diff --git "a/checkpoint-763200/trainer_state.json" "b/checkpoint-763200/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-763200/trainer_state.json"
@@ -0,0 +1,9045 @@
+{
+  "best_metric": 4.075286388397217,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/passive/lstm/0/checkpoints/checkpoint-763200",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 763200,
+  "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": 10.8198,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5522,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0566,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.995,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9475,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9066,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7464,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6405,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993292879871958e-05,
+      "loss": 6.5391,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992454285120906e-05,
+      "loss": 6.457,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.4004,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.3311,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.2652,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.1988,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.1445,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0779,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0424,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.9952,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9566,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9196,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8897,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8524,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 5.8257,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.785,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798786396158716e-05,
+      "loss": 5.7734,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 5.7366,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 5.7188,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 5.6818,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6646,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6459,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6213,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.6129,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 5.5862,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5697,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5626,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5386,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5296,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4843,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4893,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4519,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4571,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4336,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4322,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.4024,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963110020355577e-05,
+      "loss": 5.4078,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.3962,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3903,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3811,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.3518,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3426,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3434,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3293,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.324,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.2998,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.2981,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.2777,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.298,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 5.2524,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2679,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2567,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2381,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 5.2352,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2224,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.2119,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.2041,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.2186,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 5.1918,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 5.1825,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.1742,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1712,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 5.1775,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 5.1722,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.148,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1631,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1554,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 5.1293,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 5.1272,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 5.1171,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 5.1092,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 5.1037,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 5.1087,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 5.1041,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 5.1052,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 5.0813,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 5.0642,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 5.0683,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 5.0714,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 5.0689,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92622823011302e-05,
+      "loss": 5.0524,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0499,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92455267849129e-05,
+      "loss": 5.0496,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923714083740238e-05,
+      "loss": 5.0403,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922875488989185e-05,
+      "loss": 5.0311,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0276,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921199937367455e-05,
+      "loss": 5.0154,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920361342616403e-05,
+      "loss": 5.0273,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919522747865351e-05,
+      "loss": 5.0144,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918684153114299e-05,
+      "loss": 4.9927,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917845558363247e-05,
+      "loss": 4.997,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.9953,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 5.0038,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.9853,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.9823,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.9665,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.9707,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.9517,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.9615,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.95,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.9536,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.9399,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907787334991742e-05,
+      "loss": 4.9436,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90694874024069e-05,
+      "loss": 4.9305,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.9334,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905271550738586e-05,
+      "loss": 4.926,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904432955987534e-05,
+      "loss": 4.9192,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.9306,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.9147,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901920447495124e-05,
+      "loss": 4.9128,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901081852744072e-05,
+      "loss": 4.9028,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.9081,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.9029,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.8958,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.8996,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968888789888125e-05,
+      "loss": 4.8859,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960502842377605e-05,
+      "loss": 4.8855,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952116894867085e-05,
+      "loss": 4.8663,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.8723,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.8737,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.8637,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8563,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.878,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.8603,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.8522,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8518,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.8461,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.8473,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885992060866256e-05,
+      "loss": 4.8439,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885153466115204e-05,
+      "loss": 4.8398,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884314871364151e-05,
+      "loss": 4.8412,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883476276613099e-05,
+      "loss": 4.8394,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882637681862047e-05,
+      "loss": 4.8276,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.8178,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.8285,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.8157,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879284940738212e-05,
+      "loss": 4.8194,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878447983867534e-05,
+      "loss": 4.81,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.8071,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876772432245803e-05,
+      "loss": 4.8133,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875933837494751e-05,
+      "loss": 4.7999,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8750952427436986e-05,
+      "loss": 4.7967,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.765356540679932,
+      "eval_runtime": 295.2513,
+      "eval_samples_per_second": 1292.428,
+      "eval_steps_per_second": 40.389,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8742566479926466e-05,
+      "loss": 4.785,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.7835,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8725794584905426e-05,
+      "loss": 4.8069,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.7913,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.7825,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700653121177595e-05,
+      "loss": 4.7796,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692267173667075e-05,
+      "loss": 4.7803,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683881226156555e-05,
+      "loss": 4.7619,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867549527864604e-05,
+      "loss": 4.7708,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.7703,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.7729,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865035381491821e-05,
+      "loss": 4.7724,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864198424621142e-05,
+      "loss": 4.754,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86335982987009e-05,
+      "loss": 4.7539,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862521235119038e-05,
+      "loss": 4.7427,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.7441,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860844045616934e-05,
+      "loss": 4.7515,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860005450865882e-05,
+      "loss": 4.7408,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859168493995203e-05,
+      "loss": 4.7454,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.7588,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.7402,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.7456,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.7272,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549755202399435e-05,
+      "loss": 4.746,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.7239,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532999686182124e-05,
+      "loss": 4.7276,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524613738671604e-05,
+      "loss": 4.7133,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8516227791161084e-05,
+      "loss": 4.7231,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8507841843650564e-05,
+      "loss": 4.714,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8499455896140044e-05,
+      "loss": 4.7107,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8491069948629524e-05,
+      "loss": 4.7158,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482684001119e-05,
+      "loss": 4.7195,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847431443241221e-05,
+      "loss": 4.7115,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846592848490169e-05,
+      "loss": 4.7133,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.7049,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.7127,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844078702117387e-05,
+      "loss": 4.6846,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843240107366335e-05,
+      "loss": 4.7,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842401512615282e-05,
+      "loss": 4.6753,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84156291786423e-05,
+      "loss": 4.6969,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840724323113178e-05,
