556,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0073645653098248e-05,
+      "loss": 3.9519,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0065259705587728e-05,
+      "loss": 3.9621,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0056873758077208e-05,
+      "loss": 3.9599,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0048487810566688e-05,
+      "loss": 3.9573,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.00401182418599e-05,
+      "loss": 3.9642,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.003173229434938e-05,
+      "loss": 3.9552,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.002334634683886e-05,
+      "loss": 3.9694,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.001496039932834e-05,
+      "loss": 3.9587,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.016119956970215,
+      "eval_runtime": 297.6671,
+      "eval_samples_per_second": 1281.939,
+      "eval_steps_per_second": 40.062,
+      "step": 1221108
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.000657445181782e-05,
+      "loss": 3.9429,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.99981885043073e-05,
+      "loss": 3.9507,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.998980255679678e-05,
+      "loss": 3.9491,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.998141660928626e-05,
+      "loss": 3.9675,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9973030661775737e-05,
+      "loss": 3.9636,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9964644714265217e-05,
+      "loss": 3.9617,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9956258766754697e-05,
+      "loss": 3.9532,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9947872819244177e-05,
+      "loss": 3.9563,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9939486871733656e-05,
+      "loss": 3.9456,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9931100924223136e-05,
+      "loss": 3.9566,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9922714976712616e-05,
+      "loss": 3.9562,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.99143290292021e-05,
+      "loss": 3.9736,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.990595946049531e-05,
+      "loss": 3.9675,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.989757351298479e-05,
+      "loss": 3.9403,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.988918756547427e-05,
+      "loss": 3.9542,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.988080161796375e-05,
+      "loss": 3.9512,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.987241567045323e-05,
+      "loss": 3.9516,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.986402972294271e-05,
+      "loss": 3.959,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9855643775432192e-05,
+      "loss": 3.9583,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9847257827921672e-05,
+      "loss": 3.9493,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9838871880411152e-05,
+      "loss": 3.9707,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9830485932900625e-05,
+      "loss": 3.9598,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9822099985390105e-05,
+      "loss": 3.9538,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9813714037879585e-05,
+      "loss": 3.9512,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.98053444691728e-05,
+      "loss": 3.9712,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9796958521662277e-05,
+      "loss": 3.9534,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9788572574151757e-05,
+      "loss": 3.9496,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9780186626641237e-05,
+      "loss": 3.9551,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9771800679130717e-05,
+      "loss": 3.9514,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9763414731620197e-05,
+      "loss": 3.9507,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.975504516291341e-05,
+      "loss": 3.9544,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.974665921540289e-05,
+      "loss": 3.9536,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.973827326789237e-05,
+      "loss": 3.962,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.972988732038185e-05,
+      "loss": 3.9572,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.972151775167506e-05,
+      "loss": 3.9635,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9713148182968275e-05,
+      "loss": 3.9607,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.970476223545775e-05,
+      "loss": 3.9608,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.969637628794723e-05,
+      "loss": 3.9467,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.968799034043671e-05,
+      "loss": 3.9515,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.967960439292619e-05,
+      "loss": 3.9493,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.96712348242194e-05,
+      "loss": 3.9499,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9662848876708883e-05,
+      "loss": 3.9493,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9654462929198363e-05,
+      "loss": 3.9566,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9646093360491572e-05,
+      "loss": 3.9479,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9637707412981052e-05,
+      "loss": 3.9596,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9629321465470532e-05,
+      "loss": 3.9568,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9620935517960012e-05,
+      "loss": 3.961,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9612549570449492e-05,
+      "loss": 3.9553,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9604163622938976e-05,
+      "loss": 3.936,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9595777675428456e-05,
+      "loss": 3.9476,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9587391727917936e-05,
+      "loss": 3.9608,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9579005780407415e-05,
+      "loss": 3.9578,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9570619832896895e-05,
+      "loss": 3.9445,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9562233885386375e-05,
+      "loss": 3.9465,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9553847937875855e-05,
+      "loss": 3.9441,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9545461990365335e-05,
+      "loss": 3.9407,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9537076042854815e-05,
+      "loss": 3.9601,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9528706474148028e-05,
+      "loss": 3.939,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9520320526637508e-05,
+      "loss": 3.9541,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9511934579126988e-05,
+      "loss": 3.9567,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.950354863161646e-05,
+      "loss": 3.9385,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9495162684105944e-05,
+      "loss": 3.9593,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9486793115399157e-05,
+      "loss": 3.9405,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.947840716788864e-05,
+      "loss": 3.9468,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9470021220378113e-05,
+      "loss": 3.9383,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9461635272867593e-05,
+      "loss": 3.9589,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9453249325357073e-05,
+      "loss": 3.9436,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9444863377846553e-05,
+      "loss": 3.9499,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9436477430336033e-05,
+      "loss": 3.9356,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9428091482825516e-05,
+      "loss": 3.9414,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9419705535314996e-05,
+      "loss": 3.9574,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9411319587804476e-05,
+      "loss": 3.9572,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9402950019097685e-05,
+      "loss": 3.9488,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9394580450390894e-05,
+      "loss": 3.9548,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.938621088168411e-05,
+      "loss": 3.9656,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9377824934173587e-05,
+      "loss": 3.9456,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9369438986663067e-05,
+      "loss": 3.9636,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9361053039152547e-05,
+      "loss": 3.9424,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9352667091642027e-05,
+      "loss": 3.9486,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9344281144131507e-05,
+      "loss": 3.9477,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9335895196620987e-05,
+      "loss": 3.9561,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.932750924911047e-05,
+      "loss": 3.9527,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.931912330159995e-05,
+      "loss": 3.9654,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.931073735408943e-05,
+      "loss": 3.9478,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.930235140657891e-05,
+      "loss": 3.9322,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.929396545906839e-05,
+      "loss": 3.9421,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.92855958903616e-05,
+      "loss": 3.9525,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.927720994285108e-05,
+      "loss": 3.9466,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9268823995340562e-05,
+      "loss": 3.9472,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9260438047830042e-05,
+      "loss": 3.9531,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9252052100319522e-05,
+      "loss": 3.9453,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.924368253161273e-05,
+      "loss": 3.9478,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.923529658410221e-05,
+      "loss": 3.945,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.922691063659169e-05,
+      "loss": 3.9579,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.921852468908117e-05,
+      "loss": 3.9593,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9210138741570648e-05,
+      "loss": 3.9554,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9201785551667593e-05,
+      "loss": 3.9383,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9193399604157073e-05,
+      "loss": 3.9465,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9185013656646553e-05,
+      "loss": 3.9489,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9176627709136033e-05,
+      "loss": 3.9538,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9168241761625516e-05,
+      "loss": 3.9566,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9159855814114996e-05,
+      "loss": 3.9449,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9151486245408205e-05,
+      "loss": 3.9464,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9143100297897685e-05,
+      "loss": 3.9517,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9134714350387165e-05,
+      "loss": 3.9527,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9126328402876645e-05,
+      "loss": 3.9423,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.911794245536612e-05,
+      "loss": 3.9491,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.91095565078556e-05,
+      "loss": 3.9399,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.910117056034508e-05,
+      "loss": 3.9416,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.909278461283456e-05,
+      "loss": 3.9425,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.908441504412777e-05,
+      "loss": 3.9426,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9076029096617254e-05,
+      "loss": 3.9393,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.906765952791047e-05,
+      "loss": 3.9514,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.905927358039995e-05,
+      "loss": 3.9435,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9050887632889423e-05,
+      "loss": 3.947,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9042501685378903e-05,
+      "loss": 3.9592,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9034115737868383e-05,
+      "loss": 3.9607,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9025729790357863e-05,
+      "loss": 3.9548,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9017343842847346e-05,
+      "loss": 3.9462,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9008957895336826e-05,
+      "loss": 3.9479,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9000588326630035e-05,
+      "loss": 3.9464,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8992202379119515e-05,
+      "loss": 3.9403,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8983816431608995e-05,
+      "loss": 3.9458,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8975446862902207e-05,
+      "loss": 3.9432,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8967060915391687e-05,
+      "loss": 3.9494,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8958674967881167e-05,
+      "loss": 3.9457,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8950289020370647e-05,
+      "loss": 3.9499,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8941903072860127e-05,
+      "loss": 3.9468,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.893354988295707e-05,
+      "loss": 3.9449,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.892516393544655e-05,
+      "loss": 3.9415,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.891677798793603e-05,
+      "loss": 3.9573,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.890839204042551e-05,
+      "loss": 3.9511,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.890000609291499e-05,
+      "loss": 3.9527,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.889162014540447e-05,
+      "loss": 3.9518,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.888323419789395e-05,
+      "loss": 3.9533,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.887484825038343e-05,
+      "loss": 3.9553,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.886646230287291e-05,
+      "loss": 3.9525,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.885807635536239e-05,
+      "loss": 3.9453,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8849690407851872e-05,
+      "loss": 3.9493,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8841304460341352e-05,
+      "loss": 3.9551,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.883291851283083e-05,
+      "loss": 3.9451,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.882454894412404e-05,
+      "loss": 3.942,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.881617937541725e-05,
+      "loss": 3.9467,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8807793427906733e-05,
+      "loss": 3.9539,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8799407480396213e-05,
+      "loss": 3.9419,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8791021532885693e-05,
+      "loss": 3.9545,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8782635585375173e-05,
+      "loss": 3.9443,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8774266016668382e-05,
+      "loss": 3.9566,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8765880069157862e-05,
+      "loss": 3.9532,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8757494121647342e-05,
+      "loss": 3.9389,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.010923862457275,
+      "eval_runtime": 303.0168,
+      "eval_samples_per_second": 1259.306,
+      "eval_steps_per_second": 39.354,
+      "step": 1297428
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8749108174136825e-05,
+      "loss": 3.9395,
+      "step": 1297920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8740722226626305e-05,
+      "loss": 3.9385,
+      "step": 1298432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8732352657919514e-05,
+      "loss": 3.9564,
+      "step": 1298944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8723966710408994e-05,
+      "loss": 3.9475,
+      "step": 1299456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8715580762898474e-05,
+      "loss": 3.9494,
+      "step": 1299968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8707194815387954e-05,
+      "loss": 3.9477,
+      "step": 1300480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.869880886787743e-05,
+      "loss": 3.9415,
+      "step": 1300992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.869042292036691e-05,
+      "loss": 3.937,
+      "step": 1301504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.868203697285639e-05,
+      "loss": 3.9503,
+      "step": 1302016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.867365102534587e-05,
+      "loss": 3.9371,
+      "step": 1302528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.866528145663908e-05,
+      "loss": 3.9656,
+      "step": 1303040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8656895509128563e-05,
+      "loss": 3.9543,
+      "step": 1303552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.864852594042178e-05,
+      "loss": 3.9351,
+      "step": 1304064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8640139992911252e-05,
+      "loss": 3.9382,
+      "step": 1304576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8631754045400732e-05,
+      "loss": 3.9415,
+      "step": 1305088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8623368097890212e-05,
+      "loss": 3.9388,
+      "step": 1305600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8614982150379692e-05,
+      "loss": 3.948,
+      "step": 1306112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8606596202869172e-05,
+      "loss": 3.9507,
+      "step": 1306624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8598210255358655e-05,
+      "loss": 3.9366,
+      "step": 1307136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8589840686651865e-05,
+      "loss": 3.9633,
+      "step": 1307648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8581454739141344e-05,
+      "loss": 3.9445,
+      "step": 1308160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8573068791630824e-05,
+      "loss": 3.9493,
+      "step": 1308672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8564682844120304e-05,
+      "loss": 3.9422,
+      "step": 1309184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8556296896609784e-05,
+      "loss": 3.9606,
+      "step": 1309696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8547910949099264e-05,
+      "loss": 3.9413,
+      "step": 1310208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8539525001588748e-05,
+      "loss": 3.9435,
+      "step": 1310720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8531139054078228e-05,
+      "loss": 3.9409,
+      "step": 1311232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8522769485371437e-05,
+      "loss": 3.9418,
+      "step": 1311744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8514383537860917e-05,
+      "loss": 3.937,
+      "step": 1312256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8505997590350397e-05,
+      "loss": 3.9422,
+      "step": 1312768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8497611642839877e-05,
+      "loss": 3.9452,
+      "step": 1313280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.848924207413309e-05,
+      "loss": 3.9459,
+      "step": 1313792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.848085612662257e-05,
+      "loss": 3.9477,
+      "step": 1314304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.847247017911205e-05,
+      "loss": 3.9524,
+      "step": 1314816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8464100610405258e-05,
+      "loss": 3.9479,
+      "step": 1315328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8455714662894738e-05,
+      "loss": 3.9479,
+      "step": 1315840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8447328715384218e-05,
+      "loss": 3.9383,
+      "step": 1316352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.84389427678737e-05,
+      "loss": 3.9443,
+      "step": 1316864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.843055682036318e-05,
+      "loss": 3.9353,
+      "step": 1317376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.842217087285266e-05,
+      "loss": 3.9418,
+      "step": 1317888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.841378492534214e-05,
+      "loss": 3.9388,
+      "step": 1318400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8405398977831614e-05,
+      "loss": 3.9452,
+      "step": 1318912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8397013030321094e-05,
+      "loss": 3.9378,
+      "step": 1319424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.838864346161431e-05,
+      "loss": 3.9512,
+      "step": 1319936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8380257514103794e-05,
+      "loss": 3.9445,
+      "step": 1320448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8371871566593267e-05,
+      "loss": 3.9499,
+      "step": 1320960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8363485619082747e-05,
+      "loss": 3.9454,
+      "step": 1321472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8355099671572227e-05,
+      "loss": 3.9304,
+      "step": 1321984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8346713724061707e-05,
+      "loss": 3.9333,
+      "step": 1322496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.833834415535492e-05,
+      "loss": 3.954,
+      "step": 1323008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.83299582078444e-05,
+      "loss": 3.9432,
+      "step": 1323520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.832157226033388e-05,
+      "loss": 3.9361,
+      "step": 1324032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.831318631282336e-05,
+      "loss": 3.9323,
+      "step": 1324544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8304816744116568e-05,
+      "loss": 3.9411,
+      "step": 1325056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8296430796606048e-05,
+      "loss": 3.9232,
+      "step": 1325568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.828804484909553e-05,
+      "loss": 3.9534,
+      "step": 1326080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.827965890158501e-05,
+      "loss": 3.924,
+      "step": 1326592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.827127295407449e-05,
+      "loss": 3.9435,
+      "step": 1327104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.826288700656397e-05,
+      "loss": 3.9489,
+      "step": 1327616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.825451743785718e-05,
+      "loss": 3.9291,
+      "step": 1328128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.824613149034666e-05,
+      "loss": 3.9435,
+      "step": 1328640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.823774554283614e-05,
+      "loss": 3.9327,
+      "step": 1329152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.822935959532562e-05,
+      "loss": 3.9378,
+      "step": 1329664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8220973647815104e-05,
+      "loss": 3.9257,
+      "step": 1330176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8212604079108313e-05,
+      "loss": 3.9508,
+      "step": 1330688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8204218131597793e-05,
+      "loss": 3.9322,
+      "step": 1331200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8195832184087273e-05,
+      "loss": 3.9392,
+      "step": 1331712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8187446236576752e-05,
+      "loss": 3.9228,
+      "step": 1332224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8179060289066232e-05,
+      "loss": 3.9314,
+      "step": 1332736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8170674341555712e-05,
+      "loss": 3.9441,
+      "step": 1333248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8162288394045196e-05,
+      "loss": 3.9461,
+      "step": 1333760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8153902446534676e-05,
+      "loss": 3.9426,
+      "step": 1334272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8145516499024156e-05,
+      "loss": 3.9456,
+      "step": 1334784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8137146930317365e-05,
+      "loss": 3.9518,
+      "step": 1335296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8128760982806845e-05,
+      "loss": 3.9394,
+      "step": 1335808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8120375035296325e-05,
+      "loss": 3.9489,
+      "step": 1336320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.81119890877858e-05,
+      "loss": 3.9369,
+      "step": 1336832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8103619519079017e-05,
+      "loss": 3.9402,
+      "step": 1337344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8095233571568497e-05,
+      "loss": 3.9323,
+      "step": 1337856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8086847624057977e-05,
+      "loss": 3.9455,
+      "step": 1338368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.807846167654745e-05,
+      "loss": 3.944,
+      "step": 1338880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8070092107840666e-05,
+      "loss": 3.9543,
+      "step": 1339392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.806170616033015e-05,
+      "loss": 3.9442,
+      "step": 1339904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.805333659162336e-05,
+      "loss": 3.9204,
+      "step": 1340416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.804495064411284e-05,
+      "loss": 3.9295,
+      "step": 1340928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.803656469660232e-05,
+      "loss": 3.9414,
+      "step": 1341440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.80281787490918e-05,
+      "loss": 3.9392,
+      "step": 1341952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8019792801581275e-05,
+      "loss": 3.9345,
+      "step": 1342464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8011406854070755e-05,
+      "loss": 3.9463,
+      "step": 1342976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8003020906560235e-05,
+      "loss": 3.9325,
+      "step": 1343488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7994634959049715e-05,
+      "loss": 3.9374,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7986265390342924e-05,
+      "loss": 3.9393,
+      "step": 1344512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7977879442832404e-05,
+      "loss": 3.9467,
+      "step": 1345024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7969493495321887e-05,
+      "loss": 3.948,
+      "step": 1345536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7961107547811367e-05,
+      "loss": 3.95,
+      "step": 1346048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7952737979104576e-05,
+      "loss": 3.9247,
+      "step": 1346560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7944352031594056e-05,
+      "loss": 3.9398,
+      "step": 1347072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7935966084083536e-05,
+      "loss": 3.9367,
+      "step": 1347584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7927580136573016e-05,
+      "loss": 3.9418,
+      "step": 1348096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7919194189062496e-05,
+      "loss": 3.9515,
+      "step": 1348608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.791082462035571e-05,
+      "loss": 3.9287,
+      "step": 1349120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.790243867284519e-05,
+      "loss": 3.9387,
+      "step": 1349632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7894069104138398e-05,
+      "loss": 3.9437,
+      "step": 1350144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7885683156627878e-05,
+      "loss": 3.9437,
+      "step": 1350656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7877297209117358e-05,
+      "loss": 3.93,
+      "step": 1351168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.786891126160684e-05,
+      "loss": 3.939,
+      "step": 1351680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.786052531409632e-05,
+      "loss": 3.9329,
+      "step": 1352192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.78521393665858e-05,
+      "loss": 3.9311,
+      "step": 1352704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.784375341907528e-05,
+      "loss": 3.931,
+      "step": 1353216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.783538385036849e-05,
+      "loss": 3.9341,
+      "step": 1353728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.782699790285797e-05,
+      "loss": 3.9315,
+      "step": 1354240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.781861195534745e-05,
+      "loss": 3.9395,
+      "step": 1354752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7810226007836933e-05,
+      "loss": 3.9337,
+      "step": 1355264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7801840060326413e-05,
+      "loss": 3.939,
+      "step": 1355776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7793454112815893e-05,
+      "loss": 3.9531,
+      "step": 1356288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7785068165305373e-05,
+      "loss": 3.947,
+      "step": 1356800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7776682217794853e-05,
+      "loss": 3.9421,
+      "step": 1357312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7768296270284333e-05,
+      "loss": 3.9381,
+      "step": 1357824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7759926701577542e-05,
+      "loss": 3.9379,
+      "step": 1358336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7751540754067022e-05,
+      "loss": 3.9378,
+      "step": 1358848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7743171185360234e-05,
+      "loss": 3.9277,
+      "step": 1359360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7734785237849714e-05,
+      "loss": 3.9366,
+      "step": 1359872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7726399290339194e-05,
+      "loss": 3.9377,
+      "step": 1360384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7718013342828674e-05,
+      "loss": 3.94,
+      "step": 1360896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7709643774121887e-05,
+      "loss": 3.9368,
+      "step": 1361408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7701257826611367e-05,
+      "loss": 3.9406,
+      "step": 1361920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7692871879100847e-05,
+      "loss": 3.9365,
+      "step": 1362432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7684485931590327e-05,
+      "loss": 3.9366,
+      "step": 1362944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7676099984079807e-05,
+      "loss": 3.9291,
+      "step": 1363456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7667730415373016e-05,
+      "loss": 3.9503,
+      "step": 1363968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7659344467862496e-05,
+      "loss": 3.9444,
+      "step": 1364480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7650958520351976e-05,
+      "loss": 3.9435,
+      "step": 1364992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.764257257284146e-05,
+      "loss": 3.9397,
+      "step": 1365504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.763418662533094e-05,
+      "loss": 3.9416,
+      "step": 1366016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7625800677820412e-05,
+      "loss": 3.9437,
+      "step": 1366528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7617414730309892e-05,
+      "loss": 3.9489,
+      "step": 1367040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7609028782799372e-05,
+      "loss": 3.9376,
+      "step": 1367552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7600642835288852e-05,
+      "loss": 3.9396,
+      "step": 1368064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.75922896453858e-05,
+      "loss": 3.9398,
+      "step": 1368576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.758390369787528e-05,
+      "loss": 3.9354,
+      "step": 1369088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.757551775036476e-05,
+      "loss": 3.935,
+      "step": 1369600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7567131802854233e-05,
+      "loss": 3.9391,
+      "step": 1370112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7558745855343713e-05,
+      "loss": 3.9427,
+      "step": 1370624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.755037628663693e-05,
+      "loss": 3.931,
+      "step": 1371136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7541990339126406e-05,
+      "loss": 3.9471,
+      "step": 1371648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7533604391615886e-05,
+      "loss": 3.936,
+      "step": 1372160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.75252348229091e-05,
+      "loss": 3.9429,
+      "step": 1372672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.751684887539858e-05,
+      "loss": 3.942,
+      "step": 1373184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7508462927888058e-05,
+      "loss": 3.931,
+      "step": 1373696
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0059123039245605,
+      "eval_runtime": 292.5365,
+      "eval_samples_per_second": 1304.422,
+      "eval_steps_per_second": 40.764,
+      "step": 1373748
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7500076980377538e-05,
+      "loss": 3.9292,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7491691032867018e-05,
+      "loss": 3.9309,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7483305085356498e-05,
+      "loss": 3.9495,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7474919137845978e-05,
+      "loss": 3.9369,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7466533190335458e-05,
+      "loss": 3.9448,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7458147242824938e-05,
+      "loss": 3.9352,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7449761295314418e-05,
+      "loss": 3.9351,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7441375347803898e-05,
+      "loss": 3.9251,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.743298940029338e-05,
+      "loss": 3.9389,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.742460345278286e-05,
+      "loss": 3.9333,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.741621750527234e-05,
+      "loss": 3.9548,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.740783155776182e-05,
+      "loss": 3.9428,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.739946198905503e-05,
+      "loss": 3.9252,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.739107604154451e-05,
+      "loss": 3.9302,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.738269009403399e-05,
+      "loss": 3.9404,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7374304146523473e-05,
+      "loss": 3.9179,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7365918199012947e-05,
+      "loss": 3.9397,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7357532251502426e-05,
+      "loss": 3.9374,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7349146303991906e-05,
+      "loss": 3.9303,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7340760356481386e-05,
+      "loss": 3.9557,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7332374408970866e-05,
+      "loss": 3.9342,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.732398846146035e-05,
+      "loss": 3.9406,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.731560251394983e-05,
+      "loss": 3.9335,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.730723294524304e-05,
+      "loss": 3.95,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.729884699773252e-05,
+      "loss": 3.9256,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7290461050222e-05,
+      "loss": 3.937,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.728207510271148e-05,
+      "loss": 3.9274,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.727370553400469e-05,
+      "loss": 3.9356,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.726531958649417e-05,
+      "loss": 3.9324,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.725693363898365e-05,
+      "loss": 3.9299,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.724854769147313e-05,
+      "loss": 3.9356,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.724016174396261e-05,
+      "loss": 3.9396,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.723179217525582e-05,
+      "loss": 3.9331,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7223422606549033e-05,
+      "loss": 3.9466,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.721505303784224e-05,
+      "loss": 3.9383,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.720666709033172e-05,
+      "loss": 3.9363,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.71982811428212e-05,
+      "loss": 3.9307,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.718989519531068e-05,
+      "loss": 3.9333,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7181509247800165e-05,
+      "loss": 3.9265,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7173123300289645e-05,
+      "loss": 3.9332,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7164737352779125e-05,
+      "loss": 3.9287,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7156351405268605e-05,
+      "loss": 3.9395,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7147965457758085e-05,
+      "loss": 3.9271,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7139595889051294e-05,
+      "loss": 3.9374,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7131209941540774e-05,
+      "loss": 3.9393,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7122823994030254e-05,
+      "loss": 3.9381,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7114438046519737e-05,
+      "loss": 3.9379,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7106052099009217e-05,
+      "loss": 3.918,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7097666151498697e-05,
+      "loss": 3.9245,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7089280203988177e-05,
+      "loss": 3.9417,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7080910635281386e-05,
+      "loss": 3.9325,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7072524687770866e-05,
+      "loss": 3.9319,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7064138740260346e-05,
+      "loss": 3.9213,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.705575279274983e-05,
+      "loss": 3.9327,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.704738322404304e-05,
+      "loss": 3.9112,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7038997276532518e-05,
+      "loss": 3.946,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7030611329021998e-05,
+      "loss": 3.9123,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7022225381511478e-05,
+      "loss": 3.9397,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7013839434000955e-05,
+      "loss": 3.9333,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.700546986529417e-05,
+      "loss": 3.9217,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.699708391778365e-05,
+      "loss": 3.9358,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698869797027313e-05,
+      "loss": 3.9242,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6980312022762604e-05,
+      "loss": 3.924,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6971926075252084e-05,
+      "loss": 3.9163,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6963540127741567e-05,
+      "loss": 3.9417,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6955154180231047e-05,
+      "loss": 3.9233,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6946784611524256e-05,
+      "loss": 3.9298,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6938398664013736e-05,
+      "loss": 3.9127,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6930012716503216e-05,
+      "loss": 3.9228,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6921626768992696e-05,
+      "loss": 3.9331,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6913240821482176e-05,
+      "loss": 3.9383,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.690485487397166e-05,
+      "loss": 3.9279,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.689646892646114e-05,
+      "loss": 3.9401,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.688808297895062e-05,
+      "loss": 3.9395,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6879713410243828e-05,
+      "loss": 3.932,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6871343841537037e-05,
+      "loss": 3.936,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.686295789402652e-05,
+      "loss": 3.93,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6854571946516e-05,
+      "loss": 3.928,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.684618599900548e-05,
+      "loss": 3.9254,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.683780005149496e-05,
+      "loss": 3.9349,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.682941410398444e-05,
+      "loss": 3.9374,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.682102815647392e-05,
+      "loss": 3.9435,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.68126422089634e-05,
+      "loss": 3.9309,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6804272640256613e-05,
+      "loss": 3.9125,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6795886692746093e-05,
+      "loss": 3.9191,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6787517124039302e-05,
+      "loss": 3.932,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6779131176528782e-05,
+      "loss": 3.9277,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6770745229018262e-05,
+      "loss": 3.9268,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6762359281507742e-05,
+      "loss": 3.9356,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6753973333997222e-05,
+      "loss": 3.9255,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6745587386486705e-05,
+      "loss": 3.9204,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6737201438976185e-05,
+      "loss": 3.9277,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6728815491465665e-05,
+      "loss": 3.9402,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6720445922758874e-05,
+      "loss": 3.9338,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6712059975248354e-05,
+      "loss": 3.9387,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6703674027737834e-05,
+      "loss": 3.9209,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6695304459031047e-05,
+      "loss": 3.929,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6686918511520526e-05,
+      "loss": 3.9303,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6678532564010006e-05,
+      "loss": 3.9321,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6670146616499486e-05,
+      "loss": 3.94,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6661760668988966e-05,
+      "loss": 3.9231,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6653374721478443e-05,
+      "loss": 3.9282,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6644988773967923e-05,
+      "loss": 3.9316,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6636602826457403e-05,
+      "loss": 3.9354,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6628216878946883e-05,
+      "loss": 3.9187,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6619847310240092e-05,
+      "loss": 3.9279,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6611461362729572e-05,
+      "loss": 3.9268,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.660310817282652e-05,
+      "loss": 3.9249,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6594722225316e-05,
+      "loss": 3.9203,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.658633627780548e-05,
+      "loss": 3.9231,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.657795033029496e-05,
+      "loss": 3.9238,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.656956438278444e-05,
+      "loss": 3.9342,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6561178435273913e-05,
+      "loss": 3.9194,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6552792487763397e-05,
+      "loss": 3.9305,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6544406540252877e-05,
+      "loss": 3.9411,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6536020592742356e-05,
+      "loss": 3.9398,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6527651024035566e-05,
+      "loss": 3.9326,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6519265076525046e-05,
+      "loss": 3.9311,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6510879129014525e-05,
+      "loss": 3.9273,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6502493181504005e-05,
+      "loss": 3.9259,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6494107233993485e-05,
+      "loss": 3.9223,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.648572128648297e-05,
+      "loss": 3.9222,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.647733533897245e-05,
+      "loss": 3.9246,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.646894939146193e-05,
+      "loss": 3.9356,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.646056344395141e-05,
+      "loss": 3.9263,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6452193875244618e-05,
+      "loss": 3.9297,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.644382430653783e-05,
+      "loss": 3.9281,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.643543835902731e-05,
+      "loss": 3.9253,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.642705241151679e-05,
+      "loss": 3.919,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.641866646400627e-05,
+      "loss": 3.9404,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.641028051649575e-05,
+      "loss": 3.9369,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.640189456898523e-05,
+      "loss": 3.935,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.639350862147471e-05,
+      "loss": 3.9297,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6385139052767922e-05,