+      "loss": 4.6798,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839885728362126e-05,
+      "loss": 4.6931,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839048771491447e-05,
+      "loss": 4.6707,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838210176740395e-05,
+      "loss": 4.6883,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373732198697167e-05,
+      "loss": 4.6851,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365346251186646e-05,
+      "loss": 4.6823,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8356960303676126e-05,
+      "loss": 4.6896,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8348574356165606e-05,
+      "loss": 4.6592,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8340188408655086e-05,
+      "loss": 4.6627,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331802461144566e-05,
+      "loss": 4.6764,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323416513634046e-05,
+      "loss": 4.6683,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315030566123526e-05,
+      "loss": 4.6661,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306644618613006e-05,
+      "loss": 4.6588,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298258671102486e-05,
+      "loss": 4.6525,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289872723591966e-05,
+      "loss": 4.6464,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281486776081446e-05,
+      "loss": 4.6638,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273117207374655e-05,
+      "loss": 4.6391,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826474763866787e-05,
+      "loss": 4.6505,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825636169115735e-05,
+      "loss": 4.6565,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824797574364683e-05,
+      "loss": 4.6378,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823958979613631e-05,
+      "loss": 4.639,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823120384862579e-05,
+      "loss": 4.6387,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822281790111527e-05,
+      "loss": 4.6248,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821443195360475e-05,
+      "loss": 4.6308,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820604600609423e-05,
+      "loss": 4.6469,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819767643738744e-05,
+      "loss": 4.6288,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818929048987692e-05,
+      "loss": 4.6233,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81809045423664e-05,
+      "loss": 4.6215,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817253497365961e-05,
+      "loss": 4.6225,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816414902614909e-05,
+      "loss": 4.6368,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815576307863857e-05,
+      "loss": 4.6287,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814737713112805e-05,
+      "loss": 4.6186,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138991183617535e-05,
+      "loss": 4.6339,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813060523610701e-05,
+      "loss": 4.6298,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812221928859649e-05,
+      "loss": 4.6157,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113849719889704e-05,
+      "loss": 4.6125,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105463772379184e-05,
+      "loss": 4.6117,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809707782486866e-05,
+      "loss": 4.6085,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808869187735814e-05,
+      "loss": 4.6052,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808030592984762e-05,
+      "loss": 4.6101,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80719199823371e-05,
+      "loss": 4.6155,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063550413630307e-05,
+      "loss": 4.6164,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055164466119786e-05,
+      "loss": 4.6042,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804677851860927e-05,
+      "loss": 4.5821,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803839257109875e-05,
+      "loss": 4.5978,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803000662358823e-05,
+      "loss": 4.6017,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802162067607771e-05,
+      "loss": 4.6037,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801323472856719e-05,
+      "loss": 4.5955,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80048651598604e-05,
+      "loss": 4.5924,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799647921234988e-05,
+      "loss": 4.5976,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798809326483936e-05,
+      "loss": 4.5901,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797970731732884e-05,
+      "loss": 4.5851,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797132136981832e-05,
+      "loss": 4.5917,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79629354223078e-05,
+      "loss": 4.5754,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795454947479728e-05,
+      "loss": 4.5956,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794616352728676e-05,
+      "loss": 4.5863,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793777757977624e-05,
+      "loss": 4.5659,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792940801106946e-05,
+      "loss": 4.5785,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792103844236267e-05,
+      "loss": 4.5765,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791265249485215e-05,
+      "loss": 4.5881,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790426654734163e-05,
+      "loss": 4.5757,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895880599831107e-05,
+      "loss": 4.571,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887494652320587e-05,
+      "loss": 4.5641,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879108704810067e-05,
+      "loss": 4.573,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870722757299546e-05,
+      "loss": 4.5551,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862353188592756e-05,
+      "loss": 4.5646,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853967241082236e-05,
+      "loss": 4.5609,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845581293571715e-05,
+      "loss": 4.5635,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837195346061195e-05,
+      "loss": 4.5525,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828809398550675e-05,
+      "loss": 4.5622,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820423451040155e-05,
+      "loss": 4.5551,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812037503529635e-05,
+      "loss": 4.5553,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803651556019115e-05,
+      "loss": 4.5572,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795265608508595e-05,
+      "loss": 4.5466,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786896039801804e-05,
+      "loss": 4.5705,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778510092291284e-05,
+      "loss": 4.5544,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770124144780764e-05,
+      "loss": 4.5537,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761738197270244e-05,
+      "loss": 4.5438,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7753352249759724e-05,
+      "loss": 4.5499,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774498268105293e-05,
+      "loss": 4.5502,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773659673354241e-05,
+      "loss": 4.5402,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772821078603189e-05,
+      "loss": 4.5556,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771982483852137e-05,
+      "loss": 4.5424,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771145526981459e-05,
+      "loss": 4.5371,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770306932230407e-05,
+      "loss": 4.5334,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769468337479355e-05,
+      "loss": 4.5353,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768629742728303e-05,
+      "loss": 4.5338,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767792785857624e-05,
+      "loss": 4.5353,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766955828986945e-05,
+      "loss": 4.5217,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766117234235893e-05,
+      "loss": 4.5502,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765278639484841e-05,
+      "loss": 4.5402,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764440044733789e-05,
+      "loss": 4.5306,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763601449982737e-05,
+      "loss": 4.5287,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762762855231685e-05,
+      "loss": 4.5234,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761924260480633e-05,
+      "loss": 4.5343,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610856657295813e-05,
+      "loss": 4.5312,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7602470709785293e-05,
+      "loss": 4.5271,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75941011410785e-05,
+      "loss": 4.532,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758571519356798e-05,
+      "loss": 4.528,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757732924605746e-05,
+      "loss": 4.5252,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756894329854694e-05,
+      "loss": 4.5177,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756055735103642e-05,
+      "loss": 4.5303,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755218778232963e-05,
+      "loss": 4.5181,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754380183481911e-05,
+      "loss": 4.5205,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753541588730859e-05,
+      "loss": 4.5149,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752702993979807e-05,
+      "loss": 4.5163,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751864399228755e-05,
+      "loss": 4.5176,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751025804477703e-05,
+      "loss": 4.5059,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750188847607025e-05,
+      "loss": 4.5105,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.484802722930908,
+      "eval_runtime": 296.4897,
+      "eval_samples_per_second": 1287.029,
+      "eval_steps_per_second": 40.221,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749350252855973e-05,
+      "loss": 4.4981,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748511658104921e-05,
+      "loss": 4.4961,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747673063353868e-05,
+      "loss": 4.5236,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746834468602816e-05,
+      "loss": 4.5142,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745995873851764e-05,
+      "loss": 4.5038,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745157279100712e-05,
+      "loss": 4.5012,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74431868434966e-05,
+      "loss": 4.506,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743480089598608e-05,
+      "loss": 4.4896,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742641494847556e-05,
+      "loss": 4.5071,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741802900096504e-05,
+      "loss": 4.5003,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740964305345452e-05,
+      "loss": 4.5039,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401273484747736e-05,
+      "loss": 4.5054,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7392887537237216e-05,
+      "loss": 4.4864,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384501589726696e-05,
+      "loss": 4.4917,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376115642216176e-05,
+      "loss": 4.4858,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7367729694705656e-05,
+      "loss": 4.4861,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359343747195135e-05,
+      "loss": 4.4895,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350974178488345e-05,
+      "loss": 4.482,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342588230977825e-05,
+      "loss": 4.4888,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7334202283467305e-05,
+      "loss": 4.5077,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7325816335956784e-05,
+      "loss": 4.4862,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7317430388446264e-05,
+      "loss": 4.4911,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730904444093574e-05,
+      "loss": 4.4839,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730065849342522e-05,
+      "loss": 4.4964,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292272545914704e-05,
+      "loss": 4.4743,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7283886598404184e-05,
+      "loss": 4.487,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275500650893664e-05,
+      "loss": 4.4695,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267114703383144e-05,
+      "loss": 4.4742,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258728755872624e-05,
+      "loss": 4.4732,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250342808362104e-05,
+      "loss": 4.4694,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7241956860851584e-05,
+      "loss": 4.4769,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233570913341064e-05,
+      "loss": 4.4819,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225184965830544e-05,
+      "loss": 4.4747,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721681539712375e-05,
+      "loss": 4.4808,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720844582841696e-05,
+      "loss": 4.4664,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720005988090644e-05,
+      "loss": 4.4834,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719169031219965e-05,
+      "loss": 4.4518,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718330436468914e-05,
+      "loss": 4.4728,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717491841717862e-05,
+      "loss": 4.4466,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71665324696681e-05,
+      "loss": 4.4688,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715814652215758e-05,
+      "loss": 4.4563,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714976057464706e-05,
+      "loss": 4.4698,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714137462713654e-05,
+      "loss": 4.4504,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713298867962602e-05,
+      "loss": 4.4623,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71246027321155e-05,
+      "loss": 4.4651,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711623316340871e-05,
+      "loss": 4.4636,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710784721589819e-05,
+      "loss": 4.4675,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099461268387667e-05,
+      "loss": 4.4449,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091075320877147e-05,
+      "loss": 4.4427,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082705752170356e-05,
+      "loss": 4.4636,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074319804659836e-05,
+      "loss": 4.4548,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706593385714932e-05,
+      "loss": 4.4496,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70575479096388e-05,
+      "loss": 4.4486,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704916196212828e-05,
+      "loss": 4.4431,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704079239342149e-05,
+      "loss": 4.4343,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703240644591097e-05,
+      "loss": 4.4506,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702402049840045e-05,
+      "loss": 4.4322,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015634550889924e-05,
+      "loss": 4.4423,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007248603379404e-05,
+      "loss": 4.4528,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998862655868884e-05,
+      "loss": 4.4318,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69904930871621e-05,
+      "loss": 4.4342,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982107139651573e-05,
+      "loss": 4.438,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697372119214106e-05,
+      "loss": 4.4226,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696533524463054e-05,
+      "loss": 4.4251,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695694929712002e-05,
+      "loss": 4.4461,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69485633496095e-05,
+      "loss": 4.4319,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694019378090271e-05,
+      "loss": 4.4243,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693180783339219e-05,
+      "loss": 4.4247,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692342188588167e-05,
+      "loss": 4.4279,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691503593837115e-05,
+      "loss": 4.4411,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690664999086063e-05,
+      "loss": 4.4358,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689828042215384e-05,
+      "loss": 4.424,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688989447464332e-05,
+      "loss": 4.4405,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68815085271328e-05,
+      "loss": 4.4339,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687312257962228e-05,
+      "loss": 4.4245,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686473663211176e-05,
+      "loss": 4.4208,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856367063404974e-05,
+      "loss": 4.4274,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6847981115894454e-05,
+      "loss": 4.4217,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683961154718766e-05,
+      "loss": 4.4193,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683122559967714e-05,
+      "loss": 4.4202,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682283965216662e-05,
+      "loss": 4.4273,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68144537046561e-05,
+      "loss": 4.4322,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680606775714558e-05,
+      "loss": 4.418,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679768180963506e-05,
+      "loss": 4.4032,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678929586212454e-05,
+      "loss": 4.4109,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678090991461402e-05,
+      "loss": 4.4159,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67725239671035e-05,
+      "loss": 4.423,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676415439839671e-05,
+      "loss": 4.4125,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675576845088619e-05,
+      "loss": 4.4107,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674738250337568e-05,
+      "loss": 4.4209,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673899655586516e-05,
+      "loss": 4.407,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673062698715837e-05,
+      "loss": 4.4078,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722257418451576e-05,
+      "loss": 4.412,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6713871470941056e-05,
+      "loss": 4.4042,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705485523430536e-05,
+      "loss": 4.4219,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697099575920016e-05,
+      "loss": 4.4067,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688713628409496e-05,
+      "loss": 4.3949,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680327680898976e-05,
+      "loss": 4.4056,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6671958112192185e-05,
+      "loss": 4.4026,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663572164681665e-05,
+      "loss": 4.4202,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655186217171145e-05,
+      "loss": 4.4087,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664680026966063e-05,
+      "loss": 4.3994,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663841432215011e-05,
+      "loss": 4.3899,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663002837463959e-05,
+      "loss": 4.404,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621642427129065e-05,
+      "loss": 4.3874,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613256479618545e-05,
+      "loss": 4.3986,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604870532108025e-05,
+      "loss": 4.3942,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596484584597505e-05,
+      "loss": 4.3958,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588115015890714e-05,
+      "loss": 4.3849,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579729068380194e-05,
+      "loss": 4.4033,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571343120869674e-05,
+      "loss": 4.3877,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562957173359154e-05,
+      "loss": 4.3884,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554571225848634e-05,
+      "loss": 4.3971,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654620165714185e-05,
+      "loss": 4.3843,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653781570963133e-05,
+      "loss": 4.4062,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652944614092454e-05,
+      "loss": 4.3935,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652106019341402e-05,
+      "loss": 4.3938,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65126742459035e-05,
+      "loss": 4.3856,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650428829839298e-05,
+      "loss": 4.3896,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649590235088246e-05,
+      "loss": 4.3915,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648753278217567e-05,
+      "loss": 4.3804,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647914683466515e-05,
+      "loss": 4.3956,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647076088715463e-05,
+      "loss": 4.3881,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646237493964411e-05,
+      "loss": 4.3805,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645398899213359e-05,
+      "loss": 4.3823,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644560304462307e-05,
+      "loss": 4.3803,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437217097112554e-05,
+      "loss": 4.3731,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642884752840576e-05,
+      "loss": 4.3875,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642046158089524e-05,
+      "loss": 4.371,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641207563338472e-05,
+      "loss": 4.3925,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64036896858742e-05,
+      "loss": 4.3894,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639530373836368e-05,
+      "loss": 4.3771,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638691779085316e-05,
+      "loss": 4.3755,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637853184334264e-05,
+      "loss": 4.3741,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637014589583212e-05,
+      "loss": 4.3834,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361759948321596e-05,
+      "loss": 4.3815,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635339037961481e-05,
+      "loss": 4.3775,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634502081090802e-05,
+      "loss": 4.3807,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633663486339751e-05,
+      "loss": 4.3795,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632824891588699e-05,
+      "loss": 4.3808,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631986296837647e-05,
+      "loss": 4.3689,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631147702086595e-05,
+      "loss": 4.387,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630309107335542e-05,
+      "loss": 4.3739,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62947051258449e-05,
+      "loss": 4.3712,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628633555713812e-05,
+      "loss": 4.3726,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277949609627597e-05,
+      "loss": 4.3748,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626956366211707e-05,
+      "loss": 4.3699,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626117771460655e-05,
+      "loss": 4.3608,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625279176709603e-05,
+      "loss": 4.364,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.348752975463867,
+      "eval_runtime": 286.9409,
+      "eval_samples_per_second": 1329.859,
+      "eval_steps_per_second": 41.559,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244422198389246e-05,
+      "loss": 4.366,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6236036250878726e-05,
+      "loss": 4.3528,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622766668217194e-05,
+      "loss": 4.3814,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621928073466142e-05,
+      "loss": 4.3731,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210894787150895e-05,
+      "loss": 4.3682,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202508839640374e-05,
+      "loss": 4.357,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194122892129854e-05,
+      "loss": 4.3695,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185736944619334e-05,
+      "loss": 4.3493,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177350997108814e-05,
+      "loss": 4.3695,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168965049598294e-05,
+      "loss": 4.3651,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61605954808915e-05,
+      "loss": 4.3624,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615220953338098e-05,
+      "loss": 4.3725,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614382358587046e-05,
+      "loss": 4.349,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613543763835994e-05,
+      "loss": 4.3587,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612705169084942e-05,
+      "loss": 4.3471,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611866574333891e-05,
+      "loss": 4.3542,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611027979582839e-05,
+      "loss": 4.3514,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610189384831787e-05,
+      "loss": 4.3499,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609350790080735e-05,
+      "loss": 4.3532,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608512195329683e-05,
+      "loss": 4.3726,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607673600578631e-05,
+      "loss": 4.3523,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606835005827578e-05,
+      "loss": 4.3582,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605999686837273e-05,
+      "loss": 4.3521,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605161092086221e-05,
+      "loss": 4.3643,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604322497335169e-05,
+      "loss": 4.344,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603483902584117e-05,
+      "loss": 4.3524,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602645307833065e-05,
+      "loss": 4.3407,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601806713082013e-05,
+      "loss": 4.3423,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600968118330961e-05,
+      "loss": 4.3412,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600129523579909e-05,
+      "loss": 4.3459,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599290928828857e-05,
+      "loss": 4.3437,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598452334077805e-05,
+      "loss": 4.3541,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597613739326753e-05,