+      "loss": 3.9287,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6376753105257402e-05,
+      "loss": 3.9352,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6368367157746882e-05,
+      "loss": 3.9459,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6359981210236362e-05,
+      "loss": 3.928,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6351595262725842e-05,
+      "loss": 3.9272,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6343209315215322e-05,
+      "loss": 3.9279,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6334823367704802e-05,
+      "loss": 3.9286,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.632643742019428e-05,
+      "loss": 3.9255,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6318067851487495e-05,
+      "loss": 3.9308,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6309681903976975e-05,
+      "loss": 3.9352,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6301295956466448e-05,
+      "loss": 3.9196,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6292910008955928e-05,
+      "loss": 3.9354,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6284540440249144e-05,
+      "loss": 3.9318,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6276154492738624e-05,
+      "loss": 3.9289,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6267784924031836e-05,
+      "loss": 3.9336,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6259398976521316e-05,
+      "loss": 3.9193,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.001258850097656,
+      "eval_runtime": 296.7652,
+      "eval_samples_per_second": 1285.835,
+      "eval_steps_per_second": 40.183,
+      "step": 1450068
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6251013029010796e-05,
+      "loss": 3.9262,
+      "step": 1450496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6242627081500276e-05,
+      "loss": 3.9181,
+      "step": 1451008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6234241133989752e-05,
+      "loss": 3.9403,
+      "step": 1451520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6225855186479232e-05,
+      "loss": 3.9226,
+      "step": 1452032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6217469238968712e-05,
+      "loss": 3.9402,
+      "step": 1452544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6209083291458192e-05,
+      "loss": 3.9244,
+      "step": 1453056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6200697343947672e-05,
+      "loss": 3.9278,
+      "step": 1453568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6192311396437152e-05,
+      "loss": 3.917,
+      "step": 1454080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6183925448926632e-05,
+      "loss": 3.9322,
+      "step": 1454592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6175539501416112e-05,
+      "loss": 3.9204,
+      "step": 1455104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6167169932709325e-05,
+      "loss": 3.9448,
+      "step": 1455616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6158800364002534e-05,
+      "loss": 3.9322,
+      "step": 1456128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.615043079529575e-05,
+      "loss": 3.9157,
+      "step": 1456640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6142044847785223e-05,
+      "loss": 3.9219,
+      "step": 1457152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6133658900274706e-05,
+      "loss": 3.9317,
+      "step": 1457664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6125272952764186e-05,
+      "loss": 3.9144,
+      "step": 1458176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6116887005253666e-05,
+      "loss": 3.9279,
+      "step": 1458688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6108501057743146e-05,
+      "loss": 3.9276,
+      "step": 1459200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6100115110232626e-05,
+      "loss": 3.921,
+      "step": 1459712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6091729162722106e-05,
+      "loss": 3.9465,
+      "step": 1460224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6083343215211586e-05,
+      "loss": 3.9284,
+      "step": 1460736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6074957267701066e-05,
+      "loss": 3.9275,
+      "step": 1461248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6066571320190546e-05,
+      "loss": 3.9237,
+      "step": 1461760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6058185372680026e-05,
+      "loss": 3.9413,
+      "step": 1462272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6049832182776967e-05,
+      "loss": 3.9176,
+      "step": 1462784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6041446235266447e-05,
+      "loss": 3.9252,
+      "step": 1463296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6033060287755927e-05,
+      "loss": 3.9212,
+      "step": 1463808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6024674340245407e-05,
+      "loss": 3.9279,
+      "step": 1464320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6016288392734887e-05,
+      "loss": 3.9202,
+      "step": 1464832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.600790244522437e-05,
+      "loss": 3.9208,
+      "step": 1465344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.599951649771385e-05,
+      "loss": 3.9292,
+      "step": 1465856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.599113055020333e-05,
+      "loss": 3.9282,
+      "step": 1466368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.598274460269281e-05,
+      "loss": 3.9271,
+      "step": 1466880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.597437503398602e-05,
+      "loss": 3.9362,
+      "step": 1467392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5966005465279232e-05,
+      "loss": 3.9288,
+      "step": 1467904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5957619517768712e-05,
+      "loss": 3.9288,
+      "step": 1468416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5949233570258192e-05,
+      "loss": 3.9171,
+      "step": 1468928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5940847622747672e-05,
+      "loss": 3.9295,
+      "step": 1469440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5932461675237152e-05,
+      "loss": 3.9161,
+      "step": 1469952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.592409210653036e-05,
+      "loss": 3.923,
+      "step": 1470464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.591570615901984e-05,
+      "loss": 3.9212,
+      "step": 1470976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5907320211509324e-05,
+      "loss": 3.9288,
+      "step": 1471488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5898934263998804e-05,
+      "loss": 3.9159,
+      "step": 1472000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5890548316488284e-05,
+      "loss": 3.9274,
+      "step": 1472512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5882162368977757e-05,
+      "loss": 3.9328,
+      "step": 1473024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5873776421467237e-05,
+      "loss": 3.9311,
+      "step": 1473536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5865390473956717e-05,
+      "loss": 3.9337,
+      "step": 1474048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5857020905249933e-05,
+      "loss": 3.9091,
+      "step": 1474560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.584863495773941e-05,
+      "loss": 3.9153,
+      "step": 1475072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.584024901022889e-05,
+      "loss": 3.9302,
+      "step": 1475584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.583186306271837e-05,
+      "loss": 3.9252,
+      "step": 1476096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.582349349401158e-05,
+      "loss": 3.9226,
+      "step": 1476608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5815107546501062e-05,
+      "loss": 3.9138,
+      "step": 1477120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5806721598990542e-05,
+      "loss": 3.9264,
+      "step": 1477632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5798335651480022e-05,
+      "loss": 3.9045,
+      "step": 1478144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5789982461576967e-05,
+      "loss": 3.9311,
+      "step": 1478656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5781596514066447e-05,
+      "loss": 3.9108,
+      "step": 1479168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5773210566555927e-05,
+      "loss": 3.9238,
+      "step": 1479680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5764824619045407e-05,
+      "loss": 3.9286,
+      "step": 1480192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5756438671534883e-05,
+      "loss": 3.914,
+      "step": 1480704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5748052724024363e-05,
+      "loss": 3.9268,
+      "step": 1481216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5739666776513843e-05,
+      "loss": 3.9191,
+      "step": 1481728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5731280829003323e-05,
+      "loss": 3.9157,
+      "step": 1482240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5722894881492803e-05,
+      "loss": 3.9093,
+      "step": 1482752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5714508933982283e-05,
+      "loss": 3.9301,
+      "step": 1483264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5706122986471763e-05,
+      "loss": 3.9179,
+      "step": 1483776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5697753417764976e-05,
+      "loss": 3.922,
+      "step": 1484288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5689367470254455e-05,
+      "loss": 3.9066,
+      "step": 1484800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5680981522743935e-05,
+      "loss": 3.9141,
+      "step": 1485312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5672595575233415e-05,
+      "loss": 3.9241,
+      "step": 1485824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5664209627722895e-05,
+      "loss": 3.9323,
+      "step": 1486336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5655823680212375e-05,
+      "loss": 3.9166,
+      "step": 1486848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5647437732701855e-05,
+      "loss": 3.935,
+      "step": 1487360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.563905178519134e-05,
+      "loss": 3.9311,
+      "step": 1487872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5630682216484548e-05,
+      "loss": 3.9207,
+      "step": 1488384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5622312647777757e-05,
+      "loss": 3.9255,
+      "step": 1488896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5613926700267237e-05,
+      "loss": 3.9227,
+      "step": 1489408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5605540752756717e-05,
+      "loss": 3.922,
+      "step": 1489920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.55971548052462e-05,
+      "loss": 3.9174,
+      "step": 1490432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.558876885773568e-05,
+      "loss": 3.925,
+      "step": 1490944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.558038291022516e-05,
+      "loss": 3.9323,
+      "step": 1491456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.557199696271464e-05,
+      "loss": 3.9352,
+      "step": 1491968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.556362739400785e-05,
+      "loss": 3.9228,
+      "step": 1492480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.555524144649733e-05,
+      "loss": 3.9056,
+      "step": 1492992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.554685549898681e-05,
+      "loss": 3.9146,
+      "step": 1493504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5538469551476292e-05,
+      "loss": 3.92,
+      "step": 1494016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5530083603965772e-05,
+      "loss": 3.9251,
+      "step": 1494528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5521697656455245e-05,
+      "loss": 3.9135,
+      "step": 1495040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5513311708944725e-05,
+      "loss": 3.9251,
+      "step": 1495552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5504925761434205e-05,
+      "loss": 3.9176,
+      "step": 1496064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5496539813923685e-05,
+      "loss": 3.913,
+      "step": 1496576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5488170245216898e-05,
+      "loss": 3.9239,
+      "step": 1497088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5479784297706378e-05,
+      "loss": 3.9271,
+      "step": 1497600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5471414728999594e-05,
+      "loss": 3.9231,
+      "step": 1498112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5463028781489067e-05,
+      "loss": 3.9333,
+      "step": 1498624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5454642833978547e-05,
+      "loss": 3.9153,
+      "step": 1499136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.544625688646803e-05,
+      "loss": 3.9188,
+      "step": 1499648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.543787093895751e-05,
+      "loss": 3.9198,
+      "step": 1500160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.542948499144699e-05,
+      "loss": 3.9261,
+      "step": 1500672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.542109904393647e-05,
+      "loss": 3.9321,
+      "step": 1501184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.541272947522968e-05,
+      "loss": 3.914,
+      "step": 1501696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.540434352771916e-05,
+      "loss": 3.9208,
+      "step": 1502208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.539595758020864e-05,
+      "loss": 3.9245,
+      "step": 1502720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.538757163269812e-05,
+      "loss": 3.9258,
+      "step": 1503232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5379185685187602e-05,
+      "loss": 3.9099,
+      "step": 1503744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.537081611648081e-05,
+      "loss": 3.9221,
+      "step": 1504256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.536243016897029e-05,
+      "loss": 3.9217,
+      "step": 1504768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.535404422145977e-05,
+      "loss": 3.9158,
+      "step": 1505280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.534565827394925e-05,
+      "loss": 3.9119,
+      "step": 1505792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.533727232643873e-05,
+      "loss": 3.9157,
+      "step": 1506304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.532888637892821e-05,
+      "loss": 3.9218,
+      "step": 1506816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5320500431417694e-05,
+      "loss": 3.9198,
+      "step": 1507328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5312114483907174e-05,
+      "loss": 3.9136,
+      "step": 1507840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5303728536396654e-05,
+      "loss": 3.9184,
+      "step": 1508352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5295358967689863e-05,
+      "loss": 3.936,
+      "step": 1508864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5286973020179343e-05,
+      "loss": 3.9318,
+      "step": 1509376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5278587072668823e-05,
+      "loss": 3.9263,
+      "step": 1509888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5270201125158303e-05,
+      "loss": 3.9227,
+      "step": 1510400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5261831556451516e-05,
+      "loss": 3.9192,
+      "step": 1510912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5253445608940996e-05,
+      "loss": 3.9206,
+      "step": 1511424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5245059661430476e-05,
+      "loss": 3.9125,
+      "step": 1511936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5236673713919956e-05,
+      "loss": 3.9161,
+      "step": 1512448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5228304145213165e-05,
+      "loss": 3.9179,
+      "step": 1512960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5219918197702648e-05,
+      "loss": 3.9281,
+      "step": 1513472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5211532250192128e-05,
+      "loss": 3.9181,
+      "step": 1513984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.52031463026816e-05,
+      "loss": 3.9255,
+      "step": 1514496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.519476035517108e-05,
+      "loss": 3.9156,
+      "step": 1515008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.518637440766056e-05,
+      "loss": 3.9174,
+      "step": 1515520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.517798846015004e-05,
+      "loss": 3.9141,
+      "step": 1516032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5169602512639524e-05,
+      "loss": 3.9306,
+      "step": 1516544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5161232943932734e-05,
+      "loss": 3.9338,
+      "step": 1517056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5152846996422214e-05,
+      "loss": 3.9229,
+      "step": 1517568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5144461048911693e-05,
+      "loss": 3.9224,
+      "step": 1518080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5136075101401173e-05,
+      "loss": 3.9226,
+      "step": 1518592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5127705532694386e-05,
+      "loss": 3.9234,
+      "step": 1519104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5119319585183866e-05,
+      "loss": 3.9368,
+      "step": 1519616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5110933637673346e-05,
+      "loss": 3.921,
+      "step": 1520128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5102547690162826e-05,
+      "loss": 3.9208,
+      "step": 1520640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5094178121456035e-05,
+      "loss": 3.9168,
+      "step": 1521152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5085792173945515e-05,
+      "loss": 3.9242,
+      "step": 1521664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5077406226434995e-05,
+      "loss": 3.9158,
+      "step": 1522176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5069020278924478e-05,
+      "loss": 3.9215,
+      "step": 1522688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5060650710217687e-05,
+      "loss": 3.927,
+      "step": 1523200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5052264762707167e-05,
+      "loss": 3.9172,
+      "step": 1523712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5043878815196647e-05,
+      "loss": 3.9277,
+      "step": 1524224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5035492867686127e-05,
+      "loss": 3.9283,
+      "step": 1524736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5027106920175607e-05,
+      "loss": 3.9202,
+      "step": 1525248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.501873735146882e-05,
+      "loss": 3.9241,
+      "step": 1525760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.50103514039583e-05,
+      "loss": 3.9126,
+      "step": 1526272
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.997814893722534,
+      "eval_runtime": 293.7582,
+      "eval_samples_per_second": 1298.997,
+      "eval_steps_per_second": 40.595,
+      "step": 1526388
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.500196545644778e-05,
+      "loss": 3.9116,
+      "step": 1526784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.499357950893726e-05,
+      "loss": 3.9107,
+      "step": 1527296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.498519356142674e-05,
+      "loss": 3.9332,
+      "step": 1527808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.497680761391622e-05,
+      "loss": 3.9185,
+      "step": 1528320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.49684216664057e-05,
+      "loss": 3.9345,
+      "step": 1528832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.496003571889518e-05,
+      "loss": 3.9137,
+      "step": 1529344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.495164977138466e-05,
+      "loss": 3.9235,
+      "step": 1529856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.494326382387414e-05,
+      "loss": 3.9077,
+      "step": 1530368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.493487787636362e-05,
+      "loss": 3.9242,
+      "step": 1530880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.49264919288531e-05,
+      "loss": 3.9096,
+      "step": 1531392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.491810598134258e-05,
+      "loss": 3.9355,
+      "step": 1531904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4909736412635788e-05,
+      "loss": 3.9297,
+      "step": 1532416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.490135046512527e-05,
+      "loss": 3.9054,
+      "step": 1532928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.489296451761475e-05,
+      "loss": 3.9154,
+      "step": 1533440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.488457857010423e-05,
+      "loss": 3.9248,
+      "step": 1533952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4876192622593708e-05,
+      "loss": 3.91,
+      "step": 1534464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4867806675083188e-05,
+      "loss": 3.919,
+      "step": 1534976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4859420727572668e-05,
+      "loss": 3.919,
+      "step": 1535488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4851034780062148e-05,
+      "loss": 3.918,
+      "step": 1536000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4842648832551628e-05,
+      "loss": 3.9364,
+      "step": 1536512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.483426288504111e-05,
+      "loss": 3.9213,
+      "step": 1537024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.482587693753059e-05,
+      "loss": 3.9199,
+      "step": 1537536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4817490990020068e-05,
+      "loss": 3.9211,
+      "step": 1538048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4809105042509548e-05,
+      "loss": 3.9304,
+      "step": 1538560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4800719094999028e-05,
+      "loss": 3.9127,
+      "step": 1539072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4792333147488507e-05,
+      "loss": 3.923,
+      "step": 1539584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.478396357878172e-05,
+      "loss": 3.9104,
+      "step": 1540096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.47755776312712e-05,
+      "loss": 3.9175,
+      "step": 1540608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.476719168376068e-05,
+      "loss": 3.9125,
+      "step": 1541120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.475880573625016e-05,
+      "loss": 3.9163,
+      "step": 1541632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.475041978873964e-05,
+      "loss": 3.9151,
+      "step": 1542144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.474203384122912e-05,
+      "loss": 3.9268,
+      "step": 1542656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4733647893718596e-05,
+      "loss": 3.9219,
+      "step": 1543168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4725278325011812e-05,
+      "loss": 3.9304,
+      "step": 1543680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.471690875630502e-05,
+      "loss": 3.9172,
+      "step": 1544192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.47085228087945e-05,
+      "loss": 3.9229,
+      "step": 1544704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.470013686128398e-05,
+      "loss": 3.9103,
+      "step": 1545216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.469175091377346e-05,
+      "loss": 3.9219,
+      "step": 1545728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.468336496626294e-05,
+      "loss": 3.9103,
+      "step": 1546240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.467497901875242e-05,
+      "loss": 3.9168,
+      "step": 1546752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.46665930712419e-05,
+      "loss": 3.9101,
+      "step": 1547264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.465820712373138e-05,
+      "loss": 3.9255,
+      "step": 1547776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.464982117622086e-05,
+      "loss": 3.9131,
+      "step": 1548288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.464143522871034e-05,
+      "loss": 3.9138,
+      "step": 1548800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.463304928119982e-05,
+      "loss": 3.9296,
+      "step": 1549312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.46246633336893e-05,
+      "loss": 3.9217,
+      "step": 1549824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4616310143786242e-05,
+      "loss": 3.9237,
+      "step": 1550336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4607924196275722e-05,
+      "loss": 3.9068,
+      "step": 1550848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4599538248765202e-05,
+      "loss": 3.904,
+      "step": 1551360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4591152301254682e-05,
+      "loss": 3.9215,
+      "step": 1551872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4582766353744162e-05,
+      "loss": 3.9197,
+      "step": 1552384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4574396785037375e-05,
+      "loss": 3.9105,
+      "step": 1552896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4566010837526855e-05,
+      "loss": 3.9106,
+      "step": 1553408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4557624890016335e-05,
+      "loss": 3.9175,
+      "step": 1553920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4549238942505815e-05,
+      "loss": 3.8996,
+      "step": 1554432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4540852994995295e-05,
+      "loss": 3.9234,
+      "step": 1554944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4532467047484774e-05,
+      "loss": 3.9051,
+      "step": 1555456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.452408109997425e-05,
+      "loss": 3.911,
+      "step": 1555968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4515711531267467e-05,
+      "loss": 3.9223,
+      "step": 1556480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4507325583756947e-05,
+      "loss": 3.9063,
+      "step": 1556992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4498939636246423e-05,
+      "loss": 3.9159,
+      "step": 1557504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4490553688735903e-05,
+      "loss": 3.9134,
+      "step": 1558016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4482184120029116e-05,
+      "loss": 3.9061,
+      "step": 1558528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4473798172518596e-05,
+      "loss": 3.9088,
+      "step": 1559040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4465412225008076e-05,
+      "loss": 3.9179,
+      "step": 1559552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4457026277497556e-05,
+      "loss": 3.9098,
+      "step": 1560064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4448640329987036e-05,
+      "loss": 3.9125,
+      "step": 1560576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4440270761280248e-05,
+      "loss": 3.9049,
+      "step": 1561088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4431884813769725e-05,
+      "loss": 3.9035,
+      "step": 1561600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.442351524506294e-05,
+      "loss": 3.9147,
+      "step": 1562112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.441512929755242e-05,
+      "loss": 3.9271,
+      "step": 1562624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4406743350041897e-05,
+      "loss": 3.9084,
+      "step": 1563136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4398357402531377e-05,
+      "loss": 3.9277,
+      "step": 1563648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4389971455020857e-05,
+      "loss": 3.9221,
+      "step": 1564160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4381585507510337e-05,
+      "loss": 3.9129,
+      "step": 1564672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4373232317607282e-05,
+      "loss": 3.9152,
+      "step": 1565184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4364846370096762e-05,
+      "loss": 3.9216,
+      "step": 1565696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4356460422586242e-05,
+      "loss": 3.9121,
+      "step": 1566208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4348074475075722e-05,
+      "loss": 3.9112,
+      "step": 1566720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.43396885275652e-05,
+      "loss": 3.9175,
+      "step": 1567232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.433130258005468e-05,
+      "loss": 3.9202,
+      "step": 1567744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.432291663254416e-05,
+      "loss": 3.9214,
+      "step": 1568256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.431454706383737e-05,
+      "loss": 3.9178,
+      "step": 1568768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.430616111632685e-05,
+      "loss": 3.9027,
+      "step": 1569280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.429777516881633e-05,
+      "loss": 3.9013,
+      "step": 1569792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.428938922130581e-05,
+      "loss": 3.9161,
+      "step": 1570304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.428100327379529e-05,
+      "loss": 3.9164,
+      "step": 1570816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.427261732628477e-05,
+      "loss": 3.9111,
+      "step": 1571328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.426423137877425e-05,
+      "loss": 3.9167,
+      "step": 1571840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.425584543126373e-05,
+      "loss": 3.9121,
+      "step": 1572352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.424745948375321e-05,
+      "loss": 3.9077,
+      "step": 1572864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4239089915046423e-05,
+      "loss": 3.9142,
+      "step": 1573376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4230703967535903e-05,
+      "loss": 3.9156,
+      "step": 1573888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.422231802002538e-05,
+      "loss": 3.9209,
+      "step": 1574400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.421393207251486e-05,
+      "loss": 3.9206,
+      "step": 1574912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4205562503808075e-05,
+      "loss": 3.9183,
+      "step": 1575424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4197176556297552e-05,
+      "loss": 3.905,
+      "step": 1575936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4188790608787032e-05,
+      "loss": 3.9168,
+      "step": 1576448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4180404661276512e-05,
+      "loss": 3.9148,
+      "step": 1576960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4172018713765992e-05,
+      "loss": 3.9244,
+      "step": 1577472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4163632766255472e-05,
+      "loss": 3.9124,
+      "step": 1577984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4155246818744952e-05,
+      "loss": 3.9074,
+      "step": 1578496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4146860871234435e-05,
+      "loss": 3.9179,
+      "step": 1579008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4138491302527644e-05,
+      "loss": 3.9198,
+      "step": 1579520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4130105355017124e-05,
+      "loss": 3.9096,
+      "step": 1580032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4121719407506604e-05,
+      "loss": 3.9138,
+      "step": 1580544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4113349838799813e-05,
+      "loss": 3.9122,
+      "step": 1581056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4104963891289296e-05,
+      "loss": 3.9037,
+      "step": 1581568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4096577943778776e-05,
+      "loss": 3.904,
+      "step": 1582080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4088191996268256e-05,
+      "loss": 3.9137,
+      "step": 1582592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4079822427561466e-05,
+      "loss": 3.909,
+      "step": 1583104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4071452858854675e-05,
+      "loss": 3.9125,
+      "step": 1583616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4063066911344158e-05,
+      "loss": 3.9064,
+      "step": 1584128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4054680963833638e-05,
+      "loss": 3.9105,
+      "step": 1584640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4046295016323118e-05,
+      "loss": 3.9277,
+      "step": 1585152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4037925447616327e-05,
+      "loss": 3.9291,
+      "step": 1585664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4029539500105807e-05,
+      "loss": 3.9172,
+      "step": 1586176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4021153552595287e-05,
+      "loss": 3.9167,
+      "step": 1586688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4012767605084767e-05,
+      "loss": 3.9103,
+      "step": 1587200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.400438165757425e-05,
+      "loss": 3.9116,
+      "step": 1587712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.399599571006373e-05,
+      "loss": 3.9057,
+      "step": 1588224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3987609762553207e-05,
+      "loss": 3.9106,
+      "step": 1588736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3979223815042687e-05,
+      "loss": 3.9118,
+      "step": 1589248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3970837867532167e-05,
+      "loss": 3.9176,
+      "step": 1589760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3962451920021647e-05,
+      "loss": 3.9106,
+      "step": 1590272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3954065972511127e-05,
+      "loss": 3.9185,
+      "step": 1590784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3945680025000606e-05,
+      "loss": 3.9088,
+      "step": 1591296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.393729407749009e-05,
+      "loss": 3.9105,
+      "step": 1591808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.39289245087833e-05,
+      "loss": 3.9071,
+      "step": 1592320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.392053856127278e-05,
+      "loss": 3.9215,
+      "step": 1592832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3912168992565988e-05,
+      "loss": 3.9259,
+      "step": 1593344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3903783045055468e-05,
+      "loss": 3.9206,
+      "step": 1593856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.389539709754495e-05,
+      "loss": 3.9083,
+      "step": 1594368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.388701115003443e-05,
+      "loss": 3.9181,
+      "step": 1594880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.387862520252391e-05,
+      "loss": 3.9169,
+      "step": 1595392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3870239255013388e-05,
+      "loss": 3.927,
+      "step": 1595904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3861853307502868e-05,
+      "loss": 3.913,
+      "step": 1596416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3853467359992348e-05,
+      "loss": 3.913,
+      "step": 1596928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.384509779128556e-05,
+      "loss": 3.9143,
+      "step": 1597440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.383671184377504e-05,
+      "loss": 3.9196,
+      "step": 1597952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3828342275068253e-05,
+      "loss": 3.9082,
+      "step": 1598464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3819956327557733e-05,
+      "loss": 3.9131,
+      "step": 1598976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.381157038004721e-05,
+      "loss": 3.9185,
+      "step": 1599488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.380318443253669e-05,
+      "loss": 3.9091,
+      "step": 1600000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3794798485026172e-05,
+      "loss": 3.9188,
+      "step": 1600512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3786412537515652e-05,
+      "loss": 3.9204,
+      "step": 1601024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.377804296880886e-05,
+      "loss": 3.9139,
+      "step": 1601536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.376965702129834e-05,
+      "loss": 3.9131,
+      "step": 1602048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.376127107378782e-05,
+      "loss": 3.9121,
+      "step": 1602560
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.994936943054199,
+      "eval_runtime": 300.422,
+      "eval_samples_per_second": 1270.183,
+      "eval_steps_per_second": 39.694,
+      "step": 1602708
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.37528851262773e-05,
+      "loss": 3.9123,
+      "step": 1603072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.374449917876678e-05,
+      "loss": 3.9032,
+      "step": 1603584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3736129610059994e-05,
+      "loss": 3.9234,
+      "step": 1604096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3727760041353206e-05,
+      "loss": 3.9114,
+      "step": 1604608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3719374093842683e-05,
+      "loss": 3.9241,
+      "step": 1605120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3710988146332163e-05,
+      "loss": 3.9138,
+      "step": 1605632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3702618577625375e-05,
+      "loss": 3.9138,
+      "step": 1606144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.369423263011486e-05,
+      "loss": 3.9013,
+      "step": 1606656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3685846682604335e-05,
+      "loss": 3.9157,
+      "step": 1607168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3677460735093815e-05,
+      "loss": 3.9065,
+      "step": 1607680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3669074787583295e-05,
+      "loss": 3.9233,
+      "step": 1608192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3660688840072775e-05,
+      "loss": 3.9223,
+      "step": 1608704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3652302892562255e-05,
+      "loss": 3.8994,
+      "step": 1609216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3643916945051735e-05,
+      "loss": 3.9096,
+      "step": 1609728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3635530997541215e-05,
+      "loss": 3.9177,
+      "step": 1610240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3627145050030695e-05,
+      "loss": 3.9008,
+      "step": 1610752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3618759102520175e-05,
+      "loss": 3.9117,
+      "step": 1611264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3610373155009655e-05,
+      "loss": 3.9115,
+      "step": 1611776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3601987207499135e-05,
+      "loss": 3.9095,
+      "step": 1612288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3593601259988615e-05,
+      "loss": 3.9254,
+      "step": 1612800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3585215312478095e-05,
+      "loss": 3.9147,
+      "step": 1613312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3576829364967575e-05,
+      "loss": 3.9137,
+      "step": 1613824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3568459796260787e-05,
+      "loss": 3.9172,
+      "step": 1614336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3560073848750267e-05,
+      "loss": 3.9179,
+      "step": 1614848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3551704280043476e-05,
+      "loss": 3.9063,
+      "step": 1615360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3543318332532956e-05,
+      "loss": 3.9151,
+      "step": 1615872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3534932385022436e-05,
+      "loss": 3.9037,
+      "step": 1616384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3526546437511916e-05,
+      "loss": 3.91,
+      "step": 1616896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.351817686880513e-05,
+      "loss": 3.9049,
+      "step": 1617408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.350979092129461e-05,
+      "loss": 3.9117,
+      "step": 1617920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.350140497378409e-05,
+      "loss": 3.9071,
+      "step": 1618432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.349301902627357e-05,
+      "loss": 3.9152,
+      "step": 1618944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3484633078763045e-05,
+      "loss": 3.9162,
+      "step": 1619456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.347626351005626e-05,
+      "loss": 3.9182,
+      "step": 1619968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.346789394134947e-05,
+      "loss": 3.9123,
+      "step": 1620480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.345950799383895e-05,
+      "loss": 3.9149,
+      "step": 1620992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.345112204632843e-05,
+      "loss": 3.9055,
+      "step": 1621504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.344273609881791e-05,
+      "loss": 3.913,
+      "step": 1622016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.343435015130739e-05,
+      "loss": 3.9005,
+      "step": 1622528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.342596420379687e-05,
+      "loss": 3.9084,
+      "step": 1623040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.341757825628635e-05,
+      "loss": 3.9075,
+      "step": 1623552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.340919230877583e-05,
+      "loss": 3.9169,
+      "step": 1624064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3400822740069042e-05,
+      "loss": 3.9037,
+      "step": 1624576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.339243679255852e-05,
+      "loss": 3.9111,
+      "step": 1625088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3384050845048e-05,
+      "loss": 3.923,
+      "step": 1625600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3375664897537482e-05,
+      "loss": 3.9135,
+      "step": 1626112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3367278950026962e-05,
+      "loss": 3.9164,
+      "step": 1626624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3358893002516442e-05,
+      "loss": 3.9032,
+      "step": 1627136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3350507055005922e-05,
+      "loss": 3.8908,
+      "step": 1627648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3342121107495402e-05,
+      "loss": 3.9191,
+      "step": 1628160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.333375153878861e-05,
+      "loss": 3.9106,
+      "step": 1628672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3325381970081823e-05,
+      "loss": 3.903,
+      "step": 1629184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3316996022571303e-05,
+      "loss": 3.9064,
+      "step": 1629696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3308610075060783e-05,
+      "loss": 3.913,
+      "step": 1630208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3300224127550263e-05,
+      "loss": 3.8881,
+      "step": 1630720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3291838180039743e-05,
+      "loss": 3.918,
+      "step": 1631232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3283452232529223e-05,
+      "loss": 3.8955,
+      "step": 1631744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.32750662850187e-05,
+      "loss": 3.9084,
+      "step": 1632256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3266680337508183e-05,
+      "loss": 3.9124,
+      "step": 1632768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3258294389997663e-05,
+      "loss": 3.9065,
+      "step": 1633280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3249908442487143e-05,
+      "loss": 3.9043,
+      "step": 1633792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3241522494976623e-05,
+      "loss": 3.9106,
+      "step": 1634304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3233136547466103e-05,
+      "loss": 3.8967,
+      "step": 1634816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3224750599955583e-05,
+      "loss": 3.9038,
+      "step": 1635328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3216381031248792e-05,
+      "loss": 3.9096,
+      "step": 1635840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3208011462542004e-05,
+      "loss": 3.9068,
+      "step": 1636352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3199625515031484e-05,
+      "loss": 3.9065,
+      "step": 1636864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3191239567520964e-05,
+      "loss": 3.8954,
+      "step": 1637376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3182853620010444e-05,
+      "loss": 3.8967,
+      "step": 1637888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3174467672499924e-05,
+      "loss": 3.9068,
+      "step": 1638400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3166081724989404e-05,
+      "loss": 3.9188,
+      "step": 1638912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3157712156282617e-05,
+      "loss": 3.9028,
+      "step": 1639424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3149326208772097e-05,
+      "loss": 3.92,
+      "step": 1639936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3140940261261577e-05,
+      "loss": 3.917,
+      "step": 1640448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3132554313751056e-05,