+      "loss": 4.3464,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596775144575701e-05,
+      "loss": 4.3502,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595939825585395e-05,
+      "loss": 4.3434,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595101230834343e-05,
+      "loss": 4.3531,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5942626360832906e-05,
+      "loss": 4.3323,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593425679212612e-05,
+      "loss": 4.3451,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59258708446156e-05,
+      "loss": 4.3213,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591748489710508e-05,
+      "loss": 4.3408,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590909894959456e-05,
+      "loss": 4.3355,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590071300208404e-05,
+      "loss": 4.3473,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589232705457352e-05,
+      "loss": 4.3288,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5883941107063e-05,
+      "loss": 4.3392,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587555515955248e-05,
+      "loss": 4.3421,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586718559084569e-05,
+      "loss": 4.3452,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585879964333517e-05,
+      "loss": 4.3392,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585041369582465e-05,
+      "loss": 4.3278,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584202774831413e-05,
+      "loss": 4.319,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583364180080361e-05,
+      "loss": 4.3424,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582527223209682e-05,
+      "loss": 4.337,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58168862845863e-05,
+      "loss": 4.3304,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808516715879515e-05,
+      "loss": 4.327,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800130768368995e-05,
+      "loss": 4.3281,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791744820858475e-05,
+      "loss": 4.3115,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783375252151684e-05,
+      "loss": 4.3353,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5774989304641164e-05,
+      "loss": 4.3112,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766603357130644e-05,
+      "loss": 4.3256,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758217409620124e-05,
+      "loss": 4.3326,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5749831462109604e-05,
+      "loss": 4.3201,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741445514599084e-05,
+      "loss": 4.3135,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733059567088564e-05,
+      "loss": 4.3226,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724673619578044e-05,
+      "loss": 4.3064,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716287672067524e-05,
+      "loss": 4.314,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570791810336074e-05,
+      "loss": 4.3265,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569953215585022e-05,
+      "loss": 4.318,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56911462083397e-05,
+      "loss": 4.3102,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568276026082918e-05,
+      "loss": 4.306,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567437431331866e-05,
+      "loss": 4.3142,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566598836580814e-05,
+      "loss": 4.3224,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565760241829762e-05,
+      "loss": 4.3221,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564921647078709e-05,
+      "loss": 4.315,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564084690208031e-05,
+      "loss": 4.3246,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563246095456979e-05,
+      "loss": 4.3248,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562407500705927e-05,
+      "loss": 4.3102,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561568905954874e-05,
+      "loss": 4.309,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560730311203822e-05,
+      "loss": 4.3175,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559891716452771e-05,
+      "loss": 4.3128,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559053121701719e-05,
+      "loss": 4.311,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558214526950667e-05,
+      "loss": 4.3075,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557377570079988e-05,
+      "loss": 4.315,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556538975328936e-05,
+      "loss": 4.321,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555700380577884e-05,
+      "loss": 4.3115,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554861785826832e-05,
+      "loss": 4.2955,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5540248289561526e-05,
+      "loss": 4.2983,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5531862342051006e-05,
+      "loss": 4.3058,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523476394540486e-05,
+      "loss": 4.3146,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5515090447029966e-05,
+      "loss": 4.3093,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5506720878323175e-05,
+      "loss": 4.2987,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498334930812655e-05,
+      "loss": 4.3116,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548996536210587e-05,
+      "loss": 4.3041,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548157941459535e-05,
+      "loss": 4.3001,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547319346708483e-05,
+      "loss": 4.3022,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546480751957431e-05,
+      "loss": 4.3061,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545642157206379e-05,
+      "loss": 4.3153,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544803562455327e-05,
+      "loss": 4.2963,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543964967704275e-05,
+      "loss": 4.2864,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543126372953223e-05,
+      "loss": 4.3042,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542287778202171e-05,
+      "loss": 4.2981,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541450821331492e-05,
+      "loss": 4.3129,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54061222658044e-05,
+      "loss": 4.3055,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539773631829388e-05,
+      "loss": 4.2937,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538935037078336e-05,
+      "loss": 4.2878,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538096442327284e-05,
+      "loss": 4.3039,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372594854566055e-05,
+      "loss": 4.2841,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364208907055535e-05,
+      "loss": 4.2963,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5355822959545015e-05,
+      "loss": 4.2889,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347437012034495e-05,
+      "loss": 4.2973,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339051064523975e-05,
+      "loss": 4.2819,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5330665117013455e-05,
+      "loss": 4.3016,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5322295548306664e-05,
+      "loss": 4.2872,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5313909600796144e-05,
+      "loss": 4.2846,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305523653285624e-05,
+      "loss": 4.2996,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297137705775104e-05,
+      "loss": 4.2831,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528875175826458e-05,
+      "loss": 4.309,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528038218955779e-05,
+      "loss": 4.2958,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527201262085101e-05,
+      "loss": 4.2937,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526362667334049e-05,
+      "loss": 4.2863,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525524072582997e-05,
+      "loss": 4.2901,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524685477831945e-05,
+      "loss": 4.2905,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523846883080893e-05,
+      "loss": 4.2849,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52300828832984e-05,
+      "loss": 4.295,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522169693578788e-05,
+      "loss": 4.2888,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521331098827736e-05,
+      "loss": 4.2837,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520494141957057e-05,
+      "loss": 4.2852,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519655547206005e-05,
+      "loss": 4.2854,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518816952454953e-05,
+      "loss": 4.2779,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517978357703902e-05,
+      "loss": 4.293,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517141400833223e-05,
+      "loss": 4.2739,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516302806082171e-05,
+      "loss": 4.293,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515465849211492e-05,
+      "loss": 4.2889,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51462725446044e-05,
+      "loss": 4.2874,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5137886597093876e-05,
+      "loss": 4.278,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129500649583356e-05,
+      "loss": 4.2803,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121114702072836e-05,
+      "loss": 4.2898,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112728754562316e-05,
+      "loss": 4.2893,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104342807051795e-05,
+      "loss": 4.2788,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5095956859541275e-05,
+      "loss": 4.286,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087570912030755e-05,
+      "loss": 4.2844,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079184964520235e-05,
+      "loss": 4.2908,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070799017009715e-05,
+      "loss": 4.2756,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062413069499195e-05,
+      "loss": 4.2918,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505402712198868e-05,
+      "loss": 4.2837,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504565755328189e-05,
+      "loss": 4.2786,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503727160577137e-05,
+      "loss": 4.2777,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502888565826085e-05,
+      "loss": 4.2852,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502049971075033e-05,
+      "loss": 4.2756,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501211376323981e-05,
+      "loss": 4.2644,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500374419453302e-05,
+      "loss": 4.2808,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.265576362609863,
+      "eval_runtime": 284.2057,
+      "eval_samples_per_second": 1342.658,
+      "eval_steps_per_second": 41.959,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49953582470225e-05,
+      "loss": 4.2686,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498697229951198e-05,
+      "loss": 4.2613,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497858635200146e-05,
+      "loss": 4.2868,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497020040449093e-05,
+      "loss": 4.2825,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496181445698042e-05,
+      "loss": 4.2788,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49534285094699e-05,
+      "loss": 4.2714,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494504256195938e-05,
+      "loss": 4.2755,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493665661444886e-05,
+      "loss": 4.2564,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492827066693834e-05,
+      "loss": 4.2801,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491988471942782e-05,
+      "loss": 4.2775,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49114987719173e-05,
+      "loss": 4.2703,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490311282440678e-05,
+      "loss": 4.2825,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489474325569999e-05,
+      "loss": 4.2615,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488635730818947e-05,
+      "loss": 4.2711,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487797136067895e-05,
+      "loss": 4.2626,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486958541316843e-05,
+      "loss": 4.2605,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486119946565791e-05,
+      "loss": 4.2641,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485281351814739e-05,
+      "loss": 4.2616,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484442757063687e-05,
+      "loss": 4.2669,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483604162312635e-05,
+      "loss": 4.2819,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4827672054419564e-05,
+      "loss": 4.2703,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4819286106909044e-05,
+      "loss": 4.2753,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4810900159398524e-05,
+      "loss": 4.2647,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4802514211888004e-05,
+      "loss": 4.2747,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479412826437748e-05,
+      "loss": 4.26,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478575869567069e-05,
+      "loss": 4.2663,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477737274816017e-05,
+      "loss": 4.2586,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476898680064965e-05,
+      "loss": 4.2541,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4760600853139126e-05,
+      "loss": 4.2539,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752214905628606e-05,
+      "loss": 4.261,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474384533692182e-05,
+      "loss": 4.2595,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47354593894113e-05,
+      "loss": 4.2697,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472707344190078e-05,
+      "loss": 4.2638,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471868749439026e-05,
+      "loss": 4.2664,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471031792568348e-05,
+      "loss": 4.2563,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470193197817295e-05,
+      "loss": 4.2709,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469354603066243e-05,
+      "loss": 4.252,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468516008315191e-05,
+      "loss": 4.2576,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467677413564139e-05,
+      "loss": 4.2414,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466838818813087e-05,
+      "loss": 4.2553,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466000224062035e-05,
+      "loss": 4.2538,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465161629310983e-05,
+      "loss": 4.2635,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464324672440304e-05,
+      "loss": 4.2469,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4634860776892526e-05,
+      "loss": 4.2577,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626474829382006e-05,
+      "loss": 4.2597,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618088881871486e-05,
+      "loss": 4.262,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4609702934360966e-05,
+      "loss": 4.2588,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4601333365654175e-05,
+      "loss": 4.2439,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4592947418143655e-05,
+      "loss": 4.2358,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584577849436864e-05,
+      "loss": 4.2622,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576191901926344e-05,
+      "loss": 4.2546,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567805954415824e-05,
+      "loss": 4.2488,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559420006905304e-05,
+      "loss": 4.2493,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551034059394784e-05,
+      "loss": 4.2469,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542648111884264e-05,
+      "loss": 4.2312,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534262164373744e-05,
+      "loss": 4.2499,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4525876216863224e-05,
+      "loss": 4.2348,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517490269352704e-05,
+      "loss": 4.2431,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450912070064592e-05,
+      "loss": 4.2552,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45007347531354e-05,
+      "loss": 4.2395,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449234880562488e-05,
+      "loss": 4.2344,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448397923691809e-05,
+      "loss": 4.2484,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447559328940757e-05,
+      "loss": 4.2262,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446720734189705e-05,
+      "loss": 4.2383,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445882139438653e-05,
+      "loss": 4.2451,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445043544687601e-05,
+      "loss": 4.2433,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444204949936548e-05,
+      "loss": 4.2332,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443366355185496e-05,
+      "loss": 4.2317,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442527760434444e-05,
+      "loss": 4.2337,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441689165683393e-05,
+      "loss": 4.2447,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440850570932341e-05,
+      "loss": 4.2441,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440013614061662e-05,
+      "loss": 4.2392,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43917501931061e-05,
+      "loss": 4.2473,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438336424559558e-05,
+      "loss": 4.2494,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437497829808506e-05,
+      "loss": 4.2365,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436659235057454e-05,
+      "loss": 4.2291,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435820640306402e-05,
+      "loss": 4.2465,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4349836834357227e-05,
+      "loss": 4.2391,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4341450886846707e-05,
+      "loss": 4.2317,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333064939336186e-05,
+      "loss": 4.2336,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4324678991825666e-05,
+      "loss": 4.2407,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316293044315146e-05,
+      "loss": 4.248,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4307907096804626e-05,
+      "loss": 4.2341,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429952114929411e-05,
+      "loss": 4.2216,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429115158058732e-05,
+      "loss": 4.2305,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42827656330768e-05,
+      "loss": 4.2251,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427437968556628e-05,
+      "loss": 4.2401,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426599373805576e-05,
+      "loss": 4.2359,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425760779054524e-05,
+      "loss": 4.2232,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424923822183845e-05,
+      "loss": 4.239,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424085227432793e-05,
+      "loss": 4.2286,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423246632681741e-05,
+      "loss": 4.2319,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422408037930689e-05,
+      "loss": 4.2252,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421569443179637e-05,
+      "loss": 4.2309,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420730848428585e-05,
+      "loss": 4.2442,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419892253677533e-05,
+      "loss": 4.229,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190569346872276e-05,
+      "loss": 4.2115,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182183399361756e-05,
+      "loss": 4.2284,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4173797451851236e-05,
+      "loss": 4.228,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165411504340716e-05,
+      "loss": 4.2352,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157025556830196e-05,
+      "loss": 4.2376,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4148639609319676e-05,
+      "loss": 4.2229,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414025366180915e-05,
+      "loss": 4.2179,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413186771429863e-05,
+      "loss": 4.2312,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123498145591845e-05,
+      "loss": 4.2083,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411511219808132e-05,
+      "loss": 4.2252,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4106726250570805e-05,
+      "loss": 4.2173,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4098340303060284e-05,
+      "loss": 4.2269,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4089954355549764e-05,
+      "loss": 4.2133,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4081584786842974e-05,
+      "loss": 4.2298,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407321521813619e-05,
+      "loss": 4.2157,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406482927062567e-05,
+      "loss": 4.2143,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405644332311514e-05,
+      "loss": 4.2289,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404805737560462e-05,
+      "loss": 4.2173,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40396714280941e-05,
+      "loss": 4.2373,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403128548058358e-05,
+      "loss": 4.2256,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402291591187679e-05,
+      "loss": 4.2195,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401452996436627e-05,
+      "loss": 4.2261,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400614401685576e-05,
+      "loss": 4.2136,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399775806934524e-05,
+      "loss": 4.223,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398938850063845e-05,
+      "loss": 4.216,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398100255312793e-05,
+      "loss": 4.2279,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397261660561741e-05,
+      "loss": 4.2168,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396423065810689e-05,
+      "loss": 4.2133,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395584471059637e-05,
+      "loss": 4.2162,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394745876308585e-05,
+      "loss": 4.2193,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393907281557533e-05,
+      "loss": 4.2096,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393068686806481e-05,
+      "loss": 4.2229,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392230092055429e-05,
+      "loss": 4.2092,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3913931351847496e-05,
+      "loss": 4.2254,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905545404336976e-05,
+      "loss": 4.2192,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897159456826456e-05,
+      "loss": 4.2202,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3888773509315936e-05,
+      "loss": 4.2113,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388040394060915e-05,
+      "loss": 4.2122,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387201799309863e-05,
+      "loss": 4.2254,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386364842439184e-05,
+      "loss": 4.2173,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385526247688132e-05,
+      "loss": 4.2123,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38468765293708e-05,
+      "loss": 4.2219,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383849058186028e-05,
+      "loss": 4.2171,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383010463434976e-05,
+      "loss": 4.2248,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382173506564297e-05,
+      "loss": 4.2127,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381334911813245e-05,
+      "loss": 4.2229,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380496317062193e-05,
+      "loss": 4.2153,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379657722311141e-05,
+      "loss": 4.2126,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378819127560089e-05,
+      "loss": 4.2145,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779805328090376e-05,
+      "loss": 4.2153,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771419380579856e-05,
+      "loss": 4.2131,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763033433069336e-05,
+      "loss": 4.2008,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375464748555881e-05,
+      "loss": 4.2147,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.208554267883301,
+      "eval_runtime": 299.6848,
+      "eval_samples_per_second": 1273.308,
+      "eval_steps_per_second": 39.792,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374626153804829e-05,
+      "loss": 4.2084,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373787559053777e-05,
+      "loss": 4.1978,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372948964302725e-05,
+      "loss": 4.22,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372112007432046e-05,
+      "loss": 4.22,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371273412680994e-05,
+      "loss": 4.2119,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370434817929942e-05,
+      "loss": 4.2075,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36959622317889e-05,
+      "loss": 4.2095,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368757628427838e-05,
+      "loss": 4.1966,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367919033676786e-05,
+      "loss": 4.2111,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3670820768061074e-05,
+      "loss": 4.2163,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3662434820550554e-05,
+      "loss": 4.2065,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365406525184376e-05,
+      "loss": 4.2147,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364567930433324e-05,
+      "loss": 4.1994,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363729335682272e-05,
+      "loss": 4.2089,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36289074093122e-05,
+      "loss": 4.1973,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362052146180168e-05,
+      "loss": 4.1983,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361213551429116e-05,
+      "loss": 4.1998,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360374956678064e-05,
+      "loss": 4.1997,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359536361927012e-05,
+      "loss": 4.205,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358699405056333e-05,
+      "loss": 4.2149,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357860810305281e-05,
+      "loss": 4.2128,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35702221555423e-05,
+      "loss": 4.2126,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356183620803178e-05,
+      "loss": 4.2013,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355345026052126e-05,