+      "loss": 3.9084,
+      "step": 1640960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3124184745044266e-05,
+      "loss": 3.9053,
+      "step": 1641472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3115798797533746e-05,
+      "loss": 3.9135,
+      "step": 1641984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.310741285002323e-05,
+      "loss": 3.9096,
+      "step": 1642496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3099026902512705e-05,
+      "loss": 3.9024,
+      "step": 1643008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3090657333805918e-05,
+      "loss": 3.9134,
+      "step": 1643520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3082271386295398e-05,
+      "loss": 3.9117,
+      "step": 1644032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3073885438784878e-05,
+      "loss": 3.9192,
+      "step": 1644544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.306551587007809e-05,
+      "loss": 3.9129,
+      "step": 1645056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.305712992256757e-05,
+      "loss": 3.8956,
+      "step": 1645568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.304874397505705e-05,
+      "loss": 3.8971,
+      "step": 1646080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.304035802754653e-05,
+      "loss": 3.9042,
+      "step": 1646592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3031972080036007e-05,
+      "loss": 3.912,
+      "step": 1647104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3023586132525487e-05,
+      "loss": 3.907,
+      "step": 1647616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3015200185014967e-05,
+      "loss": 3.9101,
+      "step": 1648128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3006814237504447e-05,
+      "loss": 3.9042,
+      "step": 1648640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.299842828999393e-05,
+      "loss": 3.9009,
+      "step": 1649152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.299004234248341e-05,
+      "loss": 3.9081,
+      "step": 1649664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2981656394972887e-05,
+      "loss": 3.9037,
+      "step": 1650176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2973270447462366e-05,
+      "loss": 3.9144,
+      "step": 1650688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.296490087875558e-05,
+      "loss": 3.9153,
+      "step": 1651200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.295651493124506e-05,
+      "loss": 3.9118,
+      "step": 1651712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2948161741342004e-05,
+      "loss": 3.8996,
+      "step": 1652224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.293977579383148e-05,
+      "loss": 3.908,
+      "step": 1652736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.293138984632096e-05,
+      "loss": 3.9085,
+      "step": 1653248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.292300389881044e-05,
+      "loss": 3.9194,
+      "step": 1653760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.291461795129992e-05,
+      "loss": 3.9048,
+      "step": 1654272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.29062320037894e-05,
+      "loss": 3.8994,
+      "step": 1654784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2897846056278884e-05,
+      "loss": 3.9162,
+      "step": 1655296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.288946010876836e-05,
+      "loss": 3.9061,
+      "step": 1655808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.288107416125784e-05,
+      "loss": 3.9046,
+      "step": 1656320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2872704592551053e-05,
+      "loss": 3.9077,
+      "step": 1656832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2864318645040533e-05,
+      "loss": 3.9053,
+      "step": 1657344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2855932697530013e-05,
+      "loss": 3.8988,
+      "step": 1657856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2847546750019493e-05,
+      "loss": 3.8938,
+      "step": 1658368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2839160802508972e-05,
+      "loss": 3.907,
+      "step": 1658880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.283079123380218e-05,
+      "loss": 3.9072,
+      "step": 1659392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.282240528629166e-05,
+      "loss": 3.9053,
+      "step": 1659904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.281401933878114e-05,
+      "loss": 3.8985,
+      "step": 1660416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.280563339127062e-05,
+      "loss": 3.9013,
+      "step": 1660928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.27972474437601e-05,
+      "loss": 3.9214,
+      "step": 1661440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2788861496249585e-05,
+      "loss": 3.9201,
+      "step": 1661952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2780475548739065e-05,
+      "loss": 3.9163,
+      "step": 1662464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2772105980032274e-05,
+      "loss": 3.909,
+      "step": 1662976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2763720032521754e-05,
+      "loss": 3.9007,
+      "step": 1663488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2755350463814963e-05,
+      "loss": 3.9046,
+      "step": 1664000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2746964516304446e-05,
+      "loss": 3.8996,
+      "step": 1664512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2738578568793926e-05,
+      "loss": 3.9066,
+      "step": 1665024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2730192621283406e-05,
+      "loss": 3.9026,
+      "step": 1665536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2721806673772886e-05,
+      "loss": 3.9102,
+      "step": 1666048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2713420726262363e-05,
+      "loss": 3.9053,
+      "step": 1666560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2705034778751843e-05,
+      "loss": 3.9071,
+      "step": 1667072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2696665210045055e-05,
+      "loss": 3.9043,
+      "step": 1667584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.268827926253454e-05,
+      "loss": 3.9051,
+      "step": 1668096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2679893315024015e-05,
+      "loss": 3.9027,
+      "step": 1668608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2671523746317227e-05,
+      "loss": 3.9102,
+      "step": 1669120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2663137798806707e-05,
+      "loss": 3.9178,
+      "step": 1669632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2654751851296187e-05,
+      "loss": 3.9159,
+      "step": 1670144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2646365903785667e-05,
+      "loss": 3.9023,
+      "step": 1670656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2637979956275147e-05,
+      "loss": 3.9142,
+      "step": 1671168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2629594008764627e-05,
+      "loss": 3.9079,
+      "step": 1671680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2621208061254107e-05,
+      "loss": 3.9198,
+      "step": 1672192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2612822113743587e-05,
+      "loss": 3.9074,
+      "step": 1672704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2604436166233067e-05,
+      "loss": 3.9095,
+      "step": 1673216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2596050218722544e-05,
+      "loss": 3.905,
+      "step": 1673728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2587664271212024e-05,
+      "loss": 3.912,
+      "step": 1674240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2579278323701507e-05,
+      "loss": 3.9034,
+      "step": 1674752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.257090875499472e-05,
+      "loss": 3.9069,
+      "step": 1675264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2562522807484196e-05,
+      "loss": 3.9047,
+      "step": 1675776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2554136859973676e-05,
+      "loss": 3.909,
+      "step": 1676288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2545750912463156e-05,
+      "loss": 3.9119,
+      "step": 1676800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2537364964952636e-05,
+      "loss": 3.9074,
+      "step": 1677312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.252899539624585e-05,
+      "loss": 3.9125,
+      "step": 1677824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.252060944873533e-05,
+      "loss": 3.9054,
+      "step": 1678336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.251223988002854e-05,
+      "loss": 3.9067,
+      "step": 1678848
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.992156744003296,
+      "eval_runtime": 294.8424,
+      "eval_samples_per_second": 1294.22,
+      "eval_steps_per_second": 40.445,
+      "step": 1679028
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2503853932518017e-05,
+      "loss": 3.9038,
+      "step": 1679360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2495467985007497e-05,
+      "loss": 3.8955,
+      "step": 1679872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2487082037496977e-05,
+      "loss": 3.914,
+      "step": 1680384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.247869608998646e-05,
+      "loss": 3.908,
+      "step": 1680896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.247031014247594e-05,
+      "loss": 3.9174,
+      "step": 1681408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.246192419496542e-05,
+      "loss": 3.9042,
+      "step": 1681920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.24535382474549e-05,
+      "loss": 3.9003,
+      "step": 1682432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2445152299944377e-05,
+      "loss": 3.902,
+      "step": 1682944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2436766352433857e-05,
+      "loss": 3.9049,
+      "step": 1683456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2428380404923337e-05,
+      "loss": 3.9024,
+      "step": 1683968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2419994457412817e-05,
+      "loss": 3.9162,
+      "step": 1684480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.24116085099023e-05,
+      "loss": 3.9178,
+      "step": 1684992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.240323894119551e-05,
+      "loss": 3.8935,
+      "step": 1685504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.239485299368499e-05,
+      "loss": 3.9009,
+      "step": 1686016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.238646704617447e-05,
+      "loss": 3.9134,
+      "step": 1686528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.237808109866395e-05,
+      "loss": 3.8921,
+      "step": 1687040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.236969515115343e-05,
+      "loss": 3.9055,
+      "step": 1687552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.236130920364291e-05,
+      "loss": 3.905,
+      "step": 1688064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.235292325613239e-05,
+      "loss": 3.9032,
+      "step": 1688576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.234453730862187e-05,
+      "loss": 3.9192,
+      "step": 1689088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.233615136111135e-05,
+      "loss": 3.9054,
+      "step": 1689600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.232776541360083e-05,
+      "loss": 3.9106,
+      "step": 1690112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.231937946609031e-05,
+      "loss": 3.9064,
+      "step": 1690624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.231099351857979e-05,
+      "loss": 3.9122,
+      "step": 1691136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2302623949873e-05,
+      "loss": 3.9048,
+      "step": 1691648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.229423800236248e-05,
+      "loss": 3.9058,
+      "step": 1692160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.228585205485196e-05,
+      "loss": 3.9021,
+      "step": 1692672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2277466107341438e-05,
+      "loss": 3.899,
+      "step": 1693184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.226909653863465e-05,
+      "loss": 3.8998,
+      "step": 1693696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.226071059112413e-05,
+      "loss": 3.9026,
+      "step": 1694208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.225232464361361e-05,
+      "loss": 3.9065,
+      "step": 1694720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.224393869610309e-05,
+      "loss": 3.906,
+      "step": 1695232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2235569127396303e-05,
+      "loss": 3.9122,
+      "step": 1695744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2227199558689512e-05,
+      "loss": 3.9163,
+      "step": 1696256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2218829989982724e-05,
+      "loss": 3.9034,
+      "step": 1696768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2210444042472204e-05,
+      "loss": 3.9066,
+      "step": 1697280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2202058094961684e-05,
+      "loss": 3.9052,
+      "step": 1697792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2193672147451164e-05,
+      "loss": 3.9029,
+      "step": 1698304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2185286199940644e-05,
+      "loss": 3.8938,
+      "step": 1698816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2176900252430124e-05,
+      "loss": 3.9014,
+      "step": 1699328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2168514304919604e-05,
+      "loss": 3.9004,
+      "step": 1699840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2160128357409084e-05,
+      "loss": 3.9089,
+      "step": 1700352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2151742409898564e-05,
+      "loss": 3.9015,
+      "step": 1700864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2143356462388044e-05,
+      "loss": 3.9032,
+      "step": 1701376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2134970514877524e-05,
+      "loss": 3.9202,
+      "step": 1701888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2126584567367004e-05,
+      "loss": 3.9055,
+      "step": 1702400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2118214998660213e-05,
+      "loss": 3.9091,
+      "step": 1702912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2109829051149693e-05,
+      "loss": 3.8968,
+      "step": 1703424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2101443103639173e-05,
+      "loss": 3.8867,
+      "step": 1703936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2093073534932385e-05,
+      "loss": 3.9142,
+      "step": 1704448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2084687587421865e-05,
+      "loss": 3.9023,
+      "step": 1704960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2076301639911345e-05,
+      "loss": 3.8961,
+      "step": 1705472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2067915692400825e-05,
+      "loss": 3.9006,
+      "step": 1705984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2059529744890305e-05,
+      "loss": 3.9082,
+      "step": 1706496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2051143797379785e-05,
+      "loss": 3.8828,
+      "step": 1707008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2042774228672998e-05,
+      "loss": 3.9024,
+      "step": 1707520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2034388281162477e-05,
+      "loss": 3.8982,
+      "step": 1708032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2026002333651957e-05,
+      "loss": 3.8925,
+      "step": 1708544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2017616386141437e-05,
+      "loss": 3.9125,
+      "step": 1709056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2009230438630917e-05,
+      "loss": 3.8979,
+      "step": 1709568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2000844491120394e-05,
+      "loss": 3.8968,
+      "step": 1710080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1992458543609877e-05,
+      "loss": 3.9067,
+      "step": 1710592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1984072596099357e-05,
+      "loss": 3.8918,
+      "step": 1711104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1975686648588837e-05,
+      "loss": 3.8931,
+      "step": 1711616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.196733345868578e-05,
+      "loss": 3.9045,
+      "step": 1712128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.195894751117526e-05,
+      "loss": 3.8997,
+      "step": 1712640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.195057794246847e-05,
+      "loss": 3.9039,
+      "step": 1713152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.194219199495795e-05,
+      "loss": 3.8915,
+      "step": 1713664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.193380604744743e-05,
+      "loss": 3.8848,
+      "step": 1714176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.192542009993691e-05,
+      "loss": 3.8998,
+      "step": 1714688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.191703415242639e-05,
+      "loss": 3.9128,
+      "step": 1715200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1908648204915868e-05,
+      "loss": 3.8976,
+      "step": 1715712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1900262257405348e-05,
+      "loss": 3.914,
+      "step": 1716224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.189187630989483e-05,
+      "loss": 3.911,
+      "step": 1716736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.188349036238431e-05,
+      "loss": 3.9047,
+      "step": 1717248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.187512079367752e-05,
+      "loss": 3.8927,
+      "step": 1717760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1866734846167e-05,
+      "loss": 3.9115,
+      "step": 1718272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.185834889865648e-05,
+      "loss": 3.9006,
+      "step": 1718784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.184996295114596e-05,
+      "loss": 3.8959,
+      "step": 1719296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.184157700363544e-05,
+      "loss": 3.9085,
+      "step": 1719808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.183319105612492e-05,
+      "loss": 3.9054,
+      "step": 1720320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.18248051086144e-05,
+      "loss": 3.9142,
+      "step": 1720832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.181641916110388e-05,
+      "loss": 3.905,
+      "step": 1721344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1808049592397092e-05,
+      "loss": 3.8896,
+      "step": 1721856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1799663644886572e-05,
+      "loss": 3.89,
+      "step": 1722368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.179127769737605e-05,
+      "loss": 3.8969,
+      "step": 1722880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1782891749865532e-05,
+      "loss": 3.9055,
+      "step": 1723392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1774522181158745e-05,
+      "loss": 3.9017,
+      "step": 1723904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1766152612451954e-05,
+      "loss": 3.8997,
+      "step": 1724416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1757766664941434e-05,
+      "loss": 3.9021,
+      "step": 1724928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1749380717430914e-05,
+      "loss": 3.8912,
+      "step": 1725440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1740994769920393e-05,
+      "loss": 3.9011,
+      "step": 1725952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1732625201213606e-05,
+      "loss": 3.8968,
+      "step": 1726464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1724239253703086e-05,
+      "loss": 3.9141,
+      "step": 1726976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1715853306192566e-05,
+      "loss": 3.9086,
+      "step": 1727488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1707467358682046e-05,
+      "loss": 3.9045,
+      "step": 1728000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1699081411171522e-05,
+      "loss": 3.8972,
+      "step": 1728512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1690711842464735e-05,
+      "loss": 3.9039,
+      "step": 1729024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1682325894954218e-05,
+      "loss": 3.899,
+      "step": 1729536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1673939947443695e-05,
+      "loss": 3.9143,
+      "step": 1730048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1665553999933175e-05,
+      "loss": 3.8981,
+      "step": 1730560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1657168052422655e-05,
+      "loss": 3.8941,
+      "step": 1731072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1648782104912135e-05,
+      "loss": 3.9063,
+      "step": 1731584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1640396157401615e-05,
+      "loss": 3.9028,
+      "step": 1732096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1632010209891095e-05,
+      "loss": 3.8952,
+      "step": 1732608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1623640641184307e-05,
+      "loss": 3.9026,
+      "step": 1733120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1615254693673787e-05,
+      "loss": 3.8951,
+      "step": 1733632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1606868746163267e-05,
+      "loss": 3.8927,
+      "step": 1734144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1598482798652747e-05,
+      "loss": 3.886,
+      "step": 1734656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1590096851142223e-05,
+      "loss": 3.9051,
+      "step": 1735168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1581710903631703e-05,
+      "loss": 3.8986,
+      "step": 1735680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.157334133492492e-05,
+      "loss": 3.9001,
+      "step": 1736192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.15649553874144e-05,
+      "loss": 3.8951,
+      "step": 1736704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1556569439903876e-05,
+      "loss": 3.8949,
+      "step": 1737216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1548183492393356e-05,
+      "loss": 3.9127,
+      "step": 1737728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1539797544882836e-05,
+      "loss": 3.9158,
+      "step": 1738240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1531411597372316e-05,
+      "loss": 3.9114,
+      "step": 1738752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1523042028665528e-05,
+      "loss": 3.9011,
+      "step": 1739264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1514656081155008e-05,
+      "loss": 3.8963,
+      "step": 1739776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1506270133644488e-05,
+      "loss": 3.8982,
+      "step": 1740288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1497884186133968e-05,
+      "loss": 3.8886,
+      "step": 1740800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1489498238623448e-05,
+      "loss": 3.9001,
+      "step": 1741312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1481112291112928e-05,
+      "loss": 3.8995,
+      "step": 1741824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.147274272240614e-05,
+      "loss": 3.9015,
+      "step": 1742336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.146435677489562e-05,
+      "loss": 3.9033,
+      "step": 1742848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.14559708273851e-05,
+      "loss": 3.8987,
+      "step": 1743360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.144758487987458e-05,
+      "loss": 3.9034,
+      "step": 1743872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1439198932364057e-05,
+      "loss": 3.8962,
+      "step": 1744384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1430812984853537e-05,
+      "loss": 3.9004,
+      "step": 1744896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1422427037343017e-05,
+      "loss": 3.9012,
+      "step": 1745408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1414041089832497e-05,
+      "loss": 3.9108,
+      "step": 1745920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.140567152112571e-05,
+      "loss": 3.9104,
+      "step": 1746432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.139728557361519e-05,
+      "loss": 3.8984,
+      "step": 1746944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.138889962610467e-05,
+      "loss": 3.91,
+      "step": 1747456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.138051367859415e-05,
+      "loss": 3.9022,
+      "step": 1747968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1372144109887358e-05,
+      "loss": 3.9096,
+      "step": 1748480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.136375816237684e-05,
+      "loss": 3.9046,
+      "step": 1748992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.135537221486632e-05,
+      "loss": 3.9078,
+      "step": 1749504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.13469862673558e-05,
+      "loss": 3.896,
+      "step": 1750016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.133860031984528e-05,
+      "loss": 3.9068,
+      "step": 1750528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.133023075113849e-05,
+      "loss": 3.895,
+      "step": 1751040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.132184480362797e-05,
+      "loss": 3.9011,
+      "step": 1751552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.131345885611745e-05,
+      "loss": 3.9023,
+      "step": 1752064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1305072908606934e-05,
+      "loss": 3.9013,
+      "step": 1752576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1296703339900143e-05,
+      "loss": 3.9044,
+      "step": 1753088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1288317392389623e-05,
+      "loss": 3.9003,
+      "step": 1753600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1279931444879103e-05,
+      "loss": 3.9104,
+      "step": 1754112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1271545497368583e-05,
+      "loss": 3.895,
+      "step": 1754624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1263192307465528e-05,
+      "loss": 3.9079,
+      "step": 1755136
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9895548820495605,
+      "eval_runtime": 300.0652,
+      "eval_samples_per_second": 1271.693,
+      "eval_steps_per_second": 39.741,
+      "step": 1755348
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1254806359955004e-05,
+      "loss": 3.901,
+      "step": 1755648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1246420412444484e-05,
+      "loss": 3.8886,
+      "step": 1756160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1238034464933964e-05,
+      "loss": 3.9073,
+      "step": 1756672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1229648517423444e-05,
+      "loss": 3.9025,
+      "step": 1757184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1221262569912924e-05,
+      "loss": 3.9076,
+      "step": 1757696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1212876622402404e-05,
+      "loss": 3.8984,
+      "step": 1758208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1204507053695617e-05,
+      "loss": 3.8961,
+      "step": 1758720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1196121106185097e-05,
+      "loss": 3.8974,
+      "step": 1759232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1187735158674576e-05,
+      "loss": 3.8988,
+      "step": 1759744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1179349211164056e-05,
+      "loss": 3.8947,
+      "step": 1760256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1170979642457266e-05,
+      "loss": 3.9119,
+      "step": 1760768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.116259369494675e-05,
+      "loss": 3.9082,
+      "step": 1761280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.115420774743623e-05,
+      "loss": 3.8909,
+      "step": 1761792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.114582179992571e-05,
+      "loss": 3.8921,
+      "step": 1762304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1137435852415185e-05,
+      "loss": 3.9045,
+      "step": 1762816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1129066283708398e-05,
+      "loss": 3.8892,
+      "step": 1763328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1120680336197878e-05,
+      "loss": 3.8976,
+      "step": 1763840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1112294388687358e-05,
+      "loss": 3.8952,
+      "step": 1764352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1103908441176838e-05,
+      "loss": 3.9036,
+      "step": 1764864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1095522493666318e-05,
+      "loss": 3.9079,
+      "step": 1765376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.108715292495953e-05,
+      "loss": 3.9065,
+      "step": 1765888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1078766977449007e-05,
+      "loss": 3.9014,
+      "step": 1766400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1070381029938487e-05,
+      "loss": 3.9003,
+      "step": 1766912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1061995082427967e-05,
+      "loss": 3.9056,
+      "step": 1767424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.105360913491745e-05,
+      "loss": 3.8944,
+      "step": 1767936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.104523956621066e-05,
+      "loss": 3.903,
+      "step": 1768448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.103685361870014e-05,
+      "loss": 3.8934,
+      "step": 1768960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.102846767118962e-05,
+      "loss": 3.893,
+      "step": 1769472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.10200817236791e-05,
+      "loss": 3.8955,
+      "step": 1769984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.101171215497231e-05,
+      "loss": 3.9013,
+      "step": 1770496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.100332620746179e-05,
+      "loss": 3.8976,
+      "step": 1771008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0994956638755004e-05,
+      "loss": 3.9036,
+      "step": 1771520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.098657069124448e-05,
+      "loss": 3.9067,
+      "step": 1772032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.097818474373396e-05,
+      "loss": 3.909,
+      "step": 1772544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0969815175027173e-05,
+      "loss": 3.8932,
+      "step": 1773056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0961429227516656e-05,
+      "loss": 3.901,
+      "step": 1773568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0953043280006133e-05,
+      "loss": 3.9016,
+      "step": 1774080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0944657332495613e-05,
+      "loss": 3.8942,
+      "step": 1774592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0936271384985093e-05,
+      "loss": 3.888,
+      "step": 1775104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0927885437474573e-05,
+      "loss": 3.8923,
+      "step": 1775616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0919499489964053e-05,
+      "loss": 3.8966,
+      "step": 1776128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0911113542453533e-05,
+      "loss": 3.9038,
+      "step": 1776640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0902727594943013e-05,
+      "loss": 3.8917,
+      "step": 1777152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0894341647432492e-05,
+      "loss": 3.8946,
+      "step": 1777664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0885955699921972e-05,
+      "loss": 3.9133,
+      "step": 1778176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0877569752411452e-05,
+      "loss": 3.9008,
+      "step": 1778688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0869183804900932e-05,
+      "loss": 3.9067,
+      "step": 1779200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0860797857390412e-05,
+      "loss": 3.8888,
+      "step": 1779712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0852411909879892e-05,
+      "loss": 3.8792,
+      "step": 1780224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0844025962369372e-05,
+      "loss": 3.908,
+      "step": 1780736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0835656393662585e-05,
+      "loss": 3.8959,
+      "step": 1781248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0827270446152065e-05,
+      "loss": 3.8917,
+      "step": 1781760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0818884498641545e-05,
+      "loss": 3.8989,
+      "step": 1782272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.081049855113102e-05,
+      "loss": 3.8937,
+      "step": 1782784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.08021126036205e-05,
+      "loss": 3.8836,
+      "step": 1783296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0793759413717446e-05,
+      "loss": 3.892,
+      "step": 1783808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0785373466206926e-05,
+      "loss": 3.8923,
+      "step": 1784320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0776987518696406e-05,
+      "loss": 3.8893,
+      "step": 1784832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0768617949989615e-05,
+      "loss": 3.9042,
+      "step": 1785344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0760232002479095e-05,
+      "loss": 3.8926,
+      "step": 1785856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0751846054968575e-05,
+      "loss": 3.8877,
+      "step": 1786368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.074346010745806e-05,
+      "loss": 3.9031,
+      "step": 1786880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.073507415994754e-05,
+      "loss": 3.8851,
+      "step": 1787392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.072668821243702e-05,
+      "loss": 3.8909,
+      "step": 1787904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0718302264926495e-05,
+      "loss": 3.8942,
+      "step": 1788416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0709916317415975e-05,
+      "loss": 3.8924,
+      "step": 1788928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0701530369905455e-05,
+      "loss": 3.8941,
+      "step": 1789440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0693144422394935e-05,
+      "loss": 3.8933,
+      "step": 1789952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0684758474884418e-05,
+      "loss": 3.8772,
+      "step": 1790464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0676372527373898e-05,
+      "loss": 3.8914,
+      "step": 1790976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0668002958667107e-05,
+      "loss": 3.909,
+      "step": 1791488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0659633389960316e-05,
+      "loss": 3.8859,
+      "step": 1792000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0651247442449796e-05,
+      "loss": 3.9118,
+      "step": 1792512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.064286149493928e-05,
+      "loss": 3.9017,
+      "step": 1793024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.063447554742876e-05,
+      "loss": 3.9038,
+      "step": 1793536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.062610597872197e-05,
+      "loss": 3.8861,
+      "step": 1794048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.061772003121145e-05,
+      "loss": 3.901,
+      "step": 1794560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.060933408370093e-05,
+      "loss": 3.9005,
+      "step": 1795072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.060094813619041e-05,
+      "loss": 3.8859,
+      "step": 1795584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.059256218867989e-05,
+      "loss": 3.9064,
+      "step": 1796096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.05841926199731e-05,
+      "loss": 3.899,
+      "step": 1796608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.057580667246258e-05,
+      "loss": 3.9041,
+      "step": 1797120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.056742072495206e-05,
+      "loss": 3.9011,
+      "step": 1797632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.055903477744154e-05,
+      "loss": 3.8832,
+      "step": 1798144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.055064882993102e-05,
+      "loss": 3.8815,
+      "step": 1798656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0542279261224233e-05,
+      "loss": 3.8888,
+      "step": 1799168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0533893313713713e-05,
+      "loss": 3.9012,
+      "step": 1799680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0525507366203193e-05,
+      "loss": 3.8977,
+      "step": 1800192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0517121418692673e-05,
+      "loss": 3.8943,
+      "step": 1800704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.050873547118215e-05,
+      "loss": 3.8951,
+      "step": 1801216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0500365902475362e-05,
+      "loss": 3.8901,
+      "step": 1801728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0491979954964842e-05,
+      "loss": 3.8913,
+      "step": 1802240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0483594007454322e-05,
+      "loss": 3.8904,
+      "step": 1802752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0475208059943802e-05,
+      "loss": 3.9054,
+      "step": 1803264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0466822112433282e-05,
+      "loss": 3.9103,
+      "step": 1803776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0458436164922762e-05,
+      "loss": 3.8939,
+      "step": 1804288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0450050217412242e-05,
+      "loss": 3.8906,
+      "step": 1804800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0441664269901722e-05,
+      "loss": 3.8972,
+      "step": 1805312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0433294701194934e-05,
+      "loss": 3.8961,
+      "step": 1805824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0424908753684414e-05,
+      "loss": 3.9068,
+      "step": 1806336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0416522806173894e-05,
+      "loss": 3.8928,
+      "step": 1806848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0408136858663374e-05,
+      "loss": 3.891,
+      "step": 1807360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.039975091115285e-05,
+      "loss": 3.8994,
+      "step": 1807872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0391381342446063e-05,
+      "loss": 3.8984,
+      "step": 1808384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0382995394935543e-05,
+      "loss": 3.8858,
+      "step": 1808896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0374625826228756e-05,
+      "loss": 3.8951,
+      "step": 1809408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0366239878718236e-05,
+      "loss": 3.8938,
+      "step": 1809920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0357853931207716e-05,
+      "loss": 3.8834,
+      "step": 1810432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0349467983697196e-05,
+      "loss": 3.8849,
+      "step": 1810944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0341082036186675e-05,
+      "loss": 3.8959,
+      "step": 1811456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0332696088676152e-05,
+      "loss": 3.8977,
+      "step": 1811968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0324326519969368e-05,
+      "loss": 3.8886,
+      "step": 1812480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0315940572458848e-05,
+      "loss": 3.8911,
+      "step": 1812992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0307554624948324e-05,
+      "loss": 3.8904,
+      "step": 1813504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0299168677437804e-05,
+      "loss": 3.9046,
+      "step": 1814016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0290782729927284e-05,
+      "loss": 3.9093,
+      "step": 1814528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0282413161220497e-05,
+      "loss": 3.904,
+      "step": 1815040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0274027213709977e-05,
+      "loss": 3.8939,
+      "step": 1815552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0265641266199457e-05,
+      "loss": 3.8888,
+      "step": 1816064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0257255318688937e-05,
+      "loss": 3.8973,
+      "step": 1816576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0248869371178417e-05,
+      "loss": 3.8837,
+      "step": 1817088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0240483423667897e-05,
+      "loss": 3.8935,
+      "step": 1817600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0232097476157377e-05,
+      "loss": 3.8942,
+      "step": 1818112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0223711528646857e-05,
+      "loss": 3.8926,
+      "step": 1818624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0215325581136336e-05,
+      "loss": 3.8949,
+      "step": 1819136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.020695601242955e-05,
+      "loss": 3.8932,
+      "step": 1819648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.019857006491903e-05,
+      "loss": 3.8991,
+      "step": 1820160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0190184117408505e-05,
+      "loss": 3.8928,
+      "step": 1820672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0181798169897985e-05,
+      "loss": 3.8907,
+      "step": 1821184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0173412222387465e-05,
+      "loss": 3.8954,
+      "step": 1821696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.016504265368068e-05,
+      "loss": 3.8998,
+      "step": 1822208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0156656706170158e-05,
+      "loss": 3.9061,
+      "step": 1822720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0148270758659638e-05,
+      "loss": 3.891,
+      "step": 1823232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0139884811149118e-05,
+      "loss": 3.9037,
+      "step": 1823744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.013151524244233e-05,
+      "loss": 3.8988,
+      "step": 1824256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0123129294931807e-05,
+      "loss": 3.9031,
+      "step": 1824768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0114759726225023e-05,
+      "loss": 3.8936,
+      "step": 1825280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0106373778714503e-05,
+      "loss": 3.9003,
+      "step": 1825792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.009798783120398e-05,
+      "loss": 3.8899,
+      "step": 1826304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.008960188369346e-05,
+      "loss": 3.9046,
+      "step": 1826816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.008121593618294e-05,
+      "loss": 3.8858,
+      "step": 1827328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.007282998867242e-05,
+      "loss": 3.8965,
+      "step": 1827840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.00644440411619e-05,
+      "loss": 3.8969,
+      "step": 1828352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.005607447245511e-05,
+      "loss": 3.8976,
+      "step": 1828864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.004768852494459e-05,
+      "loss": 3.896,
+      "step": 1829376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0039318956237804e-05,
+      "loss": 3.8942,
+      "step": 1829888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.003093300872728e-05,
+      "loss": 3.9062,
+      "step": 1830400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.002254706121676e-05,
+      "loss": 3.8884,
+      "step": 1830912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0014161113706244e-05,
+      "loss": 3.904,
+      "step": 1831424
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.987441062927246,
+      "eval_runtime": 304.9485,
+      "eval_samples_per_second": 1251.329,
+      "eval_steps_per_second": 39.105,
+      "step": 1831668
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0005775166195724e-05,
+      "loss": 3.8895,
+      "step": 1831936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9997389218685204e-05,
+      "loss": 3.8851,
+      "step": 1832448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9989003271174684e-05,
+      "loss": 3.8991,
+      "step": 1832960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.998061732366416e-05,
+      "loss": 3.8966,
+      "step": 1833472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.997223137615364e-05,
+      "loss": 3.9023,
+      "step": 1833984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.996384542864312e-05,
+      "loss": 3.8917,
+      "step": 1834496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.99554594811326e-05,
+      "loss": 3.8939,
+      "step": 1835008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9947073533622083e-05,
+      "loss": 3.893,
+      "step": 1835520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9938687586111563e-05,
+      "loss": 3.891,
+      "step": 1836032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9930301638601043e-05,
+      "loss": 3.8926,
+      "step": 1836544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.992191569109052e-05,
+      "loss": 3.9001,
+      "step": 1837056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9913546122383732e-05,
+      "loss": 3.9025,
+      "step": 1837568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9905160174873212e-05,
+      "loss": 3.8863,
+      "step": 1838080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9896774227362692e-05,
+      "loss": 3.8849,
+      "step": 1838592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9888388279852172e-05,
+      "loss": 3.901,
+      "step": 1839104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9880002332341652e-05,
+      "loss": 3.8819,
+      "step": 1839616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9871616384831132e-05,
+      "loss": 3.8934,
+      "step": 1840128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9863230437320612e-05,
+      "loss": 3.8908,
+      "step": 1840640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9854844489810092e-05,
+      "loss": 3.8972,
+      "step": 1841152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9846458542299572e-05,
+      "loss": 3.9006,
+      "step": 1841664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9838072594789052e-05,
+      "loss": 3.9007,
+      "step": 1842176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9829686647278532e-05,
+      "loss": 3.8975,
+      "step": 1842688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9821300699768012e-05,
+      "loss": 3.8929,
+      "step": 1843200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9812914752257492e-05,
+      "loss": 3.8964,
+      "step": 1843712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.98045451835507e-05,
+      "loss": 3.8926,
+      "step": 1844224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.979615923604018e-05,
+      "loss": 3.8983,
+      "step": 1844736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.978777328852966e-05,
+      "loss": 3.8872,
+      "step": 1845248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9779387341019144e-05,
+      "loss": 3.8873,
+      "step": 1845760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9771001393508624e-05,
+      "loss": 3.8951,
+      "step": 1846272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9762615445998104e-05,
+      "loss": 3.8882,
+      "step": 1846784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.975422949848758e-05,
+      "loss": 3.8896,
+      "step": 1847296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.974584355097706e-05,
+      "loss": 3.8999,
+      "step": 1847808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9737473982270273e-05,
+      "loss": 3.9019,
+      "step": 1848320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9729088034759753e-05,
+      "loss": 3.899,
+      "step": 1848832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9720718466052966e-05,
+      "loss": 3.8889,
+      "step": 1849344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9712332518542446e-05,
+      "loss": 3.9012,
+      "step": 1849856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9703946571031926e-05,
+      "loss": 3.8967,
+      "step": 1850368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9695560623521405e-05,
+      "loss": 3.8892,
+      "step": 1850880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9687174676010882e-05,
+      "loss": 3.8842,
+      "step": 1851392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9678805107304098e-05,
+      "loss": 3.8841,
+      "step": 1851904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9670419159793578e-05,
+      "loss": 3.888,
+      "step": 1852416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9662033212283054e-05,
+      "loss": 3.8997,
+      "step": 1852928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9653647264772534e-05,
+      "loss": 3.8848,
+      "step": 1853440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9645261317262014e-05,
+      "loss": 3.8875,
+      "step": 1853952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9636875369751494e-05,
+      "loss": 3.9084,
+      "step": 1854464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9628489422240974e-05,
+      "loss": 3.8916,
+      "step": 1854976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9620103474730454e-05,
+      "loss": 3.9053,
+      "step": 1855488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9611733906023667e-05,
+      "loss": 3.8864,
+      "step": 1856000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9603347958513147e-05,
+      "loss": 3.8666,
+      "step": 1856512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9594962011002627e-05,
+      "loss": 3.9044,
+      "step": 1857024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9586576063492107e-05,
+      "loss": 3.893,
+      "step": 1857536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9578190115981587e-05,
+      "loss": 3.8884,
+      "step": 1858048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.95698205472748e-05,
+      "loss": 3.8901,
+      "step": 1858560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.956143459976428e-05,
+      "loss": 3.8877,
+      "step": 1859072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.955304865225376e-05,
+      "loss": 3.8805,
+      "step": 1859584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9544662704743235e-05,
+      "loss": 3.8847,
+      "step": 1860096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9536293136036448e-05,
+      "loss": 3.8845,
+      "step": 1860608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9527907188525928e-05,
+      "loss": 3.8853,
+      "step": 1861120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9519521241015408e-05,
+      "loss": 3.899,
+      "step": 1861632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9511135293504888e-05,
+      "loss": 3.8858,
+      "step": 1862144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9502749345994368e-05,
+      "loss": 3.8845,
+      "step": 1862656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9494363398483848e-05,
+      "loss": 3.8993,
+      "step": 1863168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9485977450973328e-05,
+      "loss": 3.8766,
+      "step": 1863680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9477591503462808e-05,
+      "loss": 3.8849,
+      "step": 1864192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9469205555952288e-05,
+      "loss": 3.8856,
+      "step": 1864704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.94608359872455e-05,
+      "loss": 3.8916,
+      "step": 1865216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.945246641853871e-05,
+      "loss": 3.8879,
+      "step": 1865728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.944408047102819e-05,
+      "loss": 3.8887,
+      "step": 1866240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.943569452351767e-05,
+      "loss": 3.8679,
+      "step": 1866752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.942730857600715e-05,
+      "loss": 3.8878,
+      "step": 1867264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.941892262849663e-05,
+      "loss": 3.9024,
+      "step": 1867776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.941053668098611e-05,
+      "loss": 3.8832,
+      "step": 1868288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9402150733475592e-05,
+      "loss": 3.9055,
+      "step": 1868800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.939376478596507e-05,
+      "loss": 3.8936,
+      "step": 1869312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.938537883845455e-05,
+      "loss": 3.9041,
+      "step": 1869824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.937702564855149e-05,
+      "loss": 3.8816,
+      "step": 1870336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.936863970104097e-05,
+      "loss": 3.8963,
+      "step": 1870848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9360253753530454e-05,
+      "loss": 3.8943,
+      "step": 1871360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9351884184823663e-05,
+      "loss": 3.8807,
+      "step": 1871872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9343498237313143e-05,
+      "loss": 3.9006,
+      "step": 1872384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9335112289802623e-05,
+      "loss": 3.8907,
+      "step": 1872896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9326726342292103e-05,
+      "loss": 3.9015,
+      "step": 1873408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9318340394781583e-05,
+      "loss": 3.8975,
+      "step": 1873920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9309954447271063e-05,
+      "loss": 3.8762,
+      "step": 1874432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9301568499760543e-05,
+      "loss": 3.8801,
+      "step": 1874944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9293182552250023e-05,
+      "loss": 3.8842,
+      "step": 1875456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9284796604739502e-05,
+      "loss": 3.8953,
+      "step": 1875968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9276410657228982e-05,
+      "loss": 3.8883,
+      "step": 1876480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9268024709718462e-05,
+      "loss": 3.8869,
+      "step": 1876992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9259638762207942e-05,
+      "loss": 3.8879,
+      "step": 1877504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9251269193501155e-05,
+      "loss": 3.8881,
+      "step": 1878016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9242899624794364e-05,
+      "loss": 3.888,
+      "step": 1878528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9234513677283844e-05,
+      "loss": 3.8827,
+      "step": 1879040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9226127729773324e-05,
+      "loss": 3.9035,
+      "step": 1879552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9217741782262804e-05,
+      "loss": 3.9004,
+      "step": 1880064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9209355834752284e-05,
+      "loss": 3.8905,
+      "step": 1880576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9200969887241764e-05,
+      "loss": 3.8821,
+      "step": 1881088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9192583939731247e-05,
+      "loss": 3.8931,
+      "step": 1881600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9184197992220724e-05,
+      "loss": 3.8844,
+      "step": 1882112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9175828423513936e-05,
+      "loss": 3.9037,
+      "step": 1882624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9167442476003416e-05,
+      "loss": 3.8884,
+      "step": 1883136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9159056528492893e-05,
+      "loss": 3.8882,
+      "step": 1883648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.915068695978611e-05,
+      "loss": 3.8945,
+      "step": 1884160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.914230101227559e-05,
+      "loss": 3.8942,
+      "step": 1884672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.913391506476507e-05,
+      "loss": 3.8805,
+      "step": 1885184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9125529117254545e-05,
+      "loss": 3.8908,
+      "step": 1885696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9117143169744025e-05,
+      "loss": 3.8855,
+      "step": 1886208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9108773601037237e-05,
+      "loss": 3.883,
+      "step": 1886720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9100387653526717e-05,
+      "loss": 3.8798,
+      "step": 1887232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9092001706016197e-05,
+      "loss": 3.8878,
+      "step": 1887744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9083615758505677e-05,
+      "loss": 3.8877,
+      "step": 1888256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9075229810995157e-05,
+      "loss": 3.8882,
+      "step": 1888768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9066843863484637e-05,
+      "loss": 3.8874,
+      "step": 1889280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9058457915974117e-05,
+      "loss": 3.882,
+      "step": 1889792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.905008834726733e-05,
+      "loss": 3.8964,
+      "step": 1890304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.904170239975681e-05,
+      "loss": 3.9073,
+      "step": 1890816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.903331645224629e-05,
+      "loss": 3.8962,
+      "step": 1891328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.902493050473577e-05,
+      "loss": 3.8938,
+      "step": 1891840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.901654455722525e-05,
+      "loss": 3.8808,
+      "step": 1892352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9008158609714726e-05,
+      "loss": 3.893,
+      "step": 1892864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8999772662204206e-05,
+      "loss": 3.8847,
+      "step": 1893376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8991386714693686e-05,
+      "loss": 3.8883,
+      "step": 1893888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.898303352479063e-05,
+      "loss": 3.8837,
+      "step": 1894400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.897464757728011e-05,
+      "loss": 3.8872,
+      "step": 1894912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.896626162976959e-05,
+      "loss": 3.8905,
+      "step": 1895424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.895787568225907e-05,
+      "loss": 3.8907,
+      "step": 1895936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8949489734748547e-05,
+      "loss": 3.893,
+      "step": 1896448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.894110378723803e-05,
+      "loss": 3.8897,
+      "step": 1896960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8932734218531243e-05,
+      "loss": 3.8822,
+      "step": 1897472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8924348271020723e-05,
+      "loss": 3.8964,
+      "step": 1897984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.89159623235102e-05,
+      "loss": 3.8919,
+      "step": 1898496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.890757637599968e-05,
+      "loss": 3.9035,
+      "step": 1899008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.889919042848916e-05,
+      "loss": 3.8834,
+      "step": 1899520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.889080448097864e-05,
+      "loss": 3.9015,
+      "step": 1900032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8882418533468123e-05,
+      "loss": 3.8919,
+      "step": 1900544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8874032585957603e-05,
+      "loss": 3.8961,
+      "step": 1901056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8865663017250812e-05,
+      "loss": 3.8939,
+      "step": 1901568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8857277069740292e-05,
+      "loss": 3.8924,
+      "step": 1902080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8848891122229772e-05,
+      "loss": 3.8852,
+      "step": 1902592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8840505174719252e-05,
+      "loss": 3.9005,
+      "step": 1903104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8832135606012464e-05,
+      "loss": 3.8816,
+      "step": 1903616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8823749658501944e-05,
+      "loss": 3.8894,
+      "step": 1904128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8815363710991424e-05,
+      "loss": 3.8935,
+      "step": 1904640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8806977763480904e-05,
+      "loss": 3.8966,
+      "step": 1905152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8798608194774113e-05,
+      "loss": 3.8859,
+      "step": 1905664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8790222247263593e-05,
+      "loss": 3.8885,
+      "step": 1906176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8781836299753077e-05,
+      "loss": 3.8979,
+      "step": 1906688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8773450352242557e-05,
+      "loss": 3.888,
+      "step": 1907200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8765064404732033e-05,
+      "loss": 3.897,
+      "step": 1907712
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.985560655593872,
+      "eval_runtime": 279.2204,
+      "eval_samples_per_second": 1366.63,
+      "eval_steps_per_second": 42.708,
+      "step": 1907988
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8756678457221513e-05,
+      "loss": 3.8824,
+      "step": 1908224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8748292509710993e-05,
+      "loss": 3.886,
+      "step": 1908736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8739906562200473e-05,
+      "loss": 3.8906,
+      "step": 1909248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8731536993493686e-05,
+      "loss": 3.8924,
+      "step": 1909760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8723151045983165e-05,
+      "loss": 3.8996,
+      "step": 1910272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8714765098472645e-05,
+      "loss": 3.8861,
+      "step": 1910784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8706395529765855e-05,
+      "loss": 3.8916,
+      "step": 1911296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8698009582255334e-05,
+      "loss": 3.8871,
+      "step": 1911808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8689623634744814e-05,
+      "loss": 3.8828,
+      "step": 1912320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8681237687234294e-05,
+      "loss": 3.8874,
+      "step": 1912832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8672851739723778e-05,
+      "loss": 3.895,
+      "step": 1913344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.866449854982072e-05,
+      "loss": 3.8958,
+      "step": 1913856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.86561126023102e-05,
+      "loss": 3.8903,
+      "step": 1914368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8647726654799676e-05,
+      "loss": 3.874,
+      "step": 1914880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8639340707289156e-05,
+      "loss": 3.8922,
+      "step": 1915392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.863095475977864e-05,
+      "loss": 3.8776,
+      "step": 1915904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.862256881226812e-05,
+      "loss": 3.8885,
+      "step": 1916416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.86141828647576e-05,
+      "loss": 3.886,
+      "step": 1916928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.860579691724708e-05,
+      "loss": 3.8893,
+      "step": 1917440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.859741096973656e-05,
+      "loss": 3.8933,
+      "step": 1917952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8589025022226036e-05,
+      "loss": 3.9013,
+      "step": 1918464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8580639074715516e-05,
+      "loss": 3.8952,
+      "step": 1918976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8572253127204995e-05,
+      "loss": 3.8891,
+      "step": 1919488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8563883558498208e-05,
+      "loss": 3.8875,
+      "step": 1920000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.855551398979142e-05,
+      "loss": 3.8928,
+      "step": 1920512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.85471280422809e-05,
+      "loss": 3.8877,
+      "step": 1921024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.853874209477038e-05,
+      "loss": 3.8899,
+      "step": 1921536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8530356147259857e-05,
+      "loss": 3.885,
+      "step": 1922048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.852197019974934e-05,
+      "loss": 3.8863,
+      "step": 1922560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.851358425223882e-05,
+      "loss": 3.8878,
+      "step": 1923072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.85051983047283e-05,
+      "loss": 3.8776,
+      "step": 1923584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.849681235721778e-05,
+      "loss": 3.8982,
+      "step": 1924096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.848844278851099e-05,
+      "loss": 3.8959,
+      "step": 1924608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.848005684100047e-05,
+      "loss": 3.8935,
+      "step": 1925120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.847168727229368e-05,
+      "loss": 3.8849,
+      "step": 1925632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8463317703586894e-05,
+      "loss": 3.8943,
+      "step": 1926144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8454931756076374e-05,
+      "loss": 3.8906,
+      "step": 1926656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8446545808565854e-05,
+      "loss": 3.8828,
+      "step": 1927168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.843815986105533e-05,
+      "loss": 3.8805,
+      "step": 1927680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.842977391354481e-05,
+      "loss": 3.8805,
+      "step": 1928192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8421404344838026e-05,
+      "loss": 3.881,
+      "step": 1928704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8413018397327506e-05,
+      "loss": 3.8919,
+      "step": 1929216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8404632449816983e-05,
+      "loss": 3.8846,
+      "step": 1929728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8396246502306463e-05,
+      "loss": 3.8851,
+      "step": 1930240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8387860554795943e-05,
+      "loss": 3.9,
+      "step": 1930752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8379474607285423e-05,
+      "loss": 3.8883,
+      "step": 1931264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8371088659774903e-05,
+      "loss": 3.8964,
+      "step": 1931776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8362702712264386e-05,
+      "loss": 3.8834,
+      "step": 1932288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8354316764753863e-05,
+      "loss": 3.863,
+      "step": 1932800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8345947196047075e-05,
+      "loss": 3.9015,
+      "step": 1933312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8337561248536555e-05,
+      "loss": 3.8838,
+      "step": 1933824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8329175301026035e-05,
+      "loss": 3.8881,
+      "step": 1934336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8320789353515515e-05,
+      "loss": 3.8833,
+      "step": 1934848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8312403406004995e-05,
+      "loss": 3.8858,
+      "step": 1935360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8304017458494475e-05,
+      "loss": 3.8753,
+      "step": 1935872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8295631510983955e-05,
+      "loss": 3.8768,
+      "step": 1936384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8287261942277164e-05,
+      "loss": 3.88,
+      "step": 1936896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8278875994766644e-05,
+      "loss": 3.8798,
+      "step": 1937408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8270506426059856e-05,
+      "loss": 3.8949,
+      "step": 1937920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8262120478549336e-05,
+      "loss": 3.8827,
+      "step": 1938432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8253734531038816e-05,
+      "loss": 3.876,
+      "step": 1938944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8245348583528296e-05,
+      "loss": 3.8957,
+      "step": 1939456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8236962636017776e-05,
+      "loss": 3.8705,
+      "step": 1939968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8228576688507256e-05,
+      "loss": 3.8749,
+      "step": 1940480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8220190740996736e-05,
+      "loss": 3.8885,
+      "step": 1940992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8211804793486216e-05,
+      "loss": 3.8852,
+      "step": 1941504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.820343522477943e-05,
+      "loss": 3.8817,
+      "step": 1942016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.819504927726891e-05,
+      "loss": 3.8856,
+      "step": 1942528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.818666332975839e-05,
+      "loss": 3.8646,
+      "step": 1943040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.817827738224787e-05,
+      "loss": 3.8776,
+      "step": 1943552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8169891434737345e-05,
+      "loss": 3.8992,
+      "step": 1944064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8161505487226825e-05,
+      "loss": 3.8792,
+      "step": 1944576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.815313591852004e-05,
+      "loss": 3.8943,
+      "step": 1945088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8144749971009517e-05,
+      "loss": 3.891,
+      "step": 1945600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8136364023498997e-05,
+      "loss": 3.9004,
+      "step": 1946112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.812799445479221e-05,
+      "loss": 3.8743,
+      "step": 1946624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.811960850728169e-05,
+      "loss": 3.8897,
+      "step": 1947136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.811122255977117e-05,
+      "loss": 3.8892,
+      "step": 1947648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.810283661226065e-05,
+      "loss": 3.8735,
+      "step": 1948160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.809445066475013e-05,
+      "loss": 3.8965,
+      "step": 1948672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.808606471723961e-05,
+      "loss": 3.8848,
+      "step": 1949184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.807767876972909e-05,
+      "loss": 3.8897,
+      "step": 1949696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.806929282221857e-05,
+      "loss": 3.8965,
+      "step": 1950208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8060906874708046e-05,
+      "loss": 3.8751,
+      "step": 1950720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8052520927197526e-05,
+      "loss": 3.8752,
+      "step": 1951232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.804413497968701e-05,
+      "loss": 3.8784,
+      "step": 1951744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.803574903217649e-05,
+      "loss": 3.8901,
+      "step": 1952256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.80273794634697e-05,
+      "loss": 3.8854,
+      "step": 1952768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.801900989476291e-05,
+      "loss": 3.8824,
+      "step": 1953280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.801062394725239e-05,
+      "loss": 3.8834,
+      "step": 1953792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.800223799974187e-05,
+      "loss": 3.8861,
+      "step": 1954304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.799385205223135e-05,
+      "loss": 3.8804,
+      "step": 1954816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.798546610472083e-05,
+      "loss": 3.8781,
+      "step": 1955328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.797708015721031e-05,
+      "loss": 3.8996,
+      "step": 1955840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.796869420969979e-05,
+      "loss": 3.8942,
+      "step": 1956352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7960324640993e-05,
+      "loss": 3.8883,
+      "step": 1956864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.795193869348248e-05,
+      "loss": 3.8759,
+      "step": 1957376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7943552745971963e-05,
+      "loss": 3.8874,
+      "step": 1957888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7935166798461443e-05,
+      "loss": 3.8788,
+      "step": 1958400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7926780850950923e-05,
+      "loss": 3.8988,
+      "step": 1958912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7918394903440403e-05,
+      "loss": 3.8857,
+      "step": 1959424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.791000895592988e-05,
+      "loss": 3.8786,
+      "step": 1959936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7901655766026825e-05,
+      "loss": 3.8874,
+      "step": 1960448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7893269818516305e-05,
+      "loss": 3.893,
+      "step": 1960960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7884883871005784e-05,
+      "loss": 3.8738,
+      "step": 1961472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7876497923495264e-05,
+      "loss": 3.8865,
+      "step": 1961984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7868111975984744e-05,
+      "loss": 3.8751,
+      "step": 1962496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7859726028474224e-05,
+      "loss": 3.8796,
+      "step": 1963008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.78513400809637e-05,
+      "loss": 3.8762,
+      "step": 1963520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.784295413345318e-05,
+      "loss": 3.8794,
+      "step": 1964032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7834568185942664e-05,
+      "loss": 3.883,
+      "step": 1964544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7826182238432144e-05,
+      "loss": 3.8888,
+      "step": 1965056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7817796290921624e-05,
+      "loss": 3.8767,
+      "step": 1965568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7809410343411104e-05,
+      "loss": 3.8778,
+      "step": 1966080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7801040774704313e-05,
+      "loss": 3.8932,
+      "step": 1966592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7792654827193793e-05,
+      "loss": 3.8992,
+      "step": 1967104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7784268879683273e-05,
+      "loss": 3.888,
+      "step": 1967616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7775882932172756e-05,
+      "loss": 3.8914,
+      "step": 1968128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7767513363465966e-05,
+      "loss": 3.8755,
+      "step": 1968640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7759127415955446e-05,
+      "loss": 3.8856,
+      "step": 1969152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7750741468444925e-05,
+      "loss": 3.8817,
+      "step": 1969664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7742371899738135e-05,
+      "loss": 3.8787,
+      "step": 1970176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7733985952227618e-05,
+      "loss": 3.8793,
+      "step": 1970688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7725600004717098e-05,
+      "loss": 3.8836,
+      "step": 1971200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7717214057206578e-05,
+      "loss": 3.8863,
+      "step": 1971712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7708828109696058e-05,
+      "loss": 3.881,
+      "step": 1972224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7700442162185534e-05,
+      "loss": 3.8886,
+      "step": 1972736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7692056214675014e-05,
+      "loss": 3.8863,
+      "step": 1973248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7683686645968227e-05,
+      "loss": 3.8782,
+      "step": 1973760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.767530069845771e-05,
+      "loss": 3.8867,
+      "step": 1974272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7666914750947187e-05,
+      "loss": 3.8927,
+      "step": 1974784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7658528803436667e-05,
+      "loss": 3.8916,
+      "step": 1975296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.765015923472988e-05,
+      "loss": 3.8833,
+      "step": 1975808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7641773287219356e-05,
+      "loss": 3.8953,
+      "step": 1976320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7633387339708836e-05,
+      "loss": 3.8883,
+      "step": 1976832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.762500139219832e-05,
+      "loss": 3.8891,
+      "step": 1977344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.76166154446878e-05,
+      "loss": 3.8873,
+      "step": 1977856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.760822949717728e-05,
+      "loss": 3.8836,
+      "step": 1978368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.759984354966676e-05,
+      "loss": 3.8824,
+      "step": 1978880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.759145760215624e-05,
+      "loss": 3.8933,
+      "step": 1979392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7583088033449448e-05,
+      "loss": 3.8761,
+      "step": 1979904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.757471846474266e-05,
+      "loss": 3.8851,
+      "step": 1980416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.756633251723214e-05,
+      "loss": 3.89,
+      "step": 1980928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.755794656972162e-05,
+      "loss": 3.8896,
+      "step": 1981440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.75495606222111e-05,
+      "loss": 3.8819,
+      "step": 1981952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.754119105350431e-05,
+      "loss": 3.8838,
+      "step": 1982464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.753280510599379e-05,
+      "loss": 3.8918,
+      "step": 1982976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7524419158483273e-05,
+      "loss": 3.8854,
+      "step": 1983488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7516033210972753e-05,
+      "loss": 3.8901,
+      "step": 1984000
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.983999013900757,
+      "eval_runtime": 278.8267,
+      "eval_samples_per_second": 1368.56,
+      "eval_steps_per_second": 42.768,
+      "step": 1984308
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7507647263462233e-05,
+      "loss": 3.8677,
+      "step": 1984512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7499261315951713e-05,
+      "loss": 3.8786,
+      "step": 1985024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.749087536844119e-05,
+      "loss": 3.8868,
+      "step": 1985536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.748248942093067e-05,
+      "loss": 3.8855,
+      "step": 1986048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.747410347342015e-05,
+      "loss": 3.8954,
+      "step": 1986560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.746571752590963e-05,
+      "loss": 3.8796,
+      "step": 1987072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7457331578399112e-05,
+      "loss": 3.8868,
+      "step": 1987584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7448945630888592e-05,
+      "loss": 3.8828,
+      "step": 1988096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.744055968337807e-05,
+      "loss": 3.8773,
+      "step": 1988608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.743217373586755e-05,
+      "loss": 3.881,
+      "step": 1989120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.742378778835703e-05,
+      "loss": 3.8905,
+      "step": 1989632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.741541821965024e-05,
+      "loss": 3.8934,
+      "step": 1990144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.740703227213972e-05,
+      "loss": 3.8893,
+      "step": 1990656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.73986463246292e-05,
+      "loss": 3.8672,
+      "step": 1991168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.739026037711868e-05,
+      "loss": 3.8906,
+      "step": 1991680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.738187442960816e-05,
+      "loss": 3.8708,
+      "step": 1992192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.737348848209764e-05,
+      "loss": 3.883,
+      "step": 1992704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.736511891339085e-05,
+      "loss": 3.8813,
+      "step": 1993216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7356732965880333e-05,
+      "loss": 3.8846,
+      "step": 1993728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7348347018369813e-05,
+      "loss": 3.8865,
+      "step": 1994240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7339961070859293e-05,
+      "loss": 3.8984,
+      "step": 1994752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7331575123348773e-05,
+      "loss": 3.8878,
+      "step": 1995264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.732318917583825e-05,
+      "loss": 3.8832,
+      "step": 1995776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7314819607131462e-05,
+      "loss": 3.879,
+      "step": 1996288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7306433659620942e-05,
+      "loss": 3.8897,
+      "step": 1996800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7298047712110426e-05,
+      "loss": 3.8839,
+      "step": 1997312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7289661764599902e-05,
+      "loss": 3.8804,
+      "step": 1997824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7281275817089382e-05,
+      "loss": 3.8805,
+      "step": 1998336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7272889869578862e-05,
+      "loss": 3.8801,
+      "step": 1998848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7264503922068342e-05,
+      "loss": 3.8797,
+      "step": 1999360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7256117974557822e-05,
+      "loss": 3.8704,
+      "step": 1999872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7247732027047302e-05,
+      "loss": 3.8949,
+      "step": 2000384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7239346079536782e-05,
+      "loss": 3.8945,
+      "step": 2000896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7230960132026262e-05,
+      "loss": 3.8864,
+      "step": 2001408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7222574184515742e-05,
+      "loss": 3.8818,
+      "step": 2001920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7214204615808954e-05,
+      "loss": 3.8821,
+      "step": 2002432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.720581866829843e-05,
+      "loss": 3.889,
+      "step": 2002944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.719743272078791e-05,
+      "loss": 3.8809,
+      "step": 2003456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.718904677327739e-05,
+      "loss": 3.8753,
+      "step": 2003968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7180660825766874e-05,
+      "loss": 3.8768,
+      "step": 2004480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7172274878256354e-05,
+      "loss": 3.8779,
+      "step": 2004992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7163888930745834e-05,
+      "loss": 3.8853,
+      "step": 2005504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7155502983235314e-05,
+      "loss": 3.8755,
+      "step": 2006016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.714711703572479e-05,
+      "loss": 3.8811,
+      "step": 2006528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7138747467018003e-05,
+      "loss": 3.8933,
+      "step": 2007040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7130361519507483e-05,
+      "loss": 3.8814,
+      "step": 2007552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7121975571996966e-05,
+      "loss": 3.8894,
+      "step": 2008064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7113606003290175e-05,
+      "loss": 3.8819,
+      "step": 2008576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7105220055779655e-05,
+      "loss": 3.8605,
+      "step": 2009088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7096834108269135e-05,
+      "loss": 3.8922,
+      "step": 2009600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7088448160758612e-05,
+      "loss": 3.879,
+      "step": 2010112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7080062213248095e-05,
+      "loss": 3.8815,
+      "step": 2010624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7071676265737575e-05,
+      "loss": 3.8828,
+      "step": 2011136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7063290318227055e-05,
+      "loss": 3.8791,
+      "step": 2011648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7054904370716535e-05,
+      "loss": 3.8677,
+      "step": 2012160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7046518423206015e-05,
+      "loss": 3.8764,
+      "step": 2012672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7038132475695495e-05,
+      "loss": 3.8754,
+      "step": 2013184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.702974652818497e-05,
+      "loss": 3.8738,
+      "step": 2013696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.702136058067445e-05,
+      "loss": 3.8903,
+      "step": 2014208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7012974633163935e-05,
+      "loss": 3.8774,
+      "step": 2014720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7004605064457147e-05,
+      "loss": 3.8722,
+      "step": 2015232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6996219116946624e-05,
+      "loss": 3.893,
+      "step": 2015744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6987849548239836e-05,
+      "loss": 3.8661,
+      "step": 2016256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6979463600729316e-05,
+      "loss": 3.8688,
+      "step": 2016768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6971077653218796e-05,
+      "loss": 3.8849,
+      "step": 2017280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6962691705708276e-05,
+      "loss": 3.878,
+      "step": 2017792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6954305758197756e-05,
+      "loss": 3.8737,
+      "step": 2018304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6945919810687236e-05,
+      "loss": 3.8806,
+      "step": 2018816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6937533863176716e-05,
+      "loss": 3.8623,
+      "step": 2019328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6929147915666196e-05,
+      "loss": 3.8702,
+      "step": 2019840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6920778346959405e-05,
+      "loss": 3.893,
+      "step": 2020352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.691239239944889e-05,
+      "loss": 3.8751,
+      "step": 2020864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.690400645193837e-05,
+      "loss": 3.8905,
+      "step": 2021376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.689562050442785e-05,
+      "loss": 3.8814,
+      "step": 2021888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6887234556917325e-05,
+      "loss": 3.9005,
+      "step": 2022400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6878881367014267e-05,
+      "loss": 3.8676,
+      "step": 2022912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6870511798307483e-05,
+      "loss": 3.8895,
+      "step": 2023424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6862125850796963e-05,
+      "loss": 3.8844,
+      "step": 2023936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6853739903286443e-05,
+      "loss": 3.8656,
+      "step": 2024448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.684535395577592e-05,
+      "loss": 3.8918,
+      "step": 2024960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.68369680082654e-05,
+      "loss": 3.8813,
+      "step": 2025472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.682858206075488e-05,
+      "loss": 3.8841,
+      "step": 2025984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.682019611324436e-05,
+      "loss": 3.8922,
+      "step": 2026496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6811810165733842e-05,
+      "loss": 3.87,
+      "step": 2027008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6803424218223322e-05,
+      "loss": 3.8691,
+      "step": 2027520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.67950382707128e-05,
+      "loss": 3.8726,
+      "step": 2028032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.678666870200601e-05,
+      "loss": 3.8857,
+      "step": 2028544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.677828275449549e-05,
+      "loss": 3.8783,
+      "step": 2029056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.676989680698497e-05,
+      "loss": 3.8753,
+      "step": 2029568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.676151085947445e-05,
+      "loss": 3.8777,
+      "step": 2030080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.675312491196393e-05,
+      "loss": 3.8828,
+      "step": 2030592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.674473896445341e-05,
+      "loss": 3.8778,
+      "step": 2031104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.673635301694289e-05,
+      "loss": 3.8731,
+      "step": 2031616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.67279834482361e-05,
+      "loss": 3.8939,
+      "step": 2032128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.671959750072558e-05,
+      "loss": 3.8883,
+      "step": 2032640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.671121155321506e-05,
+      "loss": 3.8852,
+      "step": 2033152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6702825605704543e-05,
+      "loss": 3.87,
+      "step": 2033664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6694439658194023e-05,
+      "loss": 3.8821,
+      "step": 2034176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6686053710683503e-05,
+      "loss": 3.8729,
+      "step": 2034688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.667766776317298e-05,
+      "loss": 3.8899,
+      "step": 2035200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.666928181566246e-05,