+      "loss": 4.2152,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354506431301074e-05,
+      "loss": 4.1951,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353667836550022e-05,
+      "loss": 4.2078,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352829241798969e-05,
+      "loss": 4.1973,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351990647047917e-05,
+      "loss": 4.1897,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351153690177239e-05,
+      "loss": 4.1963,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3503167333065596e-05,
+      "loss": 4.1975,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3494781385555076e-05,
+      "loss": 4.2015,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486395438044556e-05,
+      "loss": 4.2058,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478009490534036e-05,
+      "loss": 4.2071,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3469623543023516e-05,
+      "loss": 4.2053,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346125397431673e-05,
+      "loss": 4.1949,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345286802680621e-05,
+      "loss": 4.2069,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344448207929569e-05,
+      "loss": 4.1948,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3436096131785165e-05,
+      "loss": 4.1915,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3427710184274645e-05,
+      "loss": 4.1868,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419324236764125e-05,
+      "loss": 4.1923,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3410938289253605e-05,
+      "loss": 4.1977,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402552341743085e-05,
+      "loss": 4.2024,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3394182773036294e-05,
+      "loss": 4.1899,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385796825525774e-05,
+      "loss": 4.1967,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377410878015254e-05,
+      "loss": 4.2019,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3369024930504734e-05,
+      "loss": 4.2004,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3360638982994214e-05,
+      "loss": 4.2021,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335226941428743e-05,
+      "loss": 4.1864,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334388346677691e-05,
+      "loss": 4.1742,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333551389807012e-05,
+      "loss": 4.2047,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33271279505596e-05,
+      "loss": 4.196,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331874200304908e-05,
+      "loss": 4.1941,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331035605553856e-05,
+      "loss": 4.1921,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330197010802804e-05,
+      "loss": 4.1828,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329358416051752e-05,
+      "loss": 4.176,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328521459181073e-05,
+      "loss": 4.188,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327684502310394e-05,
+      "loss": 4.1776,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3268459075593424e-05,
+      "loss": 4.1862,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3260073128082904e-05,
+      "loss": 4.1972,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3251687180572383e-05,
+      "loss": 4.182,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3243301233061863e-05,
+      "loss": 4.1755,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323491528555134e-05,
+      "loss": 4.1911,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322654571684455e-05,
+      "loss": 4.1713,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321815976933403e-05,
+      "loss": 4.178,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320977382182351e-05,
+      "loss": 4.1891,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320138787431299e-05,
+      "loss": 4.1864,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319300192680247e-05,
+      "loss": 4.1772,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318463235809568e-05,
+      "loss": 4.1774,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317624641058516e-05,
+      "loss": 4.1746,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316786046307464e-05,
+      "loss": 4.185,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315947451556412e-05,
+      "loss": 4.1906,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315108856805361e-05,
+      "loss": 4.18,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314270262054309e-05,
+      "loss": 4.1883,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313431667303257e-05,
+      "loss": 4.1934,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312593072552205e-05,
+      "loss": 4.1876,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311754477801153e-05,
+      "loss": 4.1702,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3109158830501e-05,
+      "loss": 4.188,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310077288299048e-05,
+      "loss": 4.1867,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30924033142837e-05,
+      "loss": 4.1714,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308401736677318e-05,
+      "loss": 4.1835,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3075647798066386e-05,
+      "loss": 4.1798,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3067261850555866e-05,
+      "loss": 4.1925,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3058875903045346e-05,
+      "loss": 4.1805,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050489955534826e-05,
+      "loss": 4.1607,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042104008024306e-05,
+      "loss": 4.1785,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3033718060513786e-05,
+      "loss": 4.1714,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025332113003266e-05,
+      "loss": 4.1816,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3016946165492746e-05,
+      "loss": 4.1795,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008560217982226e-05,
+      "loss": 4.1717,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000174270471705e-05,
+      "loss": 4.1788,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2991804701764915e-05,
+      "loss": 4.1784,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983418754254395e-05,
+      "loss": 4.1699,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975032806743874e-05,
+      "loss": 4.1764,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2966646859233354e-05,
+      "loss": 4.1805,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2958260911722834e-05,
+      "loss": 4.1882,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2949874964212314e-05,
+      "loss": 4.175,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294150539550553e-05,
+      "loss": 4.1616,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293311944799501e-05,
+      "loss": 4.1736,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292473350048449e-05,
+      "loss": 4.1742,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291634755297397e-05,
+      "loss": 4.1843,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290797798426718e-05,
+      "loss": 4.1837,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289959203675666e-05,
+      "loss": 4.1643,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289120608924614e-05,
+      "loss": 4.1665,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288282014173562e-05,
+      "loss": 4.1805,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28744341942251e-05,
+      "loss": 4.1551,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286606462551831e-05,
+      "loss": 4.1731,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285767867800779e-05,
+      "loss": 4.1598,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284929273049727e-05,
+      "loss": 4.1748,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284090678298675e-05,
+      "loss": 4.1611,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283252083547623e-05,
+      "loss": 4.1747,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282413488796571e-05,
+      "loss": 4.1682,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2815765319258924e-05,
+      "loss": 4.1583,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807379371748404e-05,
+      "loss": 4.1774,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2798993424237884e-05,
+      "loss": 4.1622,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2790607476727364e-05,
+      "loss": 4.1849,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278222152921684e-05,
+      "loss": 4.1779,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277385196051005e-05,
+      "loss": 4.1654,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276546601299953e-05,
+      "loss": 4.1767,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275708006548901e-05,
+      "loss": 4.1612,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2748694117978486e-05,
+      "loss": 4.1697,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2740308170467966e-05,
+      "loss": 4.1671,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273193860176118e-05,
+      "loss": 4.175,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272355265425066e-05,
+      "loss": 4.1666,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271516670674014e-05,
+      "loss": 4.1616,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270678075922962e-05,
+      "loss": 4.1684,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26983948117191e-05,
+      "loss": 4.1671,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269000886420858e-05,
+      "loss": 4.1573,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268163929550179e-05,
+      "loss": 4.1715,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267325334799127e-05,
+      "loss": 4.1561,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266486740048075e-05,
+      "loss": 4.1811,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265648145297023e-05,
+      "loss": 4.1627,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264809550545971e-05,
+      "loss": 4.1719,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263972593675292e-05,
+      "loss": 4.1592,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26313399892424e-05,
+      "loss": 4.1656,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2622954041731886e-05,
+      "loss": 4.174,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2614568094221366e-05,
+      "loss": 4.165,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2606182146710846e-05,
+      "loss": 4.1625,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2597796199200326e-05,
+      "loss": 4.1734,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2589426630493535e-05,
+      "loss": 4.168,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2581040682983015e-05,
+      "loss": 4.1733,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2572654735472495e-05,
+      "loss": 4.1591,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2564268787961975e-05,
+      "loss": 4.1729,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2555882840451455e-05,
+      "loss": 4.1651,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2547496892940935e-05,
+      "loss": 4.1642,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253914370303787e-05,
+      "loss": 4.1654,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253075775552735e-05,
+      "loss": 4.1648,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252237180801684e-05,
+      "loss": 4.1705,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251398586050632e-05,
+      "loss": 4.144,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25055999129958e-05,
+      "loss": 4.1698,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.167238235473633,
+      "eval_runtime": 299.0901,
+      "eval_samples_per_second": 1275.84,
+      "eval_steps_per_second": 39.871,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249721396548528e-05,
+      "loss": 4.1583,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248882801797476e-05,
+      "loss": 4.1501,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248044207046424e-05,
+      "loss": 4.1675,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247205612295372e-05,
+      "loss": 4.1729,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24636701754432e-05,
+      "loss": 4.1636,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245528422793267e-05,
+      "loss": 4.1567,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244689828042215e-05,
+      "loss": 4.1611,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243851233291163e-05,
+      "loss": 4.1507,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243012638540111e-05,
+      "loss": 4.1634,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242174043789059e-05,
+      "loss": 4.1648,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241335449038007e-05,
+      "loss": 4.1639,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240496854286955e-05,
+      "loss": 4.1652,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239659897416277e-05,
+      "loss": 4.1566,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238821302665225e-05,
+      "loss": 4.1577,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237982707914173e-05,
+      "loss": 4.1509,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237144113163121e-05,
+      "loss": 4.145,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236305518412069e-05,
+      "loss": 4.1554,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235466923661017e-05,
+      "loss": 4.1498,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234628328909965e-05,