+      "loss": 3.8803,
+      "step": 2035712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6660912246955672e-05,
+      "loss": 3.8749,
+      "step": 2036224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6652526299445152e-05,
+      "loss": 3.8805,
+      "step": 2036736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6644140351934632e-05,
+      "loss": 3.8852,
+      "step": 2037248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6635754404424112e-05,
+      "loss": 3.8768,
+      "step": 2037760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6627401214521054e-05,
+      "loss": 3.8817,
+      "step": 2038272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6619015267010534e-05,
+      "loss": 3.8701,
+      "step": 2038784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6610629319500014e-05,
+      "loss": 3.8754,
+      "step": 2039296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6602243371989497e-05,
+      "loss": 3.8727,
+      "step": 2039808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6593857424478977e-05,
+      "loss": 3.8719,
+      "step": 2040320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6585471476968454e-05,
+      "loss": 3.8765,
+      "step": 2040832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6577085529457934e-05,
+      "loss": 3.8807,
+      "step": 2041344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6568715960751146e-05,
+      "loss": 3.8754,
+      "step": 2041856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6560330013240626e-05,
+      "loss": 3.8722,
+      "step": 2042368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6551944065730106e-05,
+      "loss": 3.8897,
+      "step": 2042880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6543558118219586e-05,
+      "loss": 3.892,
+      "step": 2043392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.65351885495128e-05,
+      "loss": 3.8848,
+      "step": 2043904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.652680260200228e-05,
+      "loss": 3.8876,
+      "step": 2044416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6518416654491755e-05,
+      "loss": 3.8722,
+      "step": 2044928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6510030706981235e-05,
+      "loss": 3.881,
+      "step": 2045440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6501644759470715e-05,
+      "loss": 3.8787,
+      "step": 2045952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6493275190763927e-05,
+      "loss": 3.8686,
+      "step": 2046464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6484889243253407e-05,
+      "loss": 3.8778,
+      "step": 2046976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6476503295742887e-05,
+      "loss": 3.878,
+      "step": 2047488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6468117348232367e-05,
+      "loss": 3.8833,
+      "step": 2048000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6459731400721847e-05,
+      "loss": 3.8787,
+      "step": 2048512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.645136183201506e-05,
+      "loss": 3.8812,
+      "step": 2049024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.644297588450454e-05,
+      "loss": 3.88,
+      "step": 2049536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.643458993699402e-05,
+      "loss": 3.8724,
+      "step": 2050048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.64262039894835e-05,
+      "loss": 3.8811,
+      "step": 2050560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.641781804197298e-05,
+      "loss": 3.8858,
+      "step": 2051072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.640943209446246e-05,
+      "loss": 3.8873,
+      "step": 2051584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6401046146951936e-05,
+      "loss": 3.8755,
+      "step": 2052096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6392676578245152e-05,
+      "loss": 3.8936,
+      "step": 2052608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6384290630734632e-05,
+      "loss": 3.881,
+      "step": 2053120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.637590468322411e-05,
+      "loss": 3.8863,
+      "step": 2053632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6367518735713588e-05,
+      "loss": 3.8862,
+      "step": 2054144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6359132788203068e-05,
+      "loss": 3.8762,
+      "step": 2054656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6350746840692548e-05,
+      "loss": 3.8756,
+      "step": 2055168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.634237727198576e-05,
+      "loss": 3.8921,
+      "step": 2055680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.633399132447524e-05,
+      "loss": 3.8682,
+      "step": 2056192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.632560537696472e-05,
+      "loss": 3.8811,
+      "step": 2056704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.63172194294542e-05,
+      "loss": 3.8849,
+      "step": 2057216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.630883348194368e-05,
+      "loss": 3.8858,
+      "step": 2057728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.630044753443316e-05,
+      "loss": 3.877,
+      "step": 2058240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.629207796572637e-05,
+      "loss": 3.8804,
+      "step": 2058752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6283692018215853e-05,
+      "loss": 3.8842,
+      "step": 2059264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6275306070705333e-05,
+      "loss": 3.8848,
+      "step": 2059776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6266920123194813e-05,
+      "loss": 3.8877,
+      "step": 2060288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9821367263793945,
+      "eval_runtime": 291.8473,
+      "eval_samples_per_second": 1307.502,
+      "eval_steps_per_second": 40.86,
+      "step": 2060628
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.625853417568429e-05,
+      "loss": 3.8685,
+      "step": 2060800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.625014822817377e-05,
+      "loss": 3.8774,
+      "step": 2061312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.624176228066325e-05,
+      "loss": 3.879,
+      "step": 2061824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.623337633315273e-05,
+      "loss": 3.879,
+      "step": 2062336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6225006764445942e-05,
+      "loss": 3.8937,
+      "step": 2062848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.621662081693542e-05,
+      "loss": 3.8775,
+      "step": 2063360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6208251248228634e-05,
+      "loss": 3.879,
+      "step": 2063872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.619986530071811e-05,
+      "loss": 3.8794,
+      "step": 2064384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.619147935320759e-05,
+      "loss": 3.8718,
+      "step": 2064896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.618309340569707e-05,
+      "loss": 3.8744,
+      "step": 2065408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6174723836990287e-05,
+      "loss": 3.8837,
+      "step": 2065920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6166337889479763e-05,
+      "loss": 3.8898,
+      "step": 2066432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6157951941969243e-05,
+      "loss": 3.8849,
+      "step": 2066944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6149565994458723e-05,
+      "loss": 3.864,
+      "step": 2067456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6141196425751935e-05,
+      "loss": 3.8817,
+      "step": 2067968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6132810478241415e-05,
+      "loss": 3.8661,
+      "step": 2068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6124424530730895e-05,
+      "loss": 3.879,
+      "step": 2068992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6116038583220375e-05,
+      "loss": 3.877,
+      "step": 2069504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6107652635709855e-05,
+      "loss": 3.8815,
+      "step": 2070016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6099266688199335e-05,
+      "loss": 3.8794,
+      "step": 2070528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6090897119492544e-05,
+      "loss": 3.8914,
+      "step": 2071040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6082511171982024e-05,
+      "loss": 3.882,
+      "step": 2071552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6074125224471508e-05,
+      "loss": 3.8842,
+      "step": 2072064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6065739276960988e-05,
+      "loss": 3.8737,
+      "step": 2072576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6057353329450468e-05,
+      "loss": 3.8826,
+      "step": 2073088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6048967381939944e-05,
+      "loss": 3.8819,
+      "step": 2073600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6040581434429424e-05,
+      "loss": 3.8741,
+      "step": 2074112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6032195486918904e-05,
+      "loss": 3.8775,
+      "step": 2074624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6023809539408384e-05,
+      "loss": 3.8758,
+      "step": 2075136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6015439970701596e-05,
+      "loss": 3.8751,
+      "step": 2075648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6007054023191076e-05,
+      "loss": 3.8717,
+      "step": 2076160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5998668075680556e-05,
+      "loss": 3.8815,
+      "step": 2076672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5990282128170036e-05,
+      "loss": 3.8893,
+      "step": 2077184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5981912559463245e-05,
+      "loss": 3.8813,
+      "step": 2077696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.597352661195273e-05,
+      "loss": 3.8793,
+      "step": 2078208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.596515704324594e-05,
+      "loss": 3.8793,
+      "step": 2078720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5956771095735418e-05,
+      "loss": 3.8858,
+      "step": 2079232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5948385148224898e-05,
+      "loss": 3.874,
+      "step": 2079744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5939999200714378e-05,
+      "loss": 3.8704,
+      "step": 2080256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.593162963200759e-05,
+      "loss": 3.8763,
+      "step": 2080768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.592324368449707e-05,
+      "loss": 3.8698,
+      "step": 2081280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.591485773698655e-05,
+      "loss": 3.8786,
+      "step": 2081792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.590647178947603e-05,
+      "loss": 3.873,
+      "step": 2082304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.589810222076924e-05,
+      "loss": 3.8798,
+      "step": 2082816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.588971627325872e-05,
+      "loss": 3.8844,
+      "step": 2083328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.58813303257482e-05,
+      "loss": 3.8842,
+      "step": 2083840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.587294437823768e-05,
+      "loss": 3.8852,
+      "step": 2084352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5864558430727162e-05,
+      "loss": 3.875,
+      "step": 2084864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5856172483216642e-05,
+      "loss": 3.8543,
+      "step": 2085376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5847786535706122e-05,
+      "loss": 3.8859,
+      "step": 2085888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.58394005881956e-05,
+      "loss": 3.8715,
+      "step": 2086400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.583101464068508e-05,
+      "loss": 3.8804,
+      "step": 2086912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.582262869317456e-05,
+      "loss": 3.8775,
+      "step": 2087424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.581424274566404e-05,
+      "loss": 3.8748,
+      "step": 2087936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.580587317695725e-05,
+      "loss": 3.8634,
+      "step": 2088448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.579748722944673e-05,
+      "loss": 3.8663,
+      "step": 2088960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.578910128193621e-05,
+      "loss": 3.8746,
+      "step": 2089472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.578071533442569e-05,
+      "loss": 3.8661,
+      "step": 2089984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.577232938691517e-05,
+      "loss": 3.8867,
+      "step": 2090496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5763959818208384e-05,
+      "loss": 3.8724,
+      "step": 2091008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5755573870697864e-05,
+      "loss": 3.8644,
+      "step": 2091520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5747187923187343e-05,
+      "loss": 3.8875,
+      "step": 2092032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5738801975676823e-05,
+      "loss": 3.8655,
+      "step": 2092544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5730416028166303e-05,
+      "loss": 3.8629,
+      "step": 2093056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.572203008065578e-05,
+      "loss": 3.8813,
+      "step": 2093568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5713660511948992e-05,
+      "loss": 3.873,
+      "step": 2094080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5705274564438476e-05,
+      "loss": 3.8686,
+      "step": 2094592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5696888616927952e-05,
+      "loss": 3.874,
+      "step": 2095104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5688502669417432e-05,
+      "loss": 3.8582,
+      "step": 2095616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5680116721906912e-05,
+      "loss": 3.8703,
+      "step": 2096128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5671730774396392e-05,
+      "loss": 3.8848,
+      "step": 2096640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.56633612056896e-05,
+      "loss": 3.8742,
+      "step": 2097152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5654991636982817e-05,
+      "loss": 3.8839,
+      "step": 2097664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5646605689472297e-05,
+      "loss": 3.8768,
+      "step": 2098176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5638219741961777e-05,
+      "loss": 3.8948,
+      "step": 2098688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5629833794451254e-05,
+      "loss": 3.8652,
+      "step": 2099200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5621447846940734e-05,
+      "loss": 3.8846,
+      "step": 2099712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5613061899430214e-05,
+      "loss": 3.8821,
+      "step": 2100224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5604675951919694e-05,
+      "loss": 3.8619,
+      "step": 2100736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5596290004409177e-05,
+      "loss": 3.8809,
+      "step": 2101248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5587904056898657e-05,
+      "loss": 3.8814,
+      "step": 2101760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5579518109388133e-05,
+      "loss": 3.8829,
+      "step": 2102272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5571132161877613e-05,
+      "loss": 3.8866,
+      "step": 2102784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5562746214367093e-05,
+      "loss": 3.8678,
+      "step": 2103296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5554376645660306e-05,
+      "loss": 3.8602,
+      "step": 2103808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5546007076953518e-05,
+      "loss": 3.8699,
+      "step": 2104320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5537621129442998e-05,
+      "loss": 3.8807,
+      "step": 2104832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5529235181932478e-05,
+      "loss": 3.8737,
+      "step": 2105344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5520849234421958e-05,
+      "loss": 3.8708,
+      "step": 2105856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5512463286911435e-05,
+      "loss": 3.87,
+      "step": 2106368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5504093718204647e-05,
+      "loss": 3.8768,
+      "step": 2106880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.549570777069413e-05,
+      "loss": 3.8724,
+      "step": 2107392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5487321823183607e-05,
+      "loss": 3.8712,
+      "step": 2107904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5478935875673087e-05,
+      "loss": 3.8839,
+      "step": 2108416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5470549928162567e-05,
+      "loss": 3.8901,
+      "step": 2108928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5462163980652047e-05,
+      "loss": 3.8777,
+      "step": 2109440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5453794411945256e-05,
+      "loss": 3.8701,
+      "step": 2109952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.544540846443474e-05,
+      "loss": 3.8745,
+      "step": 2110464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.543702251692422e-05,
+      "loss": 3.8699,
+      "step": 2110976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.54286365694137e-05,
+      "loss": 3.8851,
+      "step": 2111488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.542026700070691e-05,
+      "loss": 3.8771,
+      "step": 2112000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.541188105319639e-05,
+      "loss": 3.8732,
+      "step": 2112512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.54035114844896e-05,
+      "loss": 3.8771,
+      "step": 2113024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.539512553697908e-05,
+      "loss": 3.877,
+      "step": 2113536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.538673958946856e-05,
+      "loss": 3.8762,
+      "step": 2114048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.537835364195804e-05,
+      "loss": 3.874,
+      "step": 2114560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.536996769444752e-05,
+      "loss": 3.8678,
+      "step": 2115072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5361581746937e-05,
+      "loss": 3.8658,
+      "step": 2115584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.535319579942648e-05,
+      "loss": 3.8737,
+      "step": 2116096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.534480985191596e-05,
+      "loss": 3.8643,
+      "step": 2116608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.533642390440544e-05,
+      "loss": 3.8741,
+      "step": 2117120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.532803795689492e-05,
+      "loss": 3.8727,
+      "step": 2117632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5319668388188133e-05,
+      "loss": 3.8721,
+      "step": 2118144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5311282440677613e-05,
+      "loss": 3.8652,
+      "step": 2118656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.530289649316709e-05,
+      "loss": 3.8883,
+      "step": 2119168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5294526924460302e-05,
+      "loss": 3.889,
+      "step": 2119680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5286140976949785e-05,
+      "loss": 3.8762,
+      "step": 2120192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5277755029439262e-05,
+      "loss": 3.8801,
+      "step": 2120704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5269369081928742e-05,
+      "loss": 3.8721,
+      "step": 2121216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5260983134418222e-05,
+      "loss": 3.8742,
+      "step": 2121728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5252597186907702e-05,
+      "loss": 3.8731,
+      "step": 2122240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5244211239397183e-05,
+      "loss": 3.8638,
+      "step": 2122752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5235825291886663e-05,
+      "loss": 3.8728,
+      "step": 2123264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5227439344376143e-05,
+      "loss": 3.8758,
+      "step": 2123776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5219053396865622e-05,
+      "loss": 3.8807,
+      "step": 2124288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5210667449355102e-05,
+      "loss": 3.8749,
+      "step": 2124800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5202297880648314e-05,
+      "loss": 3.8762,
+      "step": 2125312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5193911933137792e-05,
+      "loss": 3.8764,
+      "step": 2125824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5185542364431005e-05,
+      "loss": 3.8678,
+      "step": 2126336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5177156416920485e-05,
+      "loss": 3.8767,
+      "step": 2126848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5168770469409965e-05,
+      "loss": 3.8806,
+      "step": 2127360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5160384521899443e-05,
+      "loss": 3.8842,
+      "step": 2127872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5151998574388923e-05,
+      "loss": 3.8684,
+      "step": 2128384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5143612626878403e-05,
+      "loss": 3.8923,
+      "step": 2128896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5135226679367884e-05,
+      "loss": 3.8754,
+      "step": 2129408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5126857110661094e-05,
+      "loss": 3.8862,
+      "step": 2129920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5118471163150574e-05,
+      "loss": 3.8794,
+      "step": 2130432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5110085215640055e-05,
+      "loss": 3.8764,
+      "step": 2130944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5101699268129535e-05,
+      "loss": 3.8685,
+      "step": 2131456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5093313320619015e-05,
+      "loss": 3.8904,
+      "step": 2131968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5084927373108495e-05,
+      "loss": 3.8619,
+      "step": 2132480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5076557804401706e-05,
+      "loss": 3.8776,
+      "step": 2132992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5068171856891186e-05,
+      "loss": 3.8777,
+      "step": 2133504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5059785909380666e-05,
+      "loss": 3.8783,
+      "step": 2134016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5051399961870147e-05,
+      "loss": 3.8747,
+      "step": 2134528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5043014014359624e-05,
+      "loss": 3.8744,
+      "step": 2135040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5034644445652838e-05,
+      "loss": 3.8772,
+      "step": 2135552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5026258498142318e-05,
+      "loss": 3.8881,
+      "step": 2136064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5017872550631798e-05,
+      "loss": 3.8779,
+      "step": 2136576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.980630397796631,
+      "eval_runtime": 292.0131,
+      "eval_samples_per_second": 1306.76,
+      "eval_steps_per_second": 40.837,
+      "step": 2136948
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5009486603121276e-05,
+      "loss": 3.8538,
+      "step": 2137088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5001100655610756e-05,
+      "loss": 3.8716,
+      "step": 2137600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4992714708100236e-05,
+      "loss": 3.8745,
+      "step": 2138112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4984328760589716e-05,
+      "loss": 3.8787,
+      "step": 2138624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4975942813079196e-05,
+      "loss": 3.8849,
+      "step": 2139136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4967556865568678e-05,
+      "loss": 3.8723,
+      "step": 2139648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4959170918058154e-05,
+      "loss": 3.8795,
+      "step": 2140160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4950784970547634e-05,
+      "loss": 3.8753,
+      "step": 2140672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4942399023037116e-05,
+      "loss": 3.8687,
+      "step": 2141184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4934013075526596e-05,
+      "loss": 3.8638,
+      "step": 2141696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4925627128016076e-05,
+      "loss": 3.8849,
+      "step": 2142208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4917241180505556e-05,
+      "loss": 3.8876,
+      "step": 2142720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4908871611798767e-05,
+      "loss": 3.8792,
+      "step": 2143232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4900485664288247e-05,
+      "loss": 3.8592,
+      "step": 2143744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4892099716777727e-05,
+      "loss": 3.8782,
+      "step": 2144256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4883713769267208e-05,
+      "loss": 3.8616,
+      "step": 2144768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4875327821756688e-05,
+      "loss": 3.8793,
+      "step": 2145280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4866941874246165e-05,
+      "loss": 3.8736,
+      "step": 2145792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4858555926735646e-05,
+      "loss": 3.8728,
+      "step": 2146304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4850169979225126e-05,
+      "loss": 3.8762,
+      "step": 2146816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4841784031714606e-05,
+      "loss": 3.8832,
+      "step": 2147328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4833398084204086e-05,
+      "loss": 3.8754,
+      "step": 2147840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4825012136693566e-05,
+      "loss": 3.8788,
+      "step": 2148352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4816626189183044e-05,
+      "loss": 3.8743,
+      "step": 2148864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4808240241672524e-05,
+      "loss": 3.8786,
+      "step": 2149376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4799854294162004e-05,
+      "loss": 3.8801,
+      "step": 2149888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4791484725455219e-05,
+      "loss": 3.866,
+      "step": 2150400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4783115156748428e-05,
+      "loss": 3.8708,
+      "step": 2150912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.477472920923791e-05,
+      "loss": 3.8756,
+      "step": 2151424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.476634326172739e-05,
+      "loss": 3.8698,
+      "step": 2151936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.475795731421687e-05,
+      "loss": 3.8686,
+      "step": 2152448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4749571366706347e-05,
+      "loss": 3.8782,
+      "step": 2152960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4741185419195827e-05,
+      "loss": 3.8797,
+      "step": 2153472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.473281585048904e-05,
+      "loss": 3.8815,
+      "step": 2153984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4724429902978518e-05,
+      "loss": 3.8741,
+      "step": 2154496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4716076713075461e-05,
+      "loss": 3.8749,
+      "step": 2155008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4707690765564941e-05,
+      "loss": 3.8835,
+      "step": 2155520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4699304818054421e-05,
+      "loss": 3.8668,
+      "step": 2156032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4690918870543901e-05,
+      "loss": 3.8679,
+      "step": 2156544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4682532923033381e-05,
+      "loss": 3.8703,
+      "step": 2157056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4674146975522863e-05,
+      "loss": 3.8665,
+      "step": 2157568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4665761028012343e-05,
+      "loss": 3.8716,
+      "step": 2158080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.465737508050182e-05,
+      "loss": 3.8693,
+      "step": 2158592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4648989132991301e-05,
+      "loss": 3.8749,
+      "step": 2159104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4640603185480781e-05,
+      "loss": 3.8775,
+      "step": 2159616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4632217237970261e-05,
+      "loss": 3.8808,
+      "step": 2160128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4623831290459741e-05,
+      "loss": 3.8777,
+      "step": 2160640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4615461721752952e-05,
+      "loss": 3.8766,
+      "step": 2161152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4607075774242432e-05,
+      "loss": 3.8497,
+      "step": 2161664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4598689826731912e-05,
+      "loss": 3.8811,
+      "step": 2162176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4590320258025122e-05,
+      "loss": 3.8673,
+      "step": 2162688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4581934310514602e-05,
+      "loss": 3.8754,
+      "step": 2163200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4573548363004082e-05,
+      "loss": 3.8705,
+      "step": 2163712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4565162415493564e-05,
+      "loss": 3.8713,
+      "step": 2164224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4556776467983044e-05,
+      "loss": 3.8614,
+      "step": 2164736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.454839052047252e-05,
+      "loss": 3.8642,
+      "step": 2165248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4540004572962002e-05,
+      "loss": 3.871,
+      "step": 2165760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4531618625451482e-05,
+      "loss": 3.8611,
+      "step": 2166272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4523232677940962e-05,
+      "loss": 3.8812,
+      "step": 2166784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4514863109234173e-05,
+      "loss": 3.8721,
+      "step": 2167296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4506477161723653e-05,
+      "loss": 3.8589,
+      "step": 2167808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4498091214213133e-05,
+      "loss": 3.8815,
+      "step": 2168320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4489705266702613e-05,
+      "loss": 3.8621,
+      "step": 2168832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4481319319192094e-05,
+      "loss": 3.8601,
+      "step": 2169344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4472933371681574e-05,
+      "loss": 3.8756,
+      "step": 2169856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4464563802974783e-05,
+      "loss": 3.8672,
+      "step": 2170368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4456177855464265e-05,
+      "loss": 3.8635,
+      "step": 2170880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4447791907953745e-05,
+      "loss": 3.8724,
+      "step": 2171392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4439405960443225e-05,
+      "loss": 3.8592,
+      "step": 2171904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4431020012932703e-05,
+      "loss": 3.8607,
+      "step": 2172416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4422634065422183e-05,
+      "loss": 3.8799,
+      "step": 2172928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4414264496715396e-05,
+      "loss": 3.8725,
+      "step": 2173440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4405878549204876e-05,
+      "loss": 3.878,
+      "step": 2173952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4397492601694354e-05,
+      "loss": 3.8748,
+      "step": 2174464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4389106654183834e-05,
+      "loss": 3.8877,
+      "step": 2174976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4380720706673314e-05,
+      "loss": 3.859,
+      "step": 2175488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4372334759162796e-05,
+      "loss": 3.8815,
+      "step": 2176000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4363948811652275e-05,
+      "loss": 3.8721,
+      "step": 2176512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4355562864141755e-05,
+      "loss": 3.8633,
+      "step": 2177024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4347193295434966e-05,
+      "loss": 3.8731,
+      "step": 2177536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4338807347924446e-05,
+      "loss": 3.8804,
+      "step": 2178048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4330421400413926e-05,
+      "loss": 3.876,
+      "step": 2178560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4322051831707137e-05,
+      "loss": 3.8862,
+      "step": 2179072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4313665884196617e-05,
+      "loss": 3.8638,
+      "step": 2179584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4305279936686097e-05,
+      "loss": 3.8549,
+      "step": 2180096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4296893989175578e-05,
+      "loss": 3.8649,
+      "step": 2180608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4288508041665058e-05,
+      "loss": 3.8778,
+      "step": 2181120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4280122094154535e-05,
+      "loss": 3.8735,
+      "step": 2181632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4271736146644017e-05,
+      "loss": 3.8664,
+      "step": 2182144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4263350199133497e-05,
+      "loss": 3.8684,
+      "step": 2182656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4254980630426709e-05,
+      "loss": 3.8734,
+      "step": 2183168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.424661106171992e-05,
+      "loss": 3.8664,
+      "step": 2183680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.42382251142094e-05,
+      "loss": 3.8658,
+      "step": 2184192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.422983916669888e-05,
+      "loss": 3.8783,
+      "step": 2184704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4221453219188358e-05,
+      "loss": 3.8826,
+      "step": 2185216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.421308365048157e-05,
+      "loss": 3.8752,
+      "step": 2185728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.420469770297105e-05,
+      "loss": 3.8652,
+      "step": 2186240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.419631175546053e-05,
+      "loss": 3.8705,
+      "step": 2186752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4187925807950009e-05,
+      "loss": 3.8647,
+      "step": 2187264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4179539860439489e-05,
+      "loss": 3.8799,
+      "step": 2187776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4171170291732703e-05,
+      "loss": 3.8753,
+      "step": 2188288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4162784344222183e-05,
+      "loss": 3.8679,
+      "step": 2188800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.415439839671166e-05,
+      "loss": 3.8752,
+      "step": 2189312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4146012449201141e-05,
+      "loss": 3.8683,
+      "step": 2189824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4137626501690621e-05,
+      "loss": 3.8712,
+      "step": 2190336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4129240554180101e-05,
+      "loss": 3.8687,
+      "step": 2190848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4120854606669581e-05,
+      "loss": 3.8676,
+      "step": 2191360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.411246865915906e-05,
+      "loss": 3.8612,
+      "step": 2191872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4104099090452272e-05,
+      "loss": 3.8681,
+      "step": 2192384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4095713142941752e-05,
+      "loss": 3.8676,
+      "step": 2192896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4087327195431233e-05,
+      "loss": 3.8644,
+      "step": 2193408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4078941247920713e-05,
+      "loss": 3.8675,
+      "step": 2193920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.407055530041019e-05,
+      "loss": 3.8684,
+      "step": 2194432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4062169352899671e-05,
+      "loss": 3.8608,
+      "step": 2194944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4053783405389151e-05,
+      "loss": 3.8822,
+      "step": 2195456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.404541383668236e-05,
+      "loss": 3.8856,
+      "step": 2195968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4037027889171842e-05,
+      "loss": 3.8722,
+      "step": 2196480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4028641941661322e-05,
+      "loss": 3.8764,
+      "step": 2196992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4020255994150802e-05,
+      "loss": 3.8703,
+      "step": 2197504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4011870046640282e-05,
+      "loss": 3.8702,
+      "step": 2198016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4003484099129764e-05,
+      "loss": 3.8739,
+      "step": 2198528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3995114530422973e-05,
+      "loss": 3.8555,
+      "step": 2199040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3986728582912453e-05,
+      "loss": 3.8704,
+      "step": 2199552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3978359014205663e-05,
+      "loss": 3.8688,
+      "step": 2200064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3969973066695143e-05,
+      "loss": 3.8777,
+      "step": 2200576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3961587119184625e-05,
+      "loss": 3.8675,
+      "step": 2201088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3953201171674105e-05,
+      "loss": 3.8755,
+      "step": 2201600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3944815224163585e-05,
+      "loss": 3.8704,
+      "step": 2202112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3936429276653065e-05,
+      "loss": 3.8633,
+      "step": 2202624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3928043329142543e-05,
+      "loss": 3.8768,
+      "step": 2203136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3919657381632023e-05,
+      "loss": 3.8745,
+      "step": 2203648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3911271434121503e-05,
+      "loss": 3.8794,
+      "step": 2204160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3902901865414716e-05,
+      "loss": 3.8657,
+      "step": 2204672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3894515917904194e-05,
+      "loss": 3.8836,
+      "step": 2205184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3886129970393674e-05,
+      "loss": 3.8738,
+      "step": 2205696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3877744022883154e-05,
+      "loss": 3.8775,
+      "step": 2206208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3869358075372635e-05,
+      "loss": 3.877,
+      "step": 2206720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3860972127862115e-05,
+      "loss": 3.8708,
+      "step": 2207232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3852586180351595e-05,
+      "loss": 3.8655,
+      "step": 2207744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3844216611644806e-05,
+      "loss": 3.8838,
+      "step": 2208256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3835830664134286e-05,
+      "loss": 3.86,
+      "step": 2208768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3827461095427497e-05,
+      "loss": 3.8697,
+      "step": 2209280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3819075147916977e-05,
+      "loss": 3.8761,
+      "step": 2209792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3810689200406457e-05,
+      "loss": 3.8755,
+      "step": 2210304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3802303252895937e-05,
+      "loss": 3.8706,
+      "step": 2210816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3793917305385418e-05,
+      "loss": 3.8709,
+      "step": 2211328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3785547736678627e-05,
+      "loss": 3.875,
+      "step": 2211840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3777161789168107e-05,
+      "loss": 3.8794,
+      "step": 2212352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3768775841657589e-05,
+      "loss": 3.8775,
+      "step": 2212864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9794180393218994,
+      "eval_runtime": 295.5561,
+      "eval_samples_per_second": 1291.095,
+      "eval_steps_per_second": 40.348,
+      "step": 2213268
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3760389894147069e-05,
+      "loss": 3.8665,
+      "step": 2213376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3752003946636549e-05,
+      "loss": 3.8609,
+      "step": 2213888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3743617999126027e-05,
+      "loss": 3.8691,
+      "step": 2214400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3735232051615507e-05,
+      "loss": 3.8784,
+      "step": 2214912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3726846104104987e-05,
+      "loss": 3.8751,
+      "step": 2215424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3718460156594467e-05,
+      "loss": 3.8728,
+      "step": 2215936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3710074209083947e-05,
+      "loss": 3.8745,
+      "step": 2216448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3701688261573429e-05,
+      "loss": 3.8733,
+      "step": 2216960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3693302314062905e-05,
+      "loss": 3.8644,
+      "step": 2217472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3684916366552385e-05,
+      "loss": 3.8629,
+      "step": 2217984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3676530419041867e-05,
+      "loss": 3.878,
+      "step": 2218496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.366816085033508e-05,
+      "loss": 3.8804,
+      "step": 2219008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3659774902824558e-05,
+      "loss": 3.881,
+      "step": 2219520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3651388955314038e-05,
+      "loss": 3.851,
+      "step": 2220032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3643003007803518e-05,
+      "loss": 3.8715,
+      "step": 2220544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3634617060292998e-05,
+      "loss": 3.8587,
+      "step": 2221056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3626231112782477e-05,
+      "loss": 3.8737,
+      "step": 2221568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3617845165271959e-05,
+      "loss": 3.8695,
+      "step": 2222080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3609459217761439e-05,
+      "loss": 3.87,
+      "step": 2222592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3601073270250916e-05,
+      "loss": 3.867,
+      "step": 2223104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3592687322740397e-05,
+      "loss": 3.8858,
+      "step": 2223616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3584301375229877e-05,
+      "loss": 3.8718,
+      "step": 2224128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3575915427719357e-05,
+      "loss": 3.8729,
+      "step": 2224640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3567529480208837e-05,
+      "loss": 3.8683,
+      "step": 2225152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3559143532698317e-05,
+      "loss": 3.8761,
+      "step": 2225664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3550757585187795e-05,
+      "loss": 3.8729,
+      "step": 2226176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3542371637677275e-05,
+      "loss": 3.8614,
+      "step": 2226688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.353400206897049e-05,
+      "loss": 3.87,
+      "step": 2227200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.352561612145997e-05,
+      "loss": 3.8678,
+      "step": 2227712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3517230173949446e-05,
+      "loss": 3.8648,
+      "step": 2228224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3508844226438928e-05,
+      "loss": 3.8678,
+      "step": 2228736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.350049103653587e-05,
+      "loss": 3.8722,
+      "step": 2229248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3492105089025351e-05,
+      "loss": 3.8759,
+      "step": 2229760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3483719141514831e-05,
+      "loss": 3.8763,
+      "step": 2230272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3475333194004311e-05,
+      "loss": 3.8743,
+      "step": 2230784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3466963625297522e-05,
+      "loss": 3.8709,
+      "step": 2231296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3458577677787002e-05,
+      "loss": 3.8759,
+      "step": 2231808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3450191730276482e-05,
+      "loss": 3.8617,