+      "loss": 4.1573,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233789734158913e-05,
+      "loss": 4.1668,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232951139407861e-05,
+      "loss": 4.1684,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232112544656809e-05,
+      "loss": 4.1627,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231273949905756e-05,
+      "loss": 4.1549,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230436993035078e-05,
+      "loss": 4.1659,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229598398284026e-05,
+      "loss": 4.1489,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228759803532974e-05,
+      "loss": 4.1633,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227921208781922e-05,
+      "loss": 4.1503,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22708261403087e-05,
+      "loss": 4.139,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226244019279818e-05,
+      "loss": 4.1513,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254054245287657e-05,
+      "loss": 4.1498,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245684676580866e-05,
+      "loss": 4.1496,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237298729070346e-05,
+      "loss": 4.1603,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2228912781559826e-05,
+      "loss": 4.1577,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2220526834049306e-05,
+      "loss": 4.1577,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2212157265342515e-05,
+      "loss": 4.1485,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2203771317831995e-05,
+      "loss": 4.1666,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2195385370321475e-05,
+      "loss": 4.1465,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2186999422810955e-05,
+      "loss": 4.1408,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217861347530044e-05,
+      "loss": 4.1469,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217022752778992e-05,
+      "loss": 4.1416,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21618415802794e-05,
+      "loss": 4.1515,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215345563276888e-05,
+      "loss": 4.1594,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214506968525836e-05,
+      "loss": 4.1397,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21367164953553e-05,
+      "loss": 4.1488,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212833054784478e-05,
+      "loss": 4.1564,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211994460033426e-05,
+      "loss": 4.1581,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211155865282374e-05,
+      "loss": 4.1564,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210317270531322e-05,
+      "loss": 4.1432,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20947867578027e-05,
+      "loss": 4.1193,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208640081029218e-05,
+      "loss": 4.1624,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2078031241585395e-05,
+      "loss": 4.1506,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2069645294074875e-05,
+      "loss": 4.1502,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061259346564355e-05,
+      "loss": 4.1437,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2052873399053835e-05,
+      "loss": 4.1376,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2044487451543315e-05,
+      "loss": 4.1336,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036101504032795e-05,
+      "loss": 4.1403,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2027715556522275e-05,
+      "loss": 4.1368,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201932960901175e-05,
+      "loss": 4.1404,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201097641910869e-05,
+      "loss": 4.1537,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200259047159817e-05,
+      "loss": 4.1341,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199420452408765e-05,
+      "loss": 4.1312,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198581857657713e-05,
+      "loss": 4.1484,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197744900787035e-05,
+      "loss": 4.1325,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196906306035983e-05,
+      "loss": 4.1342,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196067711284931e-05,
+      "loss": 4.1404,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195229116533879e-05,
+      "loss": 4.1426,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194390521782827e-05,
+      "loss": 4.1315,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193551927031775e-05,
+      "loss": 4.129,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192713332280722e-05,
+      "loss": 4.1297,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19187473752967e-05,
+      "loss": 4.1445,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191036142778618e-05,
+      "loss": 4.1419,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190197548027566e-05,
+      "loss": 4.139,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189358953276514e-05,
+      "loss": 4.1401,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188520358525462e-05,
+      "loss": 4.15,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18768176377441e-05,
+      "loss": 4.1453,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186844806903731e-05,
+      "loss": 4.129,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18600621215268e-05,
+      "loss": 4.1402,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1851692552820006e-05,
+      "loss": 4.1447,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843306605309486e-05,
+      "loss": 4.1243,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1834920657798966e-05,
+      "loss": 4.1412,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1826534710288446e-05,
+      "loss": 4.139,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818148762777926e-05,
+      "loss": 4.1465,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1809762815267406e-05,
+      "loss": 4.1387,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1801393246560615e-05,
+      "loss": 4.1184,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1793007299050095e-05,
+      "loss": 4.1333,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1784621351539575e-05,
+      "loss": 4.129,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776235404029055e-05,
+      "loss": 4.1427,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1767849456518535e-05,
+      "loss": 4.1329,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1759463509008015e-05,
+      "loss": 4.1291,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175109394030123e-05,
+      "loss": 4.1343,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174270799279071e-05,
+      "loss": 4.1385,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173432204528019e-05,
+      "loss": 4.1243,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172593609776967e-05,
+      "loss": 4.1328,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171755015025915e-05,
+      "loss": 4.1406,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170916420274863e-05,
+      "loss": 4.1436,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170079463404184e-05,
+      "loss": 4.1332,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169240868653132e-05,
+      "loss": 4.1203,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16840227390208e-05,
+      "loss": 4.1299,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167563679151028e-05,
+      "loss": 4.1302,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166725084399975e-05,
+      "loss": 4.1413,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165886489648923e-05,
+      "loss": 4.1434,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165047894897872e-05,
+      "loss": 4.1261,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642109380271935e-05,
+      "loss": 4.124,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163372343276141e-05,
+      "loss": 4.1334,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162533748525089e-05,
+      "loss": 4.1169,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161695153774037e-05,
+      "loss": 4.1322,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160856559022985e-05,
+      "loss": 4.1144,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160017964271933e-05,
+      "loss": 4.1342,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159179369520881e-05,
+      "loss": 4.1211,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158340774769829e-05,
+      "loss": 4.1294,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1575054557795226e-05,
+      "loss": 4.128,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1566668610284706e-05,
+      "loss": 4.1226,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558282662774186e-05,
+      "loss": 4.1317,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154989671526367e-05,
+      "loss": 4.1206,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154151076775315e-05,
+      "loss": 4.144,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153312482024263e-05,
+      "loss": 4.1354,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152473887273211e-05,
+      "loss": 4.1256,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151636930402532e-05,
+      "loss": 4.135,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15079833565148e-05,
+      "loss": 4.1189,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149959740900428e-05,
+      "loss": 4.1292,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149121146149376e-05,
+      "loss": 4.1272,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148282551398324e-05,
+      "loss": 4.1294,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147443956647272e-05,
+      "loss": 4.1277,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146606999776593e-05,
+      "loss": 4.1235,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145768405025541e-05,
+      "loss": 4.1282,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144929810274489e-05,
+      "loss": 4.1248,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144092853403811e-05,
+      "loss": 4.1187,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1432542586527587e-05,
+      "loss": 4.1315,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424156639017067e-05,
+      "loss": 4.1123,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1415770691506546e-05,
+      "loss": 4.1399,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1407384743996026e-05,
+      "loss": 4.1255,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1398998796485506e-05,
+      "loss": 4.1314,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1390612848974986e-05,
+      "loss": 4.1148,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1382226901464466e-05,
+      "loss": 4.1296,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1373840953953946e-05,
+      "loss": 4.1317,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136545500644342e-05,
+      "loss": 4.1284,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13570690589329e-05,
+      "loss": 4.1198,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134868311142238e-05,
+      "loss": 4.1344,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340313542715595e-05,
+      "loss": 4.126,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1331927595205075e-05,
+      "loss": 4.1367,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1323541647694555e-05,
+      "loss": 4.1169,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315155700184035e-05,
+      "loss": 4.1351,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1306786131477244e-05,
+      "loss": 4.1254,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1298400183966724e-05,
+      "loss": 4.1232,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1290014236456204e-05,
+      "loss": 4.1264,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1281628288945684e-05,
+      "loss": 4.1263,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127325872023889e-05,
+      "loss": 4.1284,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126487277272837e-05,
+      "loss": 4.1105,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125651958282532e-05,
+      "loss": 4.1303,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.135540008544922,
+      "eval_runtime": 296.2728,
+      "eval_samples_per_second": 1287.972,
+      "eval_steps_per_second": 40.25,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.12481336353148e-05,
+      "loss": 4.1206,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123974768780428e-05,
+      "loss": 4.1101,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123136174029376e-05,
+      "loss": 4.1271,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1222975792783245e-05,
+      "loss": 4.1317,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.121458984527272e-05,
+      "loss": 4.1237,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.12062038977622e-05,
+      "loss": 4.1192,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119781795025168e-05,
+      "loss": 4.1264,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118943200274116e-05,
+      "loss": 4.112,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118104605523064e-05,
+      "loss": 4.1241,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117266010772012e-05,
+      "loss": 4.1264,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11642741602096e-05,
+      "loss": 4.1231,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115588821269908e-05,
+      "loss": 4.1266,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114751864399229e-05,
+      "loss": 4.1149,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113913269648177e-05,