+      "step": 2232320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3441805782765962e-05,
+      "loss": 3.8644,
+      "step": 2232832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3433436214059172e-05,
+      "loss": 3.8706,
+      "step": 2233344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3425050266548652e-05,
+      "loss": 3.8604,
+      "step": 2233856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3416664319038132e-05,
+      "loss": 3.8686,
+      "step": 2234368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3408278371527614e-05,
+      "loss": 3.8659,
+      "step": 2234880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.339989242401709e-05,
+      "loss": 3.87,
+      "step": 2235392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.339150647650657e-05,
+      "loss": 3.8727,
+      "step": 2235904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3383120528996052e-05,
+      "loss": 3.8731,
+      "step": 2236416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3374734581485532e-05,
+      "loss": 3.8776,
+      "step": 2236928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3366348633975012e-05,
+      "loss": 3.8699,
+      "step": 2237440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3357962686464492e-05,
+      "loss": 3.8487,
+      "step": 2237952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3349576738953974e-05,
+      "loss": 3.872,
+      "step": 2238464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.334119079144345e-05,
+      "loss": 3.8677,
+      "step": 2238976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3332837601540393e-05,
+      "loss": 3.8716,
+      "step": 2239488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3324451654029873e-05,
+      "loss": 3.8677,
+      "step": 2240000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3316082085323086e-05,
+      "loss": 3.864,
+      "step": 2240512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3307696137812564e-05,
+      "loss": 3.8582,
+      "step": 2241024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3299310190302044e-05,
+      "loss": 3.8599,
+      "step": 2241536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3290924242791524e-05,
+      "loss": 3.8709,
+      "step": 2242048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3282538295281006e-05,
+      "loss": 3.8528,
+      "step": 2242560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3274152347770486e-05,
+      "loss": 3.878,
+      "step": 2243072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3265766400259966e-05,
+      "loss": 3.8724,
+      "step": 2243584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3257380452749446e-05,
+      "loss": 3.8498,
+      "step": 2244096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3248994505238924e-05,
+      "loss": 3.8786,
+      "step": 2244608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3240608557728404e-05,
+      "loss": 3.8592,
+      "step": 2245120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3232222610217884e-05,
+      "loss": 3.8577,
+      "step": 2245632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3223853041511098e-05,
+      "loss": 3.8683,
+      "step": 2246144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3215467094000574e-05,
+      "loss": 3.8688,
+      "step": 2246656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3207081146490054e-05,
+      "loss": 3.8588,
+      "step": 2247168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3198695198979536e-05,
+      "loss": 3.8723,
+      "step": 2247680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3190309251469016e-05,
+      "loss": 3.851,
+      "step": 2248192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3181939682762227e-05,
+      "loss": 3.8581,
+      "step": 2248704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.317357011405544e-05,
+      "loss": 3.8775,
+      "step": 2249216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.316518416654492e-05,
+      "loss": 3.8651,
+      "step": 2249728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3156798219034398e-05,
+      "loss": 3.8749,
+      "step": 2250240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3148412271523878e-05,
+      "loss": 3.869,
+      "step": 2250752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3140026324013357e-05,
+      "loss": 3.8847,
+      "step": 2251264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3131640376502837e-05,
+      "loss": 3.8584,
+      "step": 2251776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3123254428992317e-05,
+      "loss": 3.8754,
+      "step": 2252288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3114868481481799e-05,
+      "loss": 3.8663,
+      "step": 2252800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3106482533971279e-05,
+      "loss": 3.8642,
+      "step": 2253312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3098096586460756e-05,
+      "loss": 3.8684,
+      "step": 2253824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3089710638950237e-05,
+      "loss": 3.88,
+      "step": 2254336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.308134107024345e-05,
+      "loss": 3.8668,
+      "step": 2254848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3072955122732928e-05,
+      "loss": 3.8827,
+      "step": 2255360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3064569175222408e-05,
+      "loss": 3.8559,
+      "step": 2255872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3056183227711888e-05,
+      "loss": 3.8553,
+      "step": 2256384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3047797280201368e-05,
+      "loss": 3.8613,
+      "step": 2256896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3039411332690848e-05,
+      "loss": 3.8669,
+      "step": 2257408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.303102538518033e-05,
+      "loss": 3.8684,
+      "step": 2257920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3022655816473539e-05,
+      "loss": 3.8677,
+      "step": 2258432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.301426986896302e-05,
+      "loss": 3.8652,
+      "step": 2258944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.30058839214525e-05,
+      "loss": 3.8662,
+      "step": 2259456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.299751435274571e-05,
+      "loss": 3.8657,
+      "step": 2259968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2989128405235191e-05,
+      "loss": 3.862,
+      "step": 2260480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.298074245772467e-05,
+      "loss": 3.8756,
+      "step": 2260992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.297235651021415e-05,
+      "loss": 3.8786,
+      "step": 2261504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.296397056270363e-05,
+      "loss": 3.8736,
+      "step": 2262016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2955584615193109e-05,
+      "loss": 3.8564,
+      "step": 2262528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2947198667682589e-05,
+      "loss": 3.8669,
+      "step": 2263040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2938812720172069e-05,
+      "loss": 3.863,
+      "step": 2263552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2930426772661549e-05,
+      "loss": 3.8724,
+      "step": 2264064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.292204082515103e-05,
+      "loss": 3.8715,
+      "step": 2264576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.291365487764051e-05,
+      "loss": 3.8638,
+      "step": 2265088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2905285308933721e-05,
+      "loss": 3.8689,
+      "step": 2265600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2896899361423201e-05,
+      "loss": 3.8699,
+      "step": 2266112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2888513413912681e-05,
+      "loss": 3.8673,
+      "step": 2266624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2880127466402161e-05,
+      "loss": 3.8645,
+      "step": 2267136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.287174151889164e-05,
+      "loss": 3.8636,
+      "step": 2267648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2863371950184852e-05,
+      "loss": 3.8549,
+      "step": 2268160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2854986002674332e-05,
+      "loss": 3.8667,
+      "step": 2268672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2846600055163813e-05,
+      "loss": 3.8607,
+      "step": 2269184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.283821410765329e-05,
+      "loss": 3.8623,
+      "step": 2269696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.282982816014277e-05,
+      "loss": 3.8632,
+      "step": 2270208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2821458591435984e-05,
+      "loss": 3.8654,
+      "step": 2270720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2813072643925464e-05,
+      "loss": 3.859,
+      "step": 2271232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.280468669641494e-05,
+      "loss": 3.8764,
+      "step": 2271744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2796300748904422e-05,
+      "loss": 3.8788,
+      "step": 2272256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2787914801393902e-05,
+      "loss": 3.8726,
+      "step": 2272768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2779528853883382e-05,
+      "loss": 3.8717,
+      "step": 2273280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2771175663980326e-05,
+      "loss": 3.8679,
+      "step": 2273792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2762789716469806e-05,
+      "loss": 3.8636,
+      "step": 2274304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2754403768959285e-05,
+      "loss": 3.8692,
+      "step": 2274816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2746017821448764e-05,
+      "loss": 3.854,
+      "step": 2275328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2737631873938244e-05,
+      "loss": 3.8685,
+      "step": 2275840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2729245926427724e-05,
+      "loss": 3.8583,
+      "step": 2276352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2720859978917205e-05,
+      "loss": 3.8751,
+      "step": 2276864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2712474031406685e-05,
+      "loss": 3.8681,
+      "step": 2277376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2704088083896165e-05,
+      "loss": 3.8694,
+      "step": 2277888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2695702136385645e-05,
+      "loss": 3.8611,
+      "step": 2278400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2687316188875123e-05,
+      "loss": 3.8609,
+      "step": 2278912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2678930241364603e-05,
+      "loss": 3.8687,
+      "step": 2279424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2670544293854083e-05,
+      "loss": 3.8738,
+      "step": 2279936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2662174725147294e-05,
+      "loss": 3.876,
+      "step": 2280448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2653805156440507e-05,
+      "loss": 3.8638,
+      "step": 2280960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2645419208929987e-05,
+      "loss": 3.8789,
+      "step": 2281472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2637033261419468e-05,
+      "loss": 3.8673,
+      "step": 2281984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2628647313908945e-05,
+      "loss": 3.8729,
+      "step": 2282496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2620261366398425e-05,
+      "loss": 3.8745,
+      "step": 2283008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2611875418887906e-05,
+      "loss": 3.8665,
+      "step": 2283520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2603505850181119e-05,
+      "loss": 3.8632,
+      "step": 2284032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2595119902670595e-05,
+      "loss": 3.8812,
+      "step": 2284544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2586733955160077e-05,
+      "loss": 3.8605,
+      "step": 2285056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2578348007649557e-05,
+      "loss": 3.8629,
+      "step": 2285568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2569962060139037e-05,
+      "loss": 3.8748,
+      "step": 2286080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2561576112628517e-05,
+      "loss": 3.8716,
+      "step": 2286592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2553190165117999e-05,
+      "loss": 3.8649,
+      "step": 2287104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2544804217607475e-05,
+      "loss": 3.8685,
+      "step": 2287616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2536418270096955e-05,
+      "loss": 3.8673,
+      "step": 2288128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.252804870139017e-05,
+      "loss": 3.8764,
+      "step": 2288640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2519679132683378e-05,
+      "loss": 3.8691,
+      "step": 2289152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9781558513641357,
+      "eval_runtime": 288.53,
+      "eval_samples_per_second": 1322.535,
+      "eval_steps_per_second": 41.33,
+      "step": 2289588
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.251129318517286e-05,
+      "loss": 3.853,
+      "step": 2289664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.250290723766234e-05,
+      "loss": 3.8563,
+      "step": 2290176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2494521290151818e-05,
+      "loss": 3.8662,
+      "step": 2290688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2486135342641298e-05,
+      "loss": 3.8779,
+      "step": 2291200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.247774939513078e-05,
+      "loss": 3.8721,
+      "step": 2291712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2469363447620258e-05,
+      "loss": 3.8697,
+      "step": 2292224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2460977500109738e-05,
+      "loss": 3.8657,
+      "step": 2292736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2452591552599218e-05,
+      "loss": 3.8678,
+      "step": 2293248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.24442056050887e-05,
+      "loss": 3.8601,
+      "step": 2293760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2435819657578178e-05,
+      "loss": 3.8671,
+      "step": 2294272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2427433710067658e-05,
+      "loss": 3.8666,
+      "step": 2294784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2419047762557138e-05,
+      "loss": 3.881,
+      "step": 2295296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2410661815046618e-05,
+      "loss": 3.8767,
+      "step": 2295808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2402275867536098e-05,
+      "loss": 3.847,
+      "step": 2296320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.239390629882931e-05,
+      "loss": 3.869,
+      "step": 2296832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.238552035131879e-05,
+      "loss": 3.8541,
+      "step": 2297344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2377134403808268e-05,
+      "loss": 3.8693,
+      "step": 2297856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2368748456297748e-05,
+      "loss": 3.8615,
+      "step": 2298368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.236036250878723e-05,
+      "loss": 3.87,
+      "step": 2298880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2351976561276708e-05,
+      "loss": 3.8586,
+      "step": 2299392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2343590613766188e-05,
+      "loss": 3.8846,
+      "step": 2299904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2335204666255668e-05,
+      "loss": 3.8633,
+      "step": 2300416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2326818718745148e-05,
+      "loss": 3.871,
+      "step": 2300928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2318432771234628e-05,
+      "loss": 3.8642,
+      "step": 2301440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2310046823724108e-05,
+      "loss": 3.8777,
+      "step": 2301952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2301660876213588e-05,
+      "loss": 3.8649,
+      "step": 2302464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2293291307506799e-05,
+      "loss": 3.8572,
+      "step": 2302976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2284921738800011e-05,
+      "loss": 3.8629,
+      "step": 2303488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2276535791289491e-05,
+      "loss": 3.871,
+      "step": 2304000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.226814984377897e-05,
+      "loss": 3.858,
+      "step": 2304512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.225976389626845e-05,
+      "loss": 3.8636,
+      "step": 2305024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2251377948757931e-05,
+      "loss": 3.8643,
+      "step": 2305536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2242992001247411e-05,
+      "loss": 3.8721,
+      "step": 2306048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.223460605373689e-05,
+      "loss": 3.8684,
+      "step": 2306560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2226236485030102e-05,
+      "loss": 3.8715,
+      "step": 2307072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2217866916323313e-05,
+      "loss": 3.8705,
+      "step": 2307584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2209480968812793e-05,
+      "loss": 3.8736,
+      "step": 2308096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2201095021302273e-05,
+      "loss": 3.8529,
+      "step": 2308608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2192709073791753e-05,
+      "loss": 3.8621,
+      "step": 2309120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2184323126281233e-05,
+      "loss": 3.8642,
+      "step": 2309632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2175937178770712e-05,
+      "loss": 3.8531,
+      "step": 2310144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2167567610063923e-05,
+      "loss": 3.8688,
+      "step": 2310656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2159181662553403e-05,
+      "loss": 3.8673,
+      "step": 2311168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2150795715042885e-05,
+      "loss": 3.8614,
+      "step": 2311680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2142409767532363e-05,
+      "loss": 3.8708,
+      "step": 2312192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2134023820021843e-05,
+      "loss": 3.87,
+      "step": 2312704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2125637872511323e-05,
+      "loss": 3.8708,
+      "step": 2313216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2117251925000803e-05,
+      "loss": 3.8669,
+      "step": 2313728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2108865977490283e-05,
+      "loss": 3.8472,
+      "step": 2314240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2100496408783495e-05,
+      "loss": 3.8644,
+      "step": 2314752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2092126840076706e-05,
+      "loss": 3.8655,
+      "step": 2315264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2083740892566184e-05,
+      "loss": 3.8698,
+      "step": 2315776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2075354945055666e-05,
+      "loss": 3.8593,
+      "step": 2316288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2066968997545146e-05,
+      "loss": 3.8588,
+      "step": 2316800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2058583050034624e-05,
+      "loss": 3.8557,
+      "step": 2317312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2050197102524104e-05,
+      "loss": 3.8579,
+      "step": 2317824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2041811155013586e-05,
+      "loss": 3.8647,
+      "step": 2318336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2033425207503066e-05,
+      "loss": 3.8481,
+      "step": 2318848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2025055638796277e-05,
+      "loss": 3.873,
+      "step": 2319360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2016669691285757e-05,
+      "loss": 3.8721,
+      "step": 2319872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2008283743775237e-05,
+      "loss": 3.8447,
+      "step": 2320384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1999897796264715e-05,
+      "loss": 3.8758,
+      "step": 2320896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1991511848754197e-05,
+      "loss": 3.8509,
+      "step": 2321408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1983125901243676e-05,
+      "loss": 3.8572,
+      "step": 2321920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1974739953733156e-05,
+      "loss": 3.8572,
+      "step": 2322432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1966354006222635e-05,
+      "loss": 3.868,
+      "step": 2322944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1957968058712116e-05,
+      "loss": 3.8556,
+      "step": 2323456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1949582111201596e-05,
+      "loss": 3.8691,
+      "step": 2323968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1941196163691075e-05,
+      "loss": 3.8447,
+      "step": 2324480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1932810216180555e-05,
+      "loss": 3.857,
+      "step": 2324992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1924440647473767e-05,
+      "loss": 3.8713,
+      "step": 2325504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.191608745757071e-05,
+      "loss": 3.8626,
+      "step": 2326016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1907701510060189e-05,
+      "loss": 3.8698,
+      "step": 2326528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1899315562549669e-05,
+      "loss": 3.8665,
+      "step": 2327040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.189092961503915e-05,
+      "loss": 3.8797,
+      "step": 2327552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1882560046332361e-05,
+      "loss": 3.8559,
+      "step": 2328064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.187417409882184e-05,
+      "loss": 3.8757,
+      "step": 2328576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1865788151311321e-05,
+      "loss": 3.8582,
+      "step": 2329088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1857402203800801e-05,
+      "loss": 3.8599,
+      "step": 2329600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1849016256290279e-05,
+      "loss": 3.8629,
+      "step": 2330112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1840630308779759e-05,
+      "loss": 3.873,
+      "step": 2330624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1832260740072972e-05,
+      "loss": 3.8677,
+      "step": 2331136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1823874792562452e-05,
+      "loss": 3.8772,
+      "step": 2331648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1815488845051931e-05,
+      "loss": 3.8546,
+      "step": 2332160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1807102897541411e-05,
+      "loss": 3.8493,
+      "step": 2332672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1798716950030891e-05,
+      "loss": 3.8596,
+      "step": 2333184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.179033100252037e-05,
+      "loss": 3.8641,
+      "step": 2333696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1781945055009851e-05,
+      "loss": 3.8611,
+      "step": 2334208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1773559107499331e-05,
+      "loss": 3.862,
+      "step": 2334720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1765173159988811e-05,
+      "loss": 3.8648,
+      "step": 2335232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.175678721247829e-05,
+      "loss": 3.8609,
+      "step": 2335744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1748401264967771e-05,
+      "loss": 3.8601,
+      "step": 2336256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1740015317457251e-05,
+      "loss": 3.8575,
+      "step": 2336768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1731645748750462e-05,
+      "loss": 3.8724,
+      "step": 2337280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1723259801239942e-05,
+      "loss": 3.8769,
+      "step": 2337792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1714873853729422e-05,
+      "loss": 3.868,
+      "step": 2338304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.17064879062189e-05,
+      "loss": 3.8556,
+      "step": 2338816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1698118337512113e-05,
+      "loss": 3.863,
+      "step": 2339328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1689732390001592e-05,
+      "loss": 3.8602,
+      "step": 2339840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1681346442491072e-05,
+      "loss": 3.8697,
+      "step": 2340352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1672960494980552e-05,
+      "loss": 3.8714,
+      "step": 2340864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1664574547470032e-05,
+      "loss": 3.8574,
+      "step": 2341376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1656188599959512e-05,
+      "loss": 3.8652,
+      "step": 2341888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.164780265244899e-05,
+      "loss": 3.8638,
+      "step": 2342400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1639449462545934e-05,
+      "loss": 3.8651,
+      "step": 2342912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1631079893839146e-05,
+      "loss": 3.8595,
+      "step": 2343424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1622693946328626e-05,
+      "loss": 3.8606,
+      "step": 2343936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1614307998818106e-05,
+      "loss": 3.8519,
+      "step": 2344448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1605922051307586e-05,
+      "loss": 3.8631,
+      "step": 2344960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1597536103797066e-05,
+      "loss": 3.8535,
+      "step": 2345472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1589150156286546e-05,
+      "loss": 3.8584,
+      "step": 2345984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1580764208776024e-05,
+      "loss": 3.8581,
+      "step": 2346496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1572378261265506e-05,
+      "loss": 3.8615,
+      "step": 2347008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1563992313754986e-05,
+      "loss": 3.8577,
+      "step": 2347520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1555606366244464e-05,
+      "loss": 3.8681,
+      "step": 2348032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1547236797537677e-05,
+      "loss": 3.874,
+      "step": 2348544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1538850850027157e-05,
+      "loss": 3.8727,
+      "step": 2349056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1530464902516637e-05,
+      "loss": 3.8691,
+      "step": 2349568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1522078955006117e-05,
+      "loss": 3.8605,
+      "step": 2350080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1513693007495597e-05,
+      "loss": 3.8623,
+      "step": 2350592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1505307059985077e-05,
+      "loss": 3.8656,
+      "step": 2351104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1496921112474555e-05,
+      "loss": 3.8466,
+      "step": 2351616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1488551543767767e-05,
+      "loss": 3.8654,
+      "step": 2352128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1480165596257247e-05,
+      "loss": 3.8577,
+      "step": 2352640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1471779648746727e-05,
+      "loss": 3.8712,
+      "step": 2353152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1463393701236207e-05,
+      "loss": 3.8616,
+      "step": 2353664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1455007753725687e-05,
+      "loss": 3.868,
+      "step": 2354176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1446638185018898e-05,
+      "loss": 3.8567,
+      "step": 2354688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.143825223750838e-05,
+      "loss": 3.8588,
+      "step": 2355200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1429866289997858e-05,
+      "loss": 3.8684,
+      "step": 2355712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1421480342487338e-05,
+      "loss": 3.8664,
+      "step": 2356224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.141311077378055e-05,
+      "loss": 3.8737,
+      "step": 2356736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1404724826270028e-05,
+      "loss": 3.8602,
+      "step": 2357248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1396338878759508e-05,
+      "loss": 3.8725,
+      "step": 2357760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.138795293124899e-05,
+      "loss": 3.8673,
+      "step": 2358272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.137956698373847e-05,
+      "loss": 3.8653,
+      "step": 2358784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1371181036227948e-05,
+      "loss": 3.8714,
+      "step": 2359296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.136281146752116e-05,
+      "loss": 3.8635,
+      "step": 2359808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.135442552001064e-05,
+      "loss": 3.8653,
+      "step": 2360320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1346039572500119e-05,
+      "loss": 3.8744,
+      "step": 2360832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1337653624989599e-05,
+      "loss": 3.8569,
+      "step": 2361344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.132926767747908e-05,
+      "loss": 3.8599,
+      "step": 2361856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.132088172996856e-05,
+      "loss": 3.8683,
+      "step": 2362368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1312512161261771e-05,
+      "loss": 3.8667,
+      "step": 2362880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1304126213751251e-05,
+      "loss": 3.8623,
+      "step": 2363392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1295740266240731e-05,
+      "loss": 3.8646,
+      "step": 2363904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.128735431873021e-05,
+      "loss": 3.8607,
+      "step": 2364416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1278968371219691e-05,
+      "loss": 3.8795,
+      "step": 2364928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1270598802512902e-05,
+      "loss": 3.8627,
+      "step": 2365440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.977055788040161,
+      "eval_runtime": 293.5697,
+      "eval_samples_per_second": 1299.831,
+      "eval_steps_per_second": 40.621,
+      "step": 2365908
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1262212855002382e-05,
+      "loss": 3.8577,
+      "step": 2365952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1253826907491862e-05,
+      "loss": 3.8608,
+      "step": 2366464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1245440959981342e-05,
+      "loss": 3.8553,
+      "step": 2366976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1237055012470822e-05,
+      "loss": 3.876,
+      "step": 2367488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1228685443764034e-05,
+      "loss": 3.8655,
+      "step": 2368000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1220299496253513e-05,
+      "loss": 3.8689,
+      "step": 2368512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1211913548742993e-05,
+      "loss": 3.86,
+      "step": 2369024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1203527601232472e-05,
+      "loss": 3.8642,
+      "step": 2369536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1195141653721952e-05,
+      "loss": 3.8548,
+      "step": 2370048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1186755706211432e-05,
+      "loss": 3.8636,
+      "step": 2370560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1178402516308376e-05,
+      "loss": 3.8624,
+      "step": 2371072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1170032947601586e-05,
+      "loss": 3.8788,
+      "step": 2371584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1161647000091066e-05,
+      "loss": 3.8723,
+      "step": 2372096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1153261052580546e-05,
+      "loss": 3.8449,
+      "step": 2372608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1144875105070026e-05,
+      "loss": 3.8619,
+      "step": 2373120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1136489157559506e-05,
+      "loss": 3.8551,
+      "step": 2373632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1128103210048986e-05,
+      "loss": 3.8629,
+      "step": 2374144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1119717262538466e-05,
+      "loss": 3.8611,
+      "step": 2374656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1111331315027946e-05,
+      "loss": 3.8638,
+      "step": 2375168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1102945367517426e-05,
+      "loss": 3.8579,
+      "step": 2375680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1094559420006906e-05,
+      "loss": 3.8775,
+      "step": 2376192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1086173472496386e-05,
+      "loss": 3.8648,
+      "step": 2376704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1077787524985864e-05,
+      "loss": 3.8643,
+      "step": 2377216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1069417956279077e-05,
+      "loss": 3.8604,
+      "step": 2377728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1061032008768557e-05,
+      "loss": 3.8742,
+      "step": 2378240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1052646061258037e-05,
+      "loss": 3.8616,
+      "step": 2378752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1044260113747517e-05,
+      "loss": 3.854,
+      "step": 2379264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1035874166236997e-05,
+      "loss": 3.8621,
+      "step": 2379776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1027488218726477e-05,
+      "loss": 3.8599,
+      "step": 2380288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1019102271215957e-05,
+      "loss": 3.8554,
+      "step": 2380800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1010716323705436e-05,
+      "loss": 3.8614,
+      "step": 2381312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1002346754998647e-05,
+      "loss": 3.8591,
+      "step": 2381824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0993960807488129e-05,
+      "loss": 3.8694,
+      "step": 2382336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0985574859977607e-05,
+      "loss": 3.8683,
+      "step": 2382848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0977188912467087e-05,
+      "loss": 3.8676,
+      "step": 2383360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.09688193437603e-05,
+      "loss": 3.8652,
+      "step": 2383872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0960433396249778e-05,
+      "loss": 3.8687,
+      "step": 2384384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0952047448739258e-05,
+      "loss": 3.8525,
+      "step": 2384896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0943661501228738e-05,
+      "loss": 3.8608,
+      "step": 2385408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.093529193252195e-05,
+      "loss": 3.8603,
+      "step": 2385920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0926905985011429e-05,
+      "loss": 3.8519,
+      "step": 2386432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0918536416304641e-05,
+      "loss": 3.8628,
+      "step": 2386944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0910150468794121e-05,
+      "loss": 3.8647,
+      "step": 2387456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0901764521283601e-05,
+      "loss": 3.856,
+      "step": 2387968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0893378573773081e-05,
+      "loss": 3.8659,
+      "step": 2388480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.088499262626256e-05,
+      "loss": 3.8672,
+      "step": 2388992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.087660667875204e-05,
+      "loss": 3.8676,
+      "step": 2389504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0868220731241519e-05,
+      "loss": 3.8646,
+      "step": 2390016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0859834783731e-05,
+      "loss": 3.8451,
+      "step": 2390528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.085144883622048e-05,
+      "loss": 3.8602,
+      "step": 2391040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0843062888709959e-05,
+      "loss": 3.8673,
+      "step": 2391552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.083467694119944e-05,
+      "loss": 3.8635,
+      "step": 2392064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.082629099368892e-05,
+      "loss": 3.8566,
+      "step": 2392576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.08179050461784e-05,
+      "loss": 3.8549,
+      "step": 2393088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0809535477471611e-05,
+      "loss": 3.8503,
+      "step": 2393600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0801165908764822e-05,
+      "loss": 3.8519,
+      "step": 2394112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0792779961254302e-05,
+      "loss": 3.8611,
+      "step": 2394624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0784394013743784e-05,
+      "loss": 3.8489,
+      "step": 2395136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0776008066233262e-05,
+      "loss": 3.8647,
+      "step": 2395648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0767622118722742e-05,
+      "loss": 3.8654,
+      "step": 2396160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0759236171212222e-05,
+      "loss": 3.8488,
+      "step": 2396672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0750850223701702e-05,
+      "loss": 3.8684,
+      "step": 2397184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0742464276191182e-05,
+      "loss": 3.8478,
+      "step": 2397696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0734094707484393e-05,
+      "loss": 3.8534,
+      "step": 2398208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0725725138777605e-05,
+      "loss": 3.8527,
+      "step": 2398720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0717339191267083e-05,
+      "loss": 3.8656,
+      "step": 2399232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0708953243756565e-05,
+      "loss": 3.8512,
+      "step": 2399744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0700567296246045e-05,
+      "loss": 3.8645,
+      "step": 2400256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0692181348735523e-05,
+      "loss": 3.8405,
+      "step": 2400768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0683795401225003e-05,
+      "loss": 3.852,
+      "step": 2401280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0675409453714485e-05,
+      "loss": 3.8692,
+      "step": 2401792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0667039885007696e-05,
+      "loss": 3.8602,
+      "step": 2402304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0658653937497176e-05,
+      "loss": 3.863,
+      "step": 2402816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0650267989986655e-05,
+      "loss": 3.8659,
+      "step": 2403328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0641882042476135e-05,
+      "loss": 3.8748,
+      "step": 2403840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0633512473769346e-05,
+      "loss": 3.8551,
+      "step": 2404352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0625126526258826e-05,
+      "loss": 3.8723,
+      "step": 2404864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0616740578748306e-05,
+      "loss": 3.8514,
+      "step": 2405376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0608354631237786e-05,
+      "loss": 3.8568,
+      "step": 2405888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0599968683727266e-05,
+      "loss": 3.8583,
+      "step": 2406400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0591599115020477e-05,
+      "loss": 3.8659,
+      "step": 2406912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0583213167509957e-05,
+      "loss": 3.8662,
+      "step": 2407424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0574827219999438e-05,
+      "loss": 3.8733,
+      "step": 2407936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0566441272488917e-05,
+      "loss": 3.8538,
+      "step": 2408448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0558055324978397e-05,
+      "loss": 3.8445,
+      "step": 2408960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0549669377467877e-05,
+      "loss": 3.8533,
+      "step": 2409472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0541299808761087e-05,
+      "loss": 3.8608,
+      "step": 2409984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0532913861250567e-05,
+      "loss": 3.8575,
+      "step": 2410496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0524527913740049e-05,
+      "loss": 3.8551,
+      "step": 2411008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0516141966229527e-05,
+      "loss": 3.8652,
+      "step": 2411520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0507756018719007e-05,
+      "loss": 3.8539,
+      "step": 2412032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0499370071208487e-05,
+      "loss": 3.858,
+      "step": 2412544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0490984123697969e-05,
+      "loss": 3.8543,
+      "step": 2413056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0482598176187447e-05,
+      "loss": 3.8673,
+      "step": 2413568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0474228607480658e-05,
+      "loss": 3.8725,
+      "step": 2414080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.046585903877387e-05,
+      "loss": 3.8669,
+      "step": 2414592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.045747309126335e-05,
+      "loss": 3.8463,
+      "step": 2415104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.044908714375283e-05,
+      "loss": 3.861,
+      "step": 2415616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.044070119624231e-05,
+      "loss": 3.8557,
+      "step": 2416128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.043231524873179e-05,
+      "loss": 3.8641,
+      "step": 2416640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0423929301221268e-05,
+      "loss": 3.8676,
+      "step": 2417152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.041554335371075e-05,
+      "loss": 3.853,
+      "step": 2417664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.040715740620023e-05,
+      "loss": 3.8632,
+      "step": 2418176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.039878783749344e-05,
+      "loss": 3.8615,
+      "step": 2418688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.039040188998292e-05,
+      "loss": 3.8597,
+      "step": 2419200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.03820159424724e-05,
+      "loss": 3.8588,
+      "step": 2419712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.037362999496188e-05,
+      "loss": 3.8559,
+      "step": 2420224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0365260426255092e-05,
+      "loss": 3.8507,
+      "step": 2420736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0356874478744571e-05,
+      "loss": 3.8586,
+      "step": 2421248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0348488531234051e-05,
+      "loss": 3.8511,
+      "step": 2421760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0340118962527264e-05,
+      "loss": 3.8525,
+      "step": 2422272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0331733015016742e-05,
+      "loss": 3.8535,
+      "step": 2422784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0323347067506222e-05,
+      "loss": 3.8644,
+      "step": 2423296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0314977498799435e-05,
+      "loss": 3.8523,
+      "step": 2423808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0306591551288915e-05,
+      "loss": 3.8575,
+      "step": 2424320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0298205603778393e-05,
+      "loss": 3.8735,
+      "step": 2424832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0289819656267874e-05,
+      "loss": 3.8687,
+      "step": 2425344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0281433708757354e-05,
+      "loss": 3.8667,
+      "step": 2425856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0273047761246833e-05,
+      "loss": 3.8567,
+      "step": 2426368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0264661813736313e-05,
+      "loss": 3.8595,
+      "step": 2426880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0256275866225794e-05,
+      "loss": 3.8598,
+      "step": 2427392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0247906297519005e-05,
+      "loss": 3.8493,
+      "step": 2427904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0239520350008485e-05,
+      "loss": 3.8597,
+      "step": 2428416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0231134402497965e-05,
+      "loss": 3.8551,
+      "step": 2428928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0222748454987445e-05,
+      "loss": 3.8657,
+      "step": 2429440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0214362507476923e-05,
+      "loss": 3.8584,
+      "step": 2429952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0205976559966405e-05,
+      "loss": 3.8618,
+      "step": 2430464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0197590612455885e-05,
+      "loss": 3.8547,
+      "step": 2430976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0189204664945363e-05,
+      "loss": 3.8551,
+      "step": 2431488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0180835096238576e-05,
+      "loss": 3.863,
+      "step": 2432000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0172449148728056e-05,
+      "loss": 3.8604,
+      "step": 2432512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0164079580021266e-05,
+      "loss": 3.8671,
+      "step": 2433024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0155693632510746e-05,
+      "loss": 3.8582,
+      "step": 2433536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0147307685000226e-05,
+      "loss": 3.8689,
+      "step": 2434048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0138921737489706e-05,
+      "loss": 3.8643,
+      "step": 2434560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0130535789979186e-05,
+      "loss": 3.8644,
+      "step": 2435072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0122149842468666e-05,
+      "loss": 3.8627,
+      "step": 2435584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0113763894958146e-05,
+      "loss": 3.8611,
+      "step": 2436096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0105394326251359e-05,
+      "loss": 3.8618,
+      "step": 2436608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0097008378740837e-05,
+      "loss": 3.8681,
+      "step": 2437120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0088622431230317e-05,
+      "loss": 3.8568,
+      "step": 2437632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0080236483719797e-05,
+      "loss": 3.8546,
+      "step": 2438144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0071850536209278e-05,
+      "loss": 3.8644,
+      "step": 2438656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0063464588698757e-05,
+      "loss": 3.862,
+      "step": 2439168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0055078641188237e-05,
+      "loss": 3.8592,
+      "step": 2439680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0046692693677717e-05,
+      "loss": 3.8661,
+      "step": 2440192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0038323124970927e-05,
+      "loss": 3.8558,
+      "step": 2440704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0029937177460407e-05,
+      "loss": 3.8713,
+      "step": 2441216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0021551229949889e-05,
+      "loss": 3.8616,
+      "step": 2441728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.976073980331421,
+      "eval_runtime": 289.1696,
+      "eval_samples_per_second": 1319.61,
+      "eval_steps_per_second": 41.239,
+      "step": 2442228
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0013165282439369e-05,
+      "loss": 3.8406,
+      "step": 2442240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0004779334928847e-05,
+      "loss": 3.8535,
+      "step": 2442752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.996393387418327e-06,
+      "loss": 3.8508,
+      "step": 2443264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.988007439907809e-06,
+      "loss": 3.8701,
+      "step": 2443776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.979621492397287e-06,
+      "loss": 3.8644,
+      "step": 2444288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.971235544886767e-06,
+      "loss": 3.8646,
+      "step": 2444800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.962849597376247e-06,
+      "loss": 3.8573,
+      "step": 2445312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.954463649865727e-06,
+      "loss": 3.8593,
+      "step": 2445824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.946077702355207e-06,
+      "loss": 3.8492,
+      "step": 2446336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.937691754844687e-06,
+      "loss": 3.8615,
+      "step": 2446848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.929305807334167e-06,
+      "loss": 3.8576,
+      "step": 2447360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.920919859823647e-06,
+      "loss": 3.8747,
+      "step": 2447872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.912550291116857e-06,
+      "loss": 3.87,
+      "step": 2448384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.904164343606339e-06,
+      "loss": 3.8438,
+      "step": 2448896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.895778396095817e-06,
+      "loss": 3.8574,
+      "step": 2449408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.887392448585297e-06,
+      "loss": 3.8559,
+      "step": 2449920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.879006501074777e-06,
+      "loss": 3.855,
+      "step": 2450432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.870620553564257e-06,
+      "loss": 3.8601,
+      "step": 2450944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.862234606053737e-06,
+      "loss": 3.8612,
+      "step": 2451456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.853848658543217e-06,
+      "loss": 3.8532,
+      "step": 2451968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.84547908983643e-06,
+      "loss": 3.8754,
+      "step": 2452480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.837093142325908e-06,
+      "loss": 3.8627,
+      "step": 2452992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.828707194815388e-06,
+      "loss": 3.8575,
+      "step": 2453504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.820321247304868e-06,
+      "loss": 3.8549,
+      "step": 2454016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.811935299794348e-06,
+      "loss": 3.8761,
+      "step": 2454528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.803549352283828e-06,
+      "loss": 3.858,
+      "step": 2455040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.795163404773308e-06,
+      "loss": 3.8544,
+      "step": 2455552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.78677745726279e-06,
+      "loss": 3.8587,
+      "step": 2456064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.778407888555998e-06,
+      "loss": 3.8544,
+      "step": 2456576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.770021941045478e-06,
+      "loss": 3.8534,
+      "step": 2457088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.76163599353496e-06,
+      "loss": 3.8574,
+      "step": 2457600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.753250046024438e-06,
+      "loss": 3.8566,
+      "step": 2458112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.744864098513918e-06,
+      "loss": 3.864,
+      "step": 2458624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.736478151003398e-06,
+      "loss": 3.862,
+      "step": 2459136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.72809220349288e-06,
+      "loss": 3.868,
+      "step": 2459648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.719722634786089e-06,
+      "loss": 3.8625,
+      "step": 2460160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.71133668727557e-06,
+      "loss": 3.8666,
+      "step": 2460672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.70295073976505e-06,
+      "loss": 3.852,
+      "step": 2461184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.694564792254529e-06,
+      "loss": 3.8569,
+      "step": 2461696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.686195223547741e-06,
+      "loss": 3.8512,
+      "step": 2462208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.677809276037221e-06,
+      "loss": 3.8545,
+      "step": 2462720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.669423328526701e-06,
+      "loss": 3.8542,
+      "step": 2463232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.66103738101618e-06,
+      "loss": 3.8628,
+      "step": 2463744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.652651433505661e-06,
+      "loss": 3.8536,
+      "step": 2464256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.644265485995141e-06,
+      "loss": 3.8626,
+      "step": 2464768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.63587953848462e-06,
+      "loss": 3.8625,
+      "step": 2465280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.6274935909741e-06,
+      "loss": 3.8662,
+      "step": 2465792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.619140401071043e-06,
+      "loss": 3.8611,
+      "step": 2466304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.610754453560524e-06,
+      "loss": 3.8408,
+      "step": 2466816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.602368506050003e-06,
+      "loss": 3.854,
+      "step": 2467328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.593982558539482e-06,
+      "loss": 3.8661,
+      "step": 2467840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.585596611028962e-06,
+      "loss": 3.8615,
+      "step": 2468352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.577210663518444e-06,
+      "loss": 3.848,
+      "step": 2468864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.568824716007922e-06,
+      "loss": 3.8532,
+      "step": 2469376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.560455147301133e-06,
+      "loss": 3.849,
+      "step": 2469888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.552069199790615e-06,
+      "loss": 3.8464,
+      "step": 2470400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.543683252280093e-06,
+      "loss": 3.8651,
+      "step": 2470912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.535297304769573e-06,
+      "loss": 3.8433,
+      "step": 2471424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.526911357259053e-06,
+      "loss": 3.8584,
+      "step": 2471936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.518541788552265e-06,
+      "loss": 3.8619,
+      "step": 2472448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.510172219845476e-06,
+      "loss": 3.8446,
+      "step": 2472960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.501786272334956e-06,
+      "loss": 3.8646,
+      "step": 2473472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.493400324824436e-06,
+      "loss": 3.8447,
+      "step": 2473984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.485014377313916e-06,
+      "loss": 3.8537,
+      "step": 2474496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.476628429803396e-06,
+      "loss": 3.8452,
+      "step": 2475008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.468242482292876e-06,
+      "loss": 3.8643,
+      "step": 2475520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.459856534782356e-06,
+      "loss": 3.8513,
+      "step": 2476032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.451470587271836e-06,
+      "loss": 3.8553,
+      "step": 2476544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.443084639761316e-06,
+      "loss": 3.8418,
+      "step": 2477056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.434698692250796e-06,
+      "loss": 3.8477,
+      "step": 2477568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.426312744740274e-06,
+      "loss": 3.8639,
+      "step": 2478080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.417943176033487e-06,
+      "loss": 3.8648,
+      "step": 2478592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.409557228522967e-06,
+      "loss": 3.8539,
+      "step": 2479104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.401171281012447e-06,
+      "loss": 3.8594,
+      "step": 2479616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.392785333501926e-06,
+      "loss": 3.8714,
+      "step": 2480128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.384399385991406e-06,
+      "loss": 3.8505,
+      "step": 2480640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.376029817284617e-06,
+      "loss": 3.87,
+      "step": 2481152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.367643869774097e-06,
+      "loss": 3.8479,
+      "step": 2481664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.359257922263577e-06,
+      "loss": 3.8559,
+      "step": 2482176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.350871974753057e-06,
+      "loss": 3.855,
+      "step": 2482688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.34250240604627e-06,
+      "loss": 3.8628,
+      "step": 2483200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.334116458535748e-06,
+      "loss": 3.8616,
+      "step": 2483712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.325730511025228e-06,
+      "loss": 3.8712,
+      "step": 2484224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.31734456351471e-06,
+      "loss": 3.8553,
+      "step": 2484736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.30897499480792e-06,
+      "loss": 3.8382,
+      "step": 2485248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.300589047297398e-06,
+      "loss": 3.8482,
+      "step": 2485760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.29220309978688e-06,
+      "loss": 3.861,
+      "step": 2486272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.28381715227636e-06,
+      "loss": 3.8528,
+      "step": 2486784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.275431204765838e-06,
+      "loss": 3.8527,
+      "step": 2487296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.26706163605905e-06,
+      "loss": 3.8607,
+      "step": 2487808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.25867568854853e-06,
+      "loss": 3.8526,
+      "step": 2488320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.25028974103801e-06,
+      "loss": 3.8542,
+      "step": 2488832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.24190379352749e-06,
+      "loss": 3.8524,
+      "step": 2489344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.23351784601697e-06,
+      "loss": 3.8656,
+      "step": 2489856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.22513189850645e-06,
+      "loss": 3.8669,
+      "step": 2490368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.216745950995929e-06,
+      "loss": 3.8612,
+      "step": 2490880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.208376382289141e-06,
+      "loss": 3.8443,
+      "step": 2491392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.199990434778621e-06,
+      "loss": 3.8553,
+      "step": 2491904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.191604487268101e-06,
+      "loss": 3.8563,
+      "step": 2492416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.183218539757581e-06,
+      "loss": 3.8598,
+      "step": 2492928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.174832592247061e-06,
+      "loss": 3.8631,
+      "step": 2493440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.166479402344004e-06,
+      "loss": 3.8503,
+      "step": 2493952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.158093454833484e-06,
+      "loss": 3.8556,
+      "step": 2494464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.149707507322963e-06,
+      "loss": 3.8585,
+      "step": 2494976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.141321559812444e-06,
+      "loss": 3.8613,
+      "step": 2495488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.132951991105655e-06,
+      "loss": 3.8488,
+      "step": 2496000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.124566043595133e-06,
+      "loss": 3.8566,
+      "step": 2496512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.116180096084615e-06,
+      "loss": 3.8472,
+      "step": 2497024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.107810527377826e-06,
+      "loss": 3.8502,
+      "step": 2497536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.099424579867306e-06,
+      "loss": 3.8507,
+      "step": 2498048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.091038632356786e-06,
+      "loss": 3.8499,
+      "step": 2498560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.082652684846266e-06,
+      "loss": 3.8461,
+      "step": 2499072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.074266737335746e-06,
+      "loss": 3.8598,
+      "step": 2499584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.065880789825226e-06,
+      "loss": 3.8532,
+      "step": 2500096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.057494842314706e-06,
+      "loss": 3.8527,
+      "step": 2500608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.049108894804186e-06,
+      "loss": 3.868,
+      "step": 2501120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.040722947293665e-06,
+      "loss": 3.8689,
+      "step": 2501632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.032336999783145e-06,
+      "loss": 3.8617,
+      "step": 2502144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.023951052272625e-06,
+      "loss": 3.8535,
+      "step": 2502656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.015581483565836e-06,
+      "loss": 3.8556,
+      "step": 2503168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.007195536055316e-06,
+      "loss": 3.8548,
+      "step": 2503680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.998809588544796e-06,
+      "loss": 3.8489,
+      "step": 2504192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.990423641034276e-06,
+      "loss": 3.8534,
+      "step": 2504704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.982037693523756e-06,
+      "loss": 3.8519,
+      "step": 2505216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.973651746013236e-06,
+      "loss": 3.8605,
+      "step": 2505728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.965282177306447e-06,
+      "loss": 3.8539,
+      "step": 2506240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.956896229795927e-06,
+      "loss": 3.8618,
+      "step": 2506752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.948510282285407e-06,
+      "loss": 3.8571,
+      "step": 2507264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.940124334774887e-06,
+      "loss": 3.8513,
+      "step": 2507776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.931738387264367e-06,
+      "loss": 3.8492,
+      "step": 2508288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.923352439753847e-06,
+      "loss": 3.8668,
+      "step": 2508800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.914982871047057e-06,
+      "loss": 3.86,
+      "step": 2509312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.906596923536537e-06,
+      "loss": 3.8594,
+      "step": 2509824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.898210976026019e-06,
+      "loss": 3.862,
+      "step": 2510336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.889825028515497e-06,
+      "loss": 3.8619,
+      "step": 2510848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.881439081004977e-06,
+      "loss": 3.8634,
+      "step": 2511360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.873053133494457e-06,
+      "loss": 3.8609,
+      "step": 2511872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.864667185983939e-06,
+      "loss": 3.8545,
+      "step": 2512384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.856281238473417e-06,
+      "loss": 3.8604,
+      "step": 2512896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.847895290962897e-06,
+      "loss": 3.8649,
+      "step": 2513408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.83952572225611e-06,
+      "loss": 3.8548,
+      "step": 2513920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.83115615354932e-06,
+      "loss": 3.8526,
+      "step": 2514432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.8227702060388e-06,
+      "loss": 3.8588,
+      "step": 2514944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.81438425852828e-06,
+      "loss": 3.8647,
+      "step": 2515456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.80599831101776e-06,
+      "loss": 3.8503,
+      "step": 2515968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.797612363507238e-06,
+      "loss": 3.8628,
+      "step": 2516480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.789242794800451e-06,
+      "loss": 3.8544,
+      "step": 2516992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.78085684728993e-06,
+      "loss": 3.8656,
+      "step": 2517504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.77247089977941e-06,
+      "loss": 3.862,
+      "step": 2518016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.76408495226889e-06,
+      "loss": 3.8491,
+      "step": 2518528
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.974855422973633,
+      "eval_runtime": 296.3908,
+      "eval_samples_per_second": 1287.459,
+      "eval_steps_per_second": 40.234,
+      "step": 2518548
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.75569900475837e-06,
+      "loss": 3.8504,
+      "step": 2519040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.74731305724785e-06,
+      "loss": 3.8505,
+      "step": 2519552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.73892710973733e-06,
+      "loss": 3.867,
+      "step": 2520064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.730557541030541e-06,
+      "loss": 3.8601,
+      "step": 2520576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.722171593520021e-06,
+      "loss": 3.8613,
+      "step": 2521088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.713785646009501e-06,
+      "loss": 3.8584,
+      "step": 2521600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.705399698498981e-06,
+      "loss": 3.8522,
+      "step": 2522112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.697030129792192e-06,
+      "loss": 3.8477,
+      "step": 2522624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.688644182281674e-06,
+      "loss": 3.8618,
+      "step": 2523136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.680258234771152e-06,
+      "loss": 3.8489,
+      "step": 2523648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.671872287260632e-06,
+      "loss": 3.8776,
+      "step": 2524160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.663486339750112e-06,
+      "loss": 3.8642,
+      "step": 2524672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.655116771043324e-06,
+      "loss": 3.8465,
+      "step": 2525184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.646747202336535e-06,
+      "loss": 3.8503,
+      "step": 2525696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.638361254826015e-06,
+      "loss": 3.851,
+      "step": 2526208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.629975307315495e-06,
+      "loss": 3.8484,
+      "step": 2526720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.621589359804973e-06,
+      "loss": 3.8577,
+      "step": 2527232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.613203412294455e-06,
+      "loss": 3.8618,
+      "step": 2527744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.604817464783935e-06,
+      "loss": 3.8471,
+      "step": 2528256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.596431517273415e-06,
+      "loss": 3.8749,
+      "step": 2528768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.588045569762893e-06,
+      "loss": 3.8528,
+      "step": 2529280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.579676001056106e-06,
+      "loss": 3.8595,
+      "step": 2529792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.571290053545586e-06,
+      "loss": 3.8532,
+      "step": 2530304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.562904106035066e-06,
+      "loss": 3.8698,
+      "step": 2530816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.554518158524545e-06,
+      "loss": 3.8506,
+      "step": 2531328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.546132211014025e-06,
+      "loss": 3.8547,
+      "step": 2531840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.537746263503505e-06,
+      "loss": 3.8517,
+      "step": 2532352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.529360315992985e-06,
+      "loss": 3.8535,
+      "step": 2532864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.520974368482465e-06,
+      "loss": 3.8486,
+      "step": 2533376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.512604799775676e-06,
+      "loss": 3.8539,
+      "step": 2533888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.504235231068889e-06,
+      "loss": 3.8583,
+      "step": 2534400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.495849283558367e-06,
+      "loss": 3.8574,
+      "step": 2534912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.487463336047847e-06,
+      "loss": 3.86,
+      "step": 2535424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.479077388537328e-06,
+      "loss": 3.8642,
+      "step": 2535936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.470707819830538e-06,
+      "loss": 3.86,
+      "step": 2536448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.46232187232002e-06,
+      "loss": 3.8591,
+      "step": 2536960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.4539359248095e-06,
+      "loss": 3.8501,
+      "step": 2537472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.445549977298979e-06,
+      "loss": 3.8556,
+      "step": 2537984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.437164029788457e-06,
+      "loss": 3.8464,
+      "step": 2538496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.428778082277939e-06,
+      "loss": 3.8542,
+      "step": 2539008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.420392134767419e-06,
+      "loss": 3.8519,
+      "step": 2539520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.412006187256897e-06,
+      "loss": 3.8573,
+      "step": 2540032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.403620239746377e-06,
+      "loss": 3.8495,
+      "step": 2540544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.395234292235859e-06,
+      "loss": 3.8624,
+      "step": 2541056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.386848344725337e-06,
+      "loss": 3.856,
+      "step": 2541568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.378462397214817e-06,
+      "loss": 3.8622,
+      "step": 2542080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.37009282850803e-06,
+      "loss": 3.8581,
+      "step": 2542592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.36170688099751e-06,
+      "loss": 3.8404,
+      "step": 2543104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.353320933486988e-06,
+      "loss": 3.8454,
+      "step": 2543616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.34493498597647e-06,
+      "loss": 3.8656,
+      "step": 2544128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.33656541726968e-06,
+      "loss": 3.8556,
+      "step": 2544640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.328195848562893e-06,
+      "loss": 3.8476,
+      "step": 2545152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.319809901052371e-06,
+      "loss": 3.8451,
+      "step": 2545664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.311423953541851e-06,
+      "loss": 3.8535,
+      "step": 2546176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.303038006031331e-06,
+      "loss": 3.836,
+      "step": 2546688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.29465205852081e-06,
+      "loss": 3.8672,
+      "step": 2547200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.286282489814022e-06,
+      "loss": 3.8367,
+      "step": 2547712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.277896542303502e-06,
+      "loss": 3.8566,
+      "step": 2548224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.269510594792983e-06,
+      "loss": 3.8593,
+      "step": 2548736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.261141026086192e-06,
+      "loss": 3.8415,
+      "step": 2549248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.252755078575674e-06,
+      "loss": 3.8565,
+      "step": 2549760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.244369131065154e-06,
+      "loss": 3.8452,
+      "step": 2550272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.235983183554634e-06,
+      "loss": 3.8516,
+      "step": 2550784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.227597236044112e-06,
+      "loss": 3.8403,
+      "step": 2551296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.219211288533594e-06,
+      "loss": 3.8632,
+      "step": 2551808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.210825341023074e-06,
+      "loss": 3.8473,
+      "step": 2552320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.202439393512552e-06,
+      "loss": 3.8521,
+      "step": 2552832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.194053446002032e-06,
+      "loss": 3.8373,
+      "step": 2553344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.185667498491514e-06,
+      "loss": 3.8464,
+      "step": 2553856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.177281550980992e-06,
+      "loss": 3.8578,
+      "step": 2554368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.168895603470472e-06,
+      "loss": 3.86,
+      "step": 2554880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.160526034763684e-06,
+      "loss": 3.8552,
+      "step": 2555392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.152140087253164e-06,
+      "loss": 3.8567,
+      "step": 2555904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.143754139742643e-06,
+      "loss": 3.8641,
+      "step": 2556416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.135368192232124e-06,
+      "loss": 3.8519,
+      "step": 2556928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.126998623525335e-06,
+      "loss": 3.8623,
+      "step": 2557440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.118612676014813e-06,
+      "loss": 3.849,
+      "step": 2557952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.110226728504295e-06,
+      "loss": 3.854,
+      "step": 2558464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.101840780993775e-06,
+      "loss": 3.8461,
+      "step": 2558976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.093454833483255e-06,
+      "loss": 3.8607,
+      "step": 2559488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.085085264776466e-06,
+      "loss": 3.8578,
+      "step": 2560000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.076699317265946e-06,
+      "loss": 3.8669,
+      "step": 2560512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.068313369755425e-06,
+      "loss": 3.8587,
+      "step": 2561024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.059927422244905e-06,
+      "loss": 3.8348,
+      "step": 2561536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.051541474734385e-06,
+      "loss": 3.8425,
+      "step": 2562048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.043171906027596e-06,
+      "loss": 3.8549,
+      "step": 2562560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.034785958517078e-06,
+      "loss": 3.8525,
+      "step": 2563072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.026400011006556e-06,
+      "loss": 3.8493,
+      "step": 2563584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.018014063496036e-06,
+      "loss": 3.8605,
+      "step": 2564096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.009628115985516e-06,
+      "loss": 3.845,
+      "step": 2564608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.001242168474996e-06,
+      "loss": 3.8504,
+      "step": 2565120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.992856220964476e-06,
+      "loss": 3.8528,
+      "step": 2565632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.984470273453956e-06,
+      "loss": 3.861,
+      "step": 2566144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.976084325943436e-06,
+      "loss": 3.8606,
+      "step": 2566656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.967714757236647e-06,
+      "loss": 3.864,
+      "step": 2567168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.959328809726127e-06,
+      "loss": 3.8376,
+      "step": 2567680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.950942862215607e-06,
+      "loss": 3.8554,
+      "step": 2568192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.942556914705086e-06,
+      "loss": 3.8527,
+      "step": 2568704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.934187345998297e-06,
+      "loss": 3.8556,
+      "step": 2569216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.925801398487779e-06,
+      "loss": 3.867,
+      "step": 2569728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.917415450977259e-06,
+      "loss": 3.8433,
+      "step": 2570240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.909062261074202e-06,
+      "loss": 3.8515,
+      "step": 2570752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.90067631356368e-06,
+      "loss": 3.8566,
+      "step": 2571264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.89229036605316e-06,
+      "loss": 3.8577,
+      "step": 2571776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.88390441854264e-06,
+      "loss": 3.8442,
+      "step": 2572288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.87551847103212e-06,
+      "loss": 3.8561,
+      "step": 2572800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.8671325235216e-06,
+      "loss": 3.8473,
+      "step": 2573312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.85874657601108e-06,
+      "loss": 3.8486,
+      "step": 2573824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.85036062850056e-06,
+      "loss": 3.8474,
+      "step": 2574336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.84197468099004e-06,
+      "loss": 3.8491,
+      "step": 2574848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.83358873347952e-06,
+      "loss": 3.8464,
+      "step": 2575360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.825202785969e-06,
+      "loss": 3.8552,
+      "step": 2575872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.816833217262211e-06,
+      "loss": 3.8463,
+      "step": 2576384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.80844726975169e-06,
+      "loss": 3.8518,
+      "step": 2576896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.80006132224117e-06,
+      "loss": 3.8667,
+      "step": 2577408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.79167537473065e-06,
+      "loss": 3.8634,
+      "step": 2577920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.78328942722013e-06,
+      "loss": 3.8565,
+      "step": 2578432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.77490347970961e-06,
+      "loss": 3.8536,
+      "step": 2578944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.766550289806554e-06,
+      "loss": 3.8534,
+      "step": 2579456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.758180721099766e-06,
+      "loss": 3.8546,
+      "step": 2579968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.749794773589245e-06,
+      "loss": 3.8422,
+      "step": 2580480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.741408826078725e-06,
+      "loss": 3.8519,
+      "step": 2580992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.733022878568205e-06,
+      "loss": 3.851,
+      "step": 2581504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.724636931057685e-06,
+      "loss": 3.857,
+      "step": 2582016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.716250983547165e-06,
+      "loss": 3.8515,
+      "step": 2582528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.707865036036644e-06,
+      "loss": 3.856,
+      "step": 2583040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.699479088526124e-06,
+      "loss": 3.8527,
+      "step": 2583552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.691093141015604e-06,
+      "loss": 3.8507,
+      "step": 2584064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.682707193505084e-06,
+      "loss": 3.8443,
+      "step": 2584576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.674321245994564e-06,
+      "loss": 3.8665,
+      "step": 2585088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.665935298484044e-06,
+      "loss": 3.8581,
+      "step": 2585600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.657549350973524e-06,
+      "loss": 3.8579,
+      "step": 2586112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.649179782266735e-06,
+      "loss": 3.8545,
+      "step": 2586624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.640793834756215e-06,
+      "loss": 3.8571,
+      "step": 2587136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.632407887245695e-06,
+      "loss": 3.8589,
+      "step": 2587648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.624021939735175e-06,
+      "loss": 3.8631,
+      "step": 2588160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.615635992224656e-06,
+      "loss": 3.8539,
+      "step": 2588672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.607250044714134e-06,
+      "loss": 3.8557,
+      "step": 2589184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.598864097203615e-06,
+      "loss": 3.8558,
+      "step": 2589696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.590478149693095e-06,
+      "loss": 3.8519,
+      "step": 2590208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.5821085809863055e-06,
+      "loss": 3.8501,
+      "step": 2590720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.573739012279517e-06,
+      "loss": 3.8556,
+      "step": 2591232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.565353064768997e-06,
+      "loss": 3.859,
+      "step": 2591744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.556967117258477e-06,
+      "loss": 3.8472,
+      "step": 2592256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.548581169747956e-06,
+      "loss": 3.8626,
+      "step": 2592768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.540195222237437e-06,
+      "loss": 3.8514,
+      "step": 2593280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.531809274726917e-06,
+      "loss": 3.8588,
+      "step": 2593792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.523423327216396e-06,
+      "loss": 3.8575,
+      "step": 2594304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.515053758509608e-06,
+      "loss": 3.848,
+      "step": 2594816
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.97375750541687,
+      "eval_runtime": 297.3713,
+      "eval_samples_per_second": 1283.214,
+      "eval_steps_per_second": 40.101,
+      "step": 2594868
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.506667810999088e-06,
+      "loss": 3.8446,
+      "step": 2595328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.498281863488568e-06,
+      "loss": 3.8442,
+      "step": 2595840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.4898959159780475e-06,
+      "loss": 3.8651,
+      "step": 2596352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.4815099684675275e-06,
+      "loss": 3.8518,
+      "step": 2596864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.4731240209570074e-06,
+      "loss": 3.8624,
+      "step": 2597376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.4647380734464865e-06,
+      "loss": 3.852,
+      "step": 2597888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.456352125935967e-06,
+      "loss": 3.8519,
+      "step": 2598400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.447966178425447e-06,
+      "loss": 3.842,
+      "step": 2598912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.439580230914926e-06,
+      "loss": 3.8569,
+      "step": 2599424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.431194283404406e-06,
+      "loss": 3.851,
+      "step": 2599936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.422808335893887e-06,
+      "loss": 3.8705,
+      "step": 2600448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.414422388383367e-06,
+      "loss": 3.8595,
+      "step": 2600960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.406052819676577e-06,
+      "loss": 3.8421,
+      "step": 2601472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.397666872166058e-06,
+      "loss": 3.8465,
+      "step": 2601984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.389280924655538e-06,
+      "loss": 3.8571,
+      "step": 2602496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.380894977145017e-06,
+      "loss": 3.8353,
+      "step": 2603008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.372509029634497e-06,
+      "loss": 3.8578,
+      "step": 2603520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.364123082123978e-06,
+      "loss": 3.8557,
+      "step": 2604032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.355737134613458e-06,
+      "loss": 3.849,
+      "step": 2604544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.347351187102937e-06,
+      "loss": 3.8726,
+      "step": 2605056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.338981618396148e-06,
+      "loss": 3.8513,
+      "step": 2605568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.330595670885629e-06,
+      "loss": 3.8572,
+      "step": 2606080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.3222097233751075e-06,
+      "loss": 3.8498,
+      "step": 2606592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.313823775864588e-06,
+      "loss": 3.8671,
+      "step": 2607104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.305437828354068e-06,
+      "loss": 3.843,
+      "step": 2607616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.297051880843549e-06,
+      "loss": 3.8537,
+      "step": 2608128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.288682312136759e-06,
+      "loss": 3.8432,
+      "step": 2608640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.280296364626239e-06,
+      "loss": 3.8537,
+      "step": 2609152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.27191041711572e-06,
+      "loss": 3.8482,
+      "step": 2609664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.263524469605199e-06,
+      "loss": 3.8475,
+      "step": 2610176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.255138522094679e-06,
+      "loss": 3.8551,
+      "step": 2610688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.246752574584159e-06,
+      "loss": 3.8589,
+      "step": 2611200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.2383666270736395e-06,
+      "loss": 3.8533,
+      "step": 2611712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.2299970583668495e-06,
+      "loss": 3.8668,
+      "step": 2612224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.221627489660061e-06,
+      "loss": 3.8543,
+      "step": 2612736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.213241542149542e-06,
+      "loss": 3.8548,
+      "step": 2613248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.204855594639022e-06,
+      "loss": 3.8478,
+      "step": 2613760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.196469647128501e-06,
+      "loss": 3.8516,
+      "step": 2614272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.188083699617981e-06,
+      "loss": 3.8454,
+      "step": 2614784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.179697752107462e-06,
+      "loss": 3.851,
+      "step": 2615296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.171311804596941e-06,
+      "loss": 3.8462,
+      "step": 2615808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.162925857086421e-06,
+      "loss": 3.8585,
+      "step": 2616320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.154539909575901e-06,
+      "loss": 3.8441,
+      "step": 2616832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.146170340869112e-06,
+      "loss": 3.8572,
+      "step": 2617344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.1377843933585915e-06,
+      "loss": 3.8573,
+      "step": 2617856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.129398445848072e-06,
+      "loss": 3.8572,
+      "step": 2618368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.121012498337552e-06,
+      "loss": 3.8569,
+      "step": 2618880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.112626550827031e-06,
+      "loss": 3.8377,
+      "step": 2619392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.104256982120243e-06,
+      "loss": 3.8454,
+      "step": 2619904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.095871034609723e-06,
+      "loss": 3.8606,
+      "step": 2620416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.087485087099204e-06,
+      "loss": 3.852,
+      "step": 2620928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.079099139588682e-06,
+      "loss": 3.8477,
+      "step": 2621440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.070713192078163e-06,
+      "loss": 3.8406,
+      "step": 2621952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.062327244567643e-06,
+      "loss": 3.8502,
+      "step": 2622464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.053941297057122e-06,
+      "loss": 3.8311,
+      "step": 2622976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.045555349546603e-06,
+      "loss": 3.8669,
+      "step": 2623488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.037185780839814e-06,
+      "loss": 3.8342,
+      "step": 2624000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.028799833329293e-06,
+      "loss": 3.8576,
+      "step": 2624512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.020413885818773e-06,
+      "loss": 3.8513,
+      "step": 2625024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.012044317111985e-06,
+      "loss": 3.8403,
+      "step": 2625536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.003658369601465e-06,
+      "loss": 3.8565,
+      "step": 2626048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.995272422090944e-06,
+      "loss": 3.8448,
+      "step": 2626560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.986886474580424e-06,
+      "loss": 3.8457,
+      "step": 2627072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.978500527069905e-06,
+      "loss": 3.8375,
+      "step": 2627584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.970114579559384e-06,
+      "loss": 3.8624,
+      "step": 2628096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.961745010852596e-06,
+      "loss": 3.8437,
+      "step": 2628608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.953375442145807e-06,
+      "loss": 3.8489,
+      "step": 2629120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.944989494635287e-06,
+      "loss": 3.8344,
+      "step": 2629632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.936603547124766e-06,
+      "loss": 3.8449,
+      "step": 2630144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.928233978417978e-06,
+      "loss": 3.8533,
+      "step": 2630656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.919848030907458e-06,
+      "loss": 3.8599,
+      "step": 2631168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.911462083396939e-06,
+      "loss": 3.8498,
+      "step": 2631680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.903076135886417e-06,
+      "loss": 3.8586,
+      "step": 2632192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.8947065671796295e-06,
+      "loss": 3.8582,
+      "step": 2632704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.886320619669109e-06,
+      "loss": 3.85,
+      "step": 2633216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.877934672158589e-06,
+      "loss": 3.8563,
+      "step": 2633728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.8695487246480685e-06,
+      "loss": 3.8496,
+      "step": 2634240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.861162777137549e-06,
+      "loss": 3.8489,
+      "step": 2634752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.852776829627029e-06,
+      "loss": 3.8458,
+      "step": 2635264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.844390882116508e-06,
+      "loss": 3.8545,
+      "step": 2635776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.836004934605988e-06,
+      "loss": 3.8572,
+      "step": 2636288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.827618987095469e-06,
+      "loss": 3.866,
+      "step": 2636800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.819233039584947e-06,
+      "loss": 3.8513,
+      "step": 2637312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.810847092074428e-06,
+      "loss": 3.8328,
+      "step": 2637824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.802461144563908e-06,
+      "loss": 3.8405,
+      "step": 2638336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.794075197053389e-06,
+      "loss": 3.8539,
+      "step": 2638848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.785705628346599e-06,
+      "loss": 3.8493,
+      "step": 2639360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.777319680836079e-06,
+      "loss": 3.8466,
+      "step": 2639872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.76893373332556e-06,
+      "loss": 3.8555,
+      "step": 2640384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.760547785815039e-06,
+      "loss": 3.8467,
+      "step": 2640896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.752161838304519e-06,
+      "loss": 3.8435,
+      "step": 2641408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.7437758907939995e-06,
+      "loss": 3.8488,
+      "step": 2641920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.7353899432834794e-06,
+      "loss": 3.8599,
+      "step": 2642432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.7270039957729586e-06,
+      "loss": 3.8536,
+      "step": 2642944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.71863442706617e-06,
+      "loss": 3.8597,
+      "step": 2643456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.710264858359382e-06,
+      "loss": 3.8401,
+      "step": 2643968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.701878910848862e-06,
+      "loss": 3.8509,
+      "step": 2644480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.693492963338341e-06,
+      "loss": 3.8515,
+      "step": 2644992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.685107015827821e-06,
+      "loss": 3.8509,
+      "step": 2645504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.676721068317302e-06,
+      "loss": 3.8603,
+      "step": 2646016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.668351499610512e-06,
+      "loss": 3.8438,
+      "step": 2646528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.659965552099992e-06,
+      "loss": 3.8478,
+      "step": 2647040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.651579604589472e-06,
+      "loss": 3.853,
+      "step": 2647552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.643193657078952e-06,
+      "loss": 3.8565,
+      "step": 2648064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.634807709568431e-06,
+      "loss": 3.8408,
+      "step": 2648576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.626421762057912e-06,
+      "loss": 3.8485,
+      "step": 2649088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.618035814547392e-06,
+      "loss": 3.8477,
+      "step": 2649600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.609649867036871e-06,
+      "loss": 3.846,
+      "step": 2650112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.601263919526351e-06,
+      "loss": 3.8404,
+      "step": 2650624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.592877972015832e-06,
+      "loss": 3.8458,
+      "step": 2651136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.58449202450531e-06,
+      "loss": 3.8434,
+      "step": 2651648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.576122455798523e-06,
+      "loss": 3.8526,
+      "step": 2652160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.567736508288003e-06,
+      "loss": 3.8415,
+      "step": 2652672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.559350560777483e-06,
+      "loss": 3.8513,
+      "step": 2653184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.550964613266962e-06,
+      "loss": 3.8648,
+      "step": 2653696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.542578665756443e-06,
+      "loss": 3.8615,
+      "step": 2654208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.5341927182459226e-06,
+      "loss": 3.8536,
+      "step": 2654720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.525823149539134e-06,
+      "loss": 3.8533,
+      "step": 2655232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.517437202028613e-06,
+      "loss": 3.8498,
+      "step": 2655744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.509051254518093e-06,
+      "loss": 3.8472,
+      "step": 2656256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.500665307007574e-06,
+      "loss": 3.8429,
+      "step": 2656768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.492279359497052e-06,
+      "loss": 3.8445,
+      "step": 2657280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.483893411986533e-06,
+      "loss": 3.8477,
+      "step": 2657792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.475523843279745e-06,
+      "loss": 3.8603,
+      "step": 2658304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.467137895769225e-06,
+      "loss": 3.8494,
+      "step": 2658816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.458751948258704e-06,
+      "loss": 3.8514,
+      "step": 2659328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.450366000748184e-06,
+      "loss": 3.8495,
+      "step": 2659840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.441980053237665e-06,
+      "loss": 3.8458,
+      "step": 2660352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.433594105727144e-06,
+      "loss": 3.8403,
+      "step": 2660864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.425208158216624e-06,
+      "loss": 3.8613,
+      "step": 2661376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.4168222107061045e-06,
+      "loss": 3.8585,
+      "step": 2661888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.408452641999314e-06,
+      "loss": 3.8587,
+      "step": 2662400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.400083073292526e-06,
+      "loss": 3.8516,
+      "step": 2662912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.391713504585738e-06,
+      "loss": 3.8507,
+      "step": 2663424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.383327557075218e-06,
+      "loss": 3.8566,
+      "step": 2663936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.374941609564697e-06,
+      "loss": 3.8669,
+      "step": 2664448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.3665556620541776e-06,
+      "loss": 3.8495,
+      "step": 2664960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.3581697145436575e-06,
+      "loss": 3.8508,
+      "step": 2665472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.3497837670331375e-06,
+      "loss": 3.851,
+      "step": 2665984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.3413978195226166e-06,
+      "loss": 3.8512,
+      "step": 2666496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.333011872012097e-06,
+      "loss": 3.8491,
+      "step": 2667008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.324625924501577e-06,
+      "loss": 3.8524,
+      "step": 2667520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.3162399769910564e-06,
+      "loss": 3.8582,
+      "step": 2668032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.307854029480536e-06,
+      "loss": 3.8438,
+      "step": 2668544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.299484460773748e-06,
+      "loss": 3.8596,
+      "step": 2669056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.291098513263229e-06,
+      "loss": 3.8562,
+      "step": 2669568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.282712565752707e-06,
+      "loss": 3.8522,
+      "step": 2670080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.274326618242188e-06,
+      "loss": 3.856,
+      "step": 2670592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.265940670731668e-06,
+      "loss": 3.8413,
+      "step": 2671104
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.973017692565918,
+      "eval_runtime": 291.5857,
+      "eval_samples_per_second": 1308.675,
+      "eval_steps_per_second": 40.897,
+      "step": 2671188
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.257554723221147e-06,
+      "loss": 3.8497,
+      "step": 2671616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.249168775710628e-06,
+      "loss": 3.8433,
+      "step": 2672128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.240782828200108e-06,
+      "loss": 3.8647,
+      "step": 2672640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.2324132594933185e-06,
+      "loss": 3.8455,
+      "step": 2673152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.224027311982799e-06,
+      "loss": 3.8633,
+      "step": 2673664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.215641364472278e-06,
+      "loss": 3.8491,
+      "step": 2674176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.207255416961758e-06,
+      "loss": 3.8519,
+      "step": 2674688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.19888584825497e-06,
+      "loss": 3.84,
+      "step": 2675200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.190516279548182e-06,
+      "loss": 3.8553,
+      "step": 2675712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.182130332037661e-06,
+      "loss": 3.8441,
+      "step": 2676224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.173744384527141e-06,
+      "loss": 3.8685,
+      "step": 2676736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.165358437016621e-06,
+      "loss": 3.8562,
+      "step": 2677248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.156988868309832e-06,
+      "loss": 3.8392,
+      "step": 2677760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.148602920799312e-06,
+      "loss": 3.8454,
+      "step": 2678272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.140233352092523e-06,
+      "loss": 3.8563,
+      "step": 2678784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.131847404582004e-06,
+      "loss": 3.8385,
+      "step": 2679296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.123461457071483e-06,
+      "loss": 3.8503,
+      "step": 2679808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.115075509560963e-06,
+      "loss": 3.8501,
+      "step": 2680320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.106689562050443e-06,
+      "loss": 3.8457,
+      "step": 2680832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.098303614539923e-06,
+      "loss": 3.8723,
+      "step": 2681344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.089917667029403e-06,
+      "loss": 3.8516,
+      "step": 2681856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.081531719518883e-06,
+      "loss": 3.85,
+      "step": 2682368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.073162150812094e-06,
+      "loss": 3.847,
+      "step": 2682880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.064776203301574e-06,
+      "loss": 3.8652,
+      "step": 2683392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.0563902557910534e-06,
+      "loss": 3.8419,
+      "step": 2683904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.048004308280534e-06,
+      "loss": 3.8472,
+      "step": 2684416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.039618360770013e-06,
+      "loss": 3.8432,
+      "step": 2684928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.031232413259494e-06,
+      "loss": 3.8534,
+      "step": 2685440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.022846465748973e-06,
+      "loss": 3.8449,
+      "step": 2685952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.014460518238453e-06,
+      "loss": 3.8443,
+      "step": 2686464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.006090949531665e-06,
+      "loss": 3.8537,
+      "step": 2686976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.997705002021144e-06,
+      "loss": 3.8519,
+      "step": 2687488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.989335433314356e-06,
+      "loss": 3.8512,
+      "step": 2688000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.980965864607567e-06,
+      "loss": 3.8608,
+      "step": 2688512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.972596295900779e-06,
+      "loss": 3.851,
+      "step": 2689024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.964210348390259e-06,
+      "loss": 3.8544,
+      "step": 2689536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.955824400879739e-06,
+      "loss": 3.8399,
+      "step": 2690048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.947438453369218e-06,
+      "loss": 3.8548,
+      "step": 2690560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.939052505858699e-06,
+      "loss": 3.8403,
+      "step": 2691072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.930666558348178e-06,
+      "loss": 3.8456,
+      "step": 2691584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.922280610837659e-06,
+      "loss": 3.846,
+      "step": 2692096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.913894663327138e-06,
+      "loss": 3.8557,
+      "step": 2692608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.905508715816618e-06,
+      "loss": 3.8416,
+      "step": 2693120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.897139147109829e-06,
+      "loss": 3.8512,
+      "step": 2693632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.888753199599309e-06,
+      "loss": 3.8581,
+      "step": 2694144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.880367252088789e-06,
+      "loss": 3.8547,
+      "step": 2694656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.871981304578269e-06,
+      "loss": 3.8586,
+      "step": 2695168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.863595357067749e-06,
+      "loss": 3.8335,
+      "step": 2695680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.855209409557229e-06,
+      "loss": 3.8401,
+      "step": 2696192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.846823462046708e-06,
+      "loss": 3.8551,
+      "step": 2696704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.838437514536189e-06,
+      "loss": 3.8494,
+      "step": 2697216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.830051567025668e-06,
+      "loss": 3.8462,
+      "step": 2697728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.821665619515149e-06,
+      "loss": 3.8393,
+      "step": 2698240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.813279672004628e-06,
+      "loss": 3.8508,
+      "step": 2698752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.80491010329784e-06,
+      "loss": 3.828,
+      "step": 2699264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.79652415578732e-06,
+      "loss": 3.8592,
+      "step": 2699776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.7881382082768e-06,
+      "loss": 3.8376,
+      "step": 2700288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.7797522607662795e-06,
+      "loss": 3.8474,
+      "step": 2700800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.77138269205949e-06,
+      "loss": 3.8551,
+      "step": 2701312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.763013123352702e-06,
+      "loss": 3.8373,
+      "step": 2701824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.754627175842182e-06,
+      "loss": 3.8523,
+      "step": 2702336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.746241228331662e-06,
+      "loss": 3.844,
+      "step": 2702848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.737855280821142e-06,
+      "loss": 3.8412,
+      "step": 2703360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.729469333310622e-06,
+      "loss": 3.8352,
+      "step": 2703872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.721083385800102e-06,
+      "loss": 3.8564,
+      "step": 2704384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.712713817093313e-06,
+      "loss": 3.8442,
+      "step": 2704896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.7043278695827925e-06,
+      "loss": 3.8482,
+      "step": 2705408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.6959419220722724e-06,
+      "loss": 3.8335,
+      "step": 2705920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.687555974561752e-06,
+      "loss": 3.8377,
+      "step": 2706432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.679202784658696e-06,
+      "loss": 3.8493,
+      "step": 2706944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.670816837148176e-06,
+      "loss": 3.8567,
+      "step": 2707456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.662430889637655e-06,
+      "loss": 3.8442,
+      "step": 2707968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.654044942127136e-06,
+      "loss": 3.8604,
+      "step": 2708480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.645658994616615e-06,
+      "loss": 3.8554,
+      "step": 2708992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.6372730471060955e-06,
+      "loss": 3.8474,
+      "step": 2709504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.628887099595575e-06,
+      "loss": 3.851,
+      "step": 2710016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.6205011520850546e-06,
+      "loss": 3.8481,
+      "step": 2710528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.6121152045745345e-06,
+      "loss": 3.8472,
+      "step": 2711040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.6037292570640145e-06,
+      "loss": 3.8436,
+      "step": 2711552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5953433095534944e-06,
+      "loss": 3.8523,
+      "step": 2712064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.586973740846706e-06,
+      "loss": 3.8582,
+      "step": 2712576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.578587793336186e-06,
+      "loss": 3.8613,
+      "step": 2713088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.570201845825666e-06,
+      "loss": 3.8502,
+      "step": 2713600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.561815898315145e-06,
+      "loss": 3.8295,
+      "step": 2714112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.553429950804626e-06,
+      "loss": 3.84,
+      "step": 2714624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.545044003294105e-06,
+      "loss": 3.8472,
+      "step": 2715136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.536658055783586e-06,
+      "loss": 3.8527,
+      "step": 2715648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.528272108273065e-06,
+      "loss": 3.8406,
+      "step": 2716160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.519902539566276e-06,
+      "loss": 3.8528,
+      "step": 2716672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.5115165920557565e-06,
+      "loss": 3.8446,
+      "step": 2717184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.503130644545236e-06,
+      "loss": 3.8398,
+      "step": 2717696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.494744697034716e-06,
+      "loss": 3.8498,
+      "step": 2718208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.486358749524196e-06,
+      "loss": 3.8531,
+      "step": 2718720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.4779728020136755e-06,
+      "loss": 3.8502,
+      "step": 2719232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.469586854503156e-06,
+      "loss": 3.8584,
+      "step": 2719744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.461200906992635e-06,
+      "loss": 3.8401,
+      "step": 2720256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.452814959482116e-06,
+      "loss": 3.8458,
+      "step": 2720768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.444445390775327e-06,
+      "loss": 3.8484,
+      "step": 2721280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.436059443264807e-06,
+      "loss": 3.8511,
+      "step": 2721792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.427673495754287e-06,
+      "loss": 3.8579,
+      "step": 2722304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.4193039270474985e-06,
+      "loss": 3.8424,
+      "step": 2722816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.410917979536978e-06,
+      "loss": 3.8459,
+      "step": 2723328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.402532032026458e-06,
+      "loss": 3.8484,
+      "step": 2723840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.3941460845159375e-06,
+      "loss": 3.8526,
+      "step": 2724352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.3857601370054175e-06,
+      "loss": 3.837,
+      "step": 2724864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.377390568298629e-06,
+      "loss": 3.8491,
+      "step": 2725376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.369004620788109e-06,
+      "loss": 3.8481,
+      "step": 2725888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.360618673277589e-06,
+      "loss": 3.8418,
+      "step": 2726400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.352232725767069e-06,
+      "loss": 3.8389,
+      "step": 2726912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.343846778256548e-06,
+      "loss": 3.8411,
+      "step": 2727424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.335460830746029e-06,
+      "loss": 3.8467,
+      "step": 2727936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.327074883235508e-06,
+      "loss": 3.8454,
+      "step": 2728448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.318688935724989e-06,
+      "loss": 3.8385,
+      "step": 2728960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.3103193670182e-06,
+      "loss": 3.8446,
+      "step": 2729472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.30193341950768e-06,
+      "loss": 3.8622,
+      "step": 2729984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.2935474719971595e-06,
+      "loss": 3.8599,
+      "step": 2730496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.2851615244866395e-06,
+      "loss": 3.8525,
+      "step": 2731008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.276791955779851e-06,
+      "loss": 3.8461,
+      "step": 2731520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.26840600826933e-06,
+      "loss": 3.8471,
+      "step": 2732032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.260020060758811e-06,
+      "loss": 3.8469,
+      "step": 2732544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.25163411324829e-06,
+      "loss": 3.8388,
+      "step": 2733056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.243248165737771e-06,
+      "loss": 3.8424,
+      "step": 2733568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.23486221822725e-06,
+      "loss": 3.8448,
+      "step": 2734080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.22647627071673e-06,
+      "loss": 3.8557,
+      "step": 2734592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.21809032320621e-06,
+      "loss": 3.8471,
+      "step": 2735104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.209720754499421e-06,
+      "loss": 3.851,
+      "step": 2735616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.2013348069889016e-06,
+      "loss": 3.8438,
+      "step": 2736128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.192948859478381e-06,
+      "loss": 3.8442,
+      "step": 2736640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.1845629119678615e-06,
+      "loss": 3.8411,
+      "step": 2737152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.176193343261072e-06,
+      "loss": 3.858,
+      "step": 2737664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.167823774554284e-06,
+      "loss": 3.8589,
+      "step": 2738176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.159437827043764e-06,
+      "loss": 3.8497,
+      "step": 2738688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.151051879533244e-06,
+      "loss": 3.8483,
+      "step": 2739200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.142665932022724e-06,
+      "loss": 3.8489,
+      "step": 2739712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.134296363315935e-06,
+      "loss": 3.85,
+      "step": 2740224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.1259104158054145e-06,
+      "loss": 3.8638,
+      "step": 2740736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.1175244682948945e-06,
+      "loss": 3.8467,
+      "step": 2741248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.1091385207843744e-06,
+      "loss": 3.85,
+      "step": 2741760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.100768952077586e-06,
+      "loss": 3.846,
+      "step": 2742272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.092383004567066e-06,
+      "loss": 3.8521,
+      "step": 2742784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.084013435860277e-06,
+      "loss": 3.8444,
+      "step": 2743296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.075627488349758e-06,
+      "loss": 3.85,
+      "step": 2743808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.067241540839237e-06,
+      "loss": 3.8562,
+      "step": 2744320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.0588555933287175e-06,
+      "loss": 3.8454,
+      "step": 2744832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.050469645818197e-06,
+      "loss": 3.8531,
+      "step": 2745344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.042083698307677e-06,
+      "loss": 3.8569,
+      "step": 2745856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.0336977507971565e-06,
+      "loss": 3.8483,
+      "step": 2746368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.0253118032866365e-06,
+      "loss": 3.8533,
+      "step": 2746880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.0169258557761165e-06,
+      "loss": 3.842,
+      "step": 2747392
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.972050905227661,
+      "eval_runtime": 291.8462,
+      "eval_samples_per_second": 1307.507,
+      "eval_steps_per_second": 40.861,
+      "step": 2747508
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.008539908265596e-06,
+      "loss": 3.838,
+      "step": 2747904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.000153960755076e-06,
+      "loss": 3.8388,
+      "step": 2748416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991768013244556e-06,
+      "loss": 3.8599,
+      "step": 2748928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983382065734036e-06,
+      "loss": 3.8472,
+      "step": 2749440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.974996118223516e-06,
+      "loss": 3.8623,
+      "step": 2749952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.966610170712996e-06,
+      "loss": 3.8418,
+      "step": 2750464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.958224223202475e-06,
+      "loss": 3.8522,
+      "step": 2750976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.949838275691956e-06,
+      "loss": 3.8356,
+      "step": 2751488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.941452328181435e-06,
+      "loss": 3.8539,
+      "step": 2752000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.933066380670915e-06,
+      "loss": 3.8396,
+      "step": 2752512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.924680433160395e-06,
+      "loss": 3.8633,
+      "step": 2753024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.916294485649875e-06,
+      "loss": 3.8585,
+      "step": 2753536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.907924916943087e-06,
+      "loss": 3.8332,
+      "step": 2754048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.899538969432566e-06,
+      "loss": 3.8442,
+      "step": 2754560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.891153021922047e-06,
+      "loss": 3.8518,
+      "step": 2755072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.882767074411526e-06,
+      "loss": 3.8368,
+      "step": 2755584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.874381126901006e-06,
+      "loss": 3.8469,
+      "step": 2756096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8659951793904865e-06,
+      "loss": 3.8486,
+      "step": 2756608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.857609231879966e-06,
+      "loss": 3.8463,
+      "step": 2757120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.849223284369446e-06,
+      "loss": 3.8663,
+      "step": 2757632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.840853715662657e-06,
+      "loss": 3.8481,
+      "step": 2758144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.832467768152137e-06,
+      "loss": 3.8482,
+      "step": 2758656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.824081820641617e-06,
+      "loss": 3.8491,
+      "step": 2759168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.815695873131096e-06,
+      "loss": 3.859,
+      "step": 2759680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.807309925620577e-06,
+      "loss": 3.8411,
+      "step": 2760192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.798923978110056e-06,
+      "loss": 3.8515,
+      "step": 2760704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.790538030599537e-06,
+      "loss": 3.8395,
+      "step": 2761216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.782152083089016e-06,
+      "loss": 3.8464,
+      "step": 2761728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.773782514382228e-06,
+      "loss": 3.8415,
+      "step": 2762240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.765396566871708e-06,
+      "loss": 3.8456,
+      "step": 2762752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.757026998164919e-06,
+      "loss": 3.8453,
+      "step": 2763264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.748641050654399e-06,
+      "loss": 3.8553,
+      "step": 2763776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.740255103143878e-06,
+      "loss": 3.8515,
+      "step": 2764288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.731869155633359e-06,
+      "loss": 3.8586,
+      "step": 2764800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723515965730302e-06,
+      "loss": 3.8437,
+      "step": 2765312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7151300182197816e-06,
+      "loss": 3.8524,
+      "step": 2765824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7067440707092615e-06,
+      "loss": 3.8399,
+      "step": 2766336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6983581231987415e-06,
+      "loss": 3.8511,
+      "step": 2766848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6899721756882214e-06,
+      "loss": 3.8376,
+      "step": 2767360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681602606981433e-06,
+      "loss": 3.8452,
+      "step": 2767872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.673216659470912e-06,
+      "loss": 3.8408,
+      "step": 2768384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.664830711960393e-06,
+      "loss": 3.8546,
+      "step": 2768896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.656444764449872e-06,
+      "loss": 3.844,
+      "step": 2769408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.648058816939352e-06,
+      "loss": 3.844,
+      "step": 2769920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.639672869428832e-06,
+      "loss": 3.8593,
+      "step": 2770432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.631286921918312e-06,
+      "loss": 3.8504,
+      "step": 2770944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.622900974407792e-06,
+      "loss": 3.8545,
+      "step": 2771456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.614515026897272e-06,
+      "loss": 3.8362,
+      "step": 2771968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.606129079386751e-06,
+      "loss": 3.8333,
+      "step": 2772480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.597759510679963e-06,
+      "loss": 3.8523,
+      "step": 2772992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5893735631694426e-06,
+      "loss": 3.8477,
+      "step": 2773504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5809876156589225e-06,
+      "loss": 3.8393,
+      "step": 2774016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5726016681484025e-06,
+      "loss": 3.8384,
+      "step": 2774528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564215720637883e-06,
+      "loss": 3.8485,
+      "step": 2775040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.555829773127362e-06,
+      "loss": 3.8299,
+      "step": 2775552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.547443825616842e-06,
+      "loss": 3.8534,
+      "step": 2776064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.539074256910054e-06,
+      "loss": 3.834,
+      "step": 2776576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.530688309399533e-06,
+      "loss": 3.844,
+      "step": 2777088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.522302361889014e-06,
+      "loss": 3.8533,
+      "step": 2777600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.513916414378493e-06,
+      "loss": 3.8364,
+      "step": 2778112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.505530466867974e-06,
+      "loss": 3.8468,
+      "step": 2778624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.497160898161185e-06,
+      "loss": 3.844,
+      "step": 2779136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4887749506506645e-06,
+      "loss": 3.836,
+      "step": 2779648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4803890031401445e-06,
+      "loss": 3.8396,
+      "step": 2780160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4720030556296245e-06,
+      "loss": 3.8479,
+      "step": 2780672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463617108119104e-06,
+      "loss": 3.841,
+      "step": 2781184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455231160608584e-06,
+      "loss": 3.8437,
+      "step": 2781696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446845213098064e-06,
+      "loss": 3.8343,
+      "step": 2782208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438459265587544e-06,
+      "loss": 3.8349,
+      "step": 2782720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430089696880756e-06,
+      "loss": 3.846,
+      "step": 2783232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.421703749370235e-06,
+      "loss": 3.8571,
+      "step": 2783744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.413317801859715e-06,
+      "loss": 3.8412,
+      "step": 2784256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.404948233152927e-06,
+      "loss": 3.8571,
+      "step": 2784768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3965622856424066e-06,
+      "loss": 3.852,
+      "step": 2785280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3881763381318865e-06,
+      "loss": 3.8451,
+      "step": 2785792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.379806769425097e-06,
+      "loss": 3.8472,
+      "step": 2786304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.371420821914578e-06,
+      "loss": 3.8532,
+      "step": 2786816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.363034874404057e-06,
+      "loss": 3.8434,
+      "step": 2787328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.354648926893537e-06,
+      "loss": 3.8425,
+      "step": 2787840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346262979383017e-06,
+      "loss": 3.8496,
+      "step": 2788352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337877031872497e-06,
+      "loss": 3.8505,
+      "step": 2788864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329491084361977e-06,
+      "loss": 3.8531,
+      "step": 2789376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321105136851457e-06,
+      "loss": 3.8489,
+      "step": 2789888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312719189340936e-06,
+      "loss": 3.8329,
+      "step": 2790400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304349620634148e-06,
+      "loss": 3.8312,
+      "step": 2790912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.295963673123628e-06,
+      "loss": 3.846,
+      "step": 2791424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.287577725613108e-06,
+      "loss": 3.8496,
+      "step": 2791936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.279208156906319e-06,
+      "loss": 3.8429,
+      "step": 2792448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.270822209395799e-06,
+      "loss": 3.8489,
+      "step": 2792960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262436261885279e-06,
+      "loss": 3.8426,
+      "step": 2793472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254050314374759e-06,
+      "loss": 3.8386,
+      "step": 2793984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.245664366864239e-06,
+      "loss": 3.8443,
+      "step": 2794496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.237278419353718e-06,
+      "loss": 3.8484,
+      "step": 2795008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.228892471843199e-06,
+      "loss": 3.8519,
+      "step": 2795520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.220506524332678e-06,
+      "loss": 3.8502,
+      "step": 2796032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.212120576822159e-06,
+      "loss": 3.8484,
+      "step": 2796544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.203734629311638e-06,
+      "loss": 3.8387,
+      "step": 2797056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.19536506060485e-06,
+      "loss": 3.8495,
+      "step": 2797568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.18697911309433e-06,
+      "loss": 3.8462,
+      "step": 2798080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17859316558381e-06,
+      "loss": 3.8537,
+      "step": 2798592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1702072180732896e-06,
+      "loss": 3.8432,
+      "step": 2799104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1618212705627695e-06,
+      "loss": 3.8375,
+      "step": 2799616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1534353230522495e-06,
+      "loss": 3.8488,
+      "step": 2800128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14506575434546e-06,
+      "loss": 3.8505,
+      "step": 2800640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136679806834941e-06,
+      "loss": 3.8393,
+      "step": 2801152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12829385932442e-06,
+      "loss": 3.845,
+      "step": 2801664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1199079118139e-06,
+      "loss": 3.8449,
+      "step": 2802176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.11152196430338e-06,
+      "loss": 3.8359,
+      "step": 2802688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.10313601679286e-06,
+      "loss": 3.8341,
+      "step": 2803200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.09475006928234e-06,
+      "loss": 3.8438,
+      "step": 2803712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.08636412177182e-06,
+      "loss": 3.8413,
+      "step": 2804224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0779781742613e-06,
+      "loss": 3.8455,
+      "step": 2804736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.069624984358242e-06,
+      "loss": 3.8372,
+      "step": 2805248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.061239036847723e-06,
+      "loss": 3.8418,
+      "step": 2805760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.052853089337202e-06,
+      "loss": 3.8621,
+      "step": 2806272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044467141826682e-06,
+      "loss": 3.8597,
+      "step": 2806784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036081194316162e-06,
+      "loss": 3.8473,
+      "step": 2807296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027711625609373e-06,
+      "loss": 3.8472,
+      "step": 2807808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.019325678098854e-06,
+      "loss": 3.8426,
+      "step": 2808320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010939730588333e-06,
+      "loss": 3.8438,
+      "step": 2808832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0025701618815445e-06,
+      "loss": 3.8376,
+      "step": 2809344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9941842143710245e-06,
+      "loss": 3.8432,
+      "step": 2809856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9857982668605045e-06,
+      "loss": 3.8453,
+      "step": 2810368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.977412319349984e-06,
+      "loss": 3.8502,
+      "step": 2810880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.969026371839464e-06,
+      "loss": 3.8443,
+      "step": 2811392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.960640424328944e-06,
+      "loss": 3.8493,
+      "step": 2811904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.952254476818424e-06,
+      "loss": 3.8412,
+      "step": 2812416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.943868529307904e-06,
+      "loss": 3.8428,
+      "step": 2812928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.935498960601115e-06,
+      "loss": 3.8382,
+      "step": 2813440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.927113013090596e-06,
+      "loss": 3.8566,
+      "step": 2813952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918743444383807e-06,
+      "loss": 3.8559,
+      "step": 2814464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9103574968732866e-06,
+      "loss": 3.8517,
+      "step": 2814976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9019715493627665e-06,
+      "loss": 3.8415,
+      "step": 2815488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8935856018522465e-06,
+      "loss": 3.8503,
+      "step": 2816000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8851996543417264e-06,
+      "loss": 3.8463,
+      "step": 2816512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876813706831206e-06,
+      "loss": 3.859,
+      "step": 2817024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.868427759320686e-06,
+      "loss": 3.8437,
+      "step": 2817536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.860041811810166e-06,
+      "loss": 3.8464,
+      "step": 2818048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.851672243103378e-06,
+      "loss": 3.8459,
+      "step": 2818560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.843286295592857e-06,
+      "loss": 3.8517,
+      "step": 2819072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.834900348082337e-06,
+      "loss": 3.8407,
+      "step": 2819584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.826514400571817e-06,
+      "loss": 3.8456,
+      "step": 2820096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.818128453061297e-06,
+      "loss": 3.8501,
+      "step": 2820608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8097425055507773e-06,
+      "loss": 3.8432,
+      "step": 2821120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.801356558040257e-06,
+      "loss": 3.851,
+      "step": 2821632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7929706105297364e-06,
+      "loss": 3.8544,
+      "step": 2822144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7845846630192167e-06,
+      "loss": 3.8471,
+      "step": 2822656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7761987155086963e-06,
+      "loss": 3.8454,
+      "step": 2823168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.767812767998176e-06,
+      "loss": 3.8472,
+      "step": 2823680
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.971450090408325,
+      "eval_runtime": 291.081,
+      "eval_samples_per_second": 1310.944,
+      "eval_steps_per_second": 40.968,
+      "step": 2823828
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.1662509259763602e+18,
+  "trial_name": null,
+  "trial_params": null
+}