+      "loss": 4.1181,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1130746748971247e-05,
+      "loss": 4.1156,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1122360801460727e-05,
+      "loss": 4.1065,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111397485395021e-05,
+      "loss": 4.117,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110558890643969e-05,
+      "loss": 4.1127,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109720295892917e-05,
+      "loss": 4.117,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108881701141865e-05,
+      "loss": 4.1254,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108043106390813e-05,
+      "loss": 4.1315,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1072045116397606e-05,
+      "loss": 4.1301,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1063659168887086e-05,
+      "loss": 4.1134,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10552896001803e-05,
+      "loss": 4.1251,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1046903652669775e-05,
+      "loss": 4.1151,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1038517705159255e-05,
+      "loss": 4.1219,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1030131757648735e-05,
+      "loss": 4.1155,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1021745810138215e-05,
+      "loss": 4.1031,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1013359862627695e-05,
+      "loss": 4.1157,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100497391511718e-05,
+      "loss": 4.1065,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099658796760666e-05,
+      "loss": 4.1135,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098821839889987e-05,
+      "loss": 4.1235,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097983245138935e-05,
+      "loss": 4.1188,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097144650387883e-05,
+      "loss": 4.1186,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.096306055636831e-05,
+      "loss": 4.109,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.095469098766152e-05,
+      "loss": 4.1307,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0946305040151e-05,
+      "loss": 4.1139,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093791909264048e-05,
+      "loss": 4.1029,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092953314512996e-05,
+      "loss": 4.1115,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092114719761944e-05,
+      "loss": 4.1041,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.091276125010892e-05,
+      "loss": 4.1126,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.09043753025984e-05,
+      "loss": 4.1165,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.089598935508788e-05,
+      "loss": 4.1105,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0887619786381095e-05,
+      "loss": 4.1088,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0879233838870575e-05,
+      "loss": 4.1221,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0870864270163784e-05,
+      "loss": 4.1185,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0862478322653264e-05,
+      "loss": 4.1225,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0854092375142744e-05,
+      "loss": 4.1048,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0845706427632224e-05,
+      "loss": 4.0816,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0837336858925433e-05,
+      "loss": 4.1292,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0828950911414913e-05,
+      "loss": 4.111,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082056496390439e-05,
+      "loss": 4.118,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081217901639387e-05,
+      "loss": 4.1082,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080379306888335e-05,
+      "loss": 4.0983,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079540712137283e-05,
+      "loss": 4.0986,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078702117386231e-05,
+      "loss": 4.1043,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077863522635179e-05,
+      "loss": 4.0977,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077026565764501e-05,
+      "loss": 4.1022,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076189608893822e-05,
+      "loss": 4.1214,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.07535101414277e-05,
+      "loss": 4.0966,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.074512419391718e-05,
+      "loss": 4.0936,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.073673824640666e-05,
+      "loss": 4.1154,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072835229889614e-05,
+      "loss": 4.0919,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071996635138561e-05,
+      "loss": 4.0971,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071159678267883e-05,
+      "loss": 4.1057,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.070321083516831e-05,
+      "loss": 4.1069,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069482488765779e-05,
+      "loss": 4.0943,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068643894014727e-05,
+      "loss": 4.1012,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067805299263675e-05,
+      "loss": 4.0887,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066966704512623e-05,
+      "loss": 4.1083,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066128109761571e-05,
+      "loss": 4.1076,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065289515010519e-05,
+      "loss": 4.1034,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064450920259467e-05,
+      "loss": 4.1054,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0636139633887876e-05,
+      "loss": 4.11,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0627753686377356e-05,
+      "loss": 4.1147,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0619367738866836e-05,
+      "loss": 4.0942,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0610998170160045e-05,
+      "loss": 4.1048,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0602612222649525e-05,
+      "loss": 4.1102,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0594226275139005e-05,
+      "loss": 4.0878,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.058584032762849e-05,
+      "loss": 4.1087,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.057745438011797e-05,
+      "loss": 4.1003,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056906843260745e-05,
+      "loss": 4.1126,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056069886390066e-05,
+      "loss": 4.1069,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055231291639014e-05,
+      "loss": 4.087,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054392696887962e-05,
+      "loss": 4.0899,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05355410213691e-05,
+      "loss": 4.0972,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.052717145266231e-05,
+      "loss": 4.1098,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051878550515179e-05,
+      "loss": 4.0956,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051039955764127e-05,
+      "loss": 4.0953,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.050202998893448e-05,
+      "loss": 4.0994,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049364404142396e-05,
+      "loss": 4.1026,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0485258093913445e-05,
+      "loss": 4.0927,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0476872146402925e-05,
+      "loss": 4.0963,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0468486198892405e-05,
+      "loss": 4.1055,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0460100251381885e-05,
+      "loss": 4.108,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0451714303871365e-05,
+      "loss": 4.1004,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0443328356360845e-05,
+      "loss": 4.0875,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0434942408850325e-05,
+      "loss": 4.0993,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04265564613398e-05,
+      "loss": 4.093,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0418186892633014e-05,
+      "loss": 4.1081,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0409800945122494e-05,
+      "loss": 4.1034,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0401414997611974e-05,
+      "loss": 4.0964,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039302905010145e-05,
+      "loss": 4.0943,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038464310259093e-05,
+      "loss": 4.0938,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0376257155080414e-05,
+      "loss": 4.0871,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0367871207569894e-05,
+      "loss": 4.0964,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0359485260059374e-05,
+      "loss": 4.0822,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0351099312548853e-05,
+      "loss": 4.1021,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.03427461226458e-05,
+      "loss": 4.0901,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033436017513527e-05,
+      "loss": 4.0931,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032597422762475e-05,
+      "loss": 4.0916,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031758828011423e-05,
+      "loss": 4.0887,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030920233260371e-05,
+      "loss": 4.1008,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030081638509319e-05,
+      "loss": 4.0889,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.029243043758267e-05,
+      "loss": 4.1114,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028404449007215e-05,
+      "loss": 4.1029,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027567492136537e-05,
+      "loss": 4.0925,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0267305352658576e-05,
+      "loss": 4.0986,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0258919405148056e-05,
+      "loss": 4.0883,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0250533457637536e-05,
+      "loss": 4.0947,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0242147510127016e-05,
+      "loss": 4.1,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0233761562616496e-05,
+      "loss": 4.0946,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0225391993909705e-05,
+      "loss": 4.0945,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0217006046399185e-05,
+      "loss": 4.0928,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0208620098888665e-05,
+      "loss": 4.0957,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0200234151378145e-05,
+      "loss": 4.0944,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0191848203867625e-05,
+      "loss": 4.083,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0183478635160834e-05,
+      "loss": 4.0973,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0175092687650314e-05,
+      "loss": 4.0848,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01667067401398e-05,
+      "loss": 4.1054,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015833717143301e-05,
+      "loss": 4.0927,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014995122392249e-05,
+      "loss": 4.0958,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014156527641197e-05,
+      "loss": 4.0835,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013317932890145e-05,
+      "loss": 4.0966,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.012479338139093e-05,
+      "loss": 4.0983,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011640743388041e-05,
+      "loss": 4.0946,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010802148636989e-05,
+      "loss": 4.0901,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009963553885937e-05,
+      "loss": 4.0998,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009124959134885e-05,
+      "loss": 4.0931,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008286364383833e-05,
+      "loss": 4.1085,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007449407513154e-05,
+      "loss": 4.0846,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006610812762102e-05,
+      "loss": 4.1023,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00577221801105e-05,
+      "loss": 4.0927,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0049336232599985e-05,
+      "loss": 4.0912,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0040950285089465e-05,
+      "loss": 4.0925,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.003256433757894e-05,
+      "loss": 4.093,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0024194768872154e-05,
+      "loss": 4.0957,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0015808821361634e-05,
+      "loss": 4.0795,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000742287385111e-05,
+      "loss": 4.0968,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.111083984375,
+      "eval_runtime": 297.8115,
+      "eval_samples_per_second": 1281.317,
+      "eval_steps_per_second": 40.042,
+      "step": 610560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999903692634059e-05,
+      "loss": 4.084,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999065097883007e-05,
+      "loss": 4.0795,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.998226503131955e-05,
+      "loss": 4.0928,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.997387908380903e-05,
+      "loss": 4.103,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9965509515102236e-05,
+      "loss": 4.0929,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.995712356759172e-05,
+      "loss": 4.0862,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.99487376200812e-05,
+      "loss": 4.0994,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994035167257068e-05,
+      "loss": 4.0776,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.993196572506016e-05,
+      "loss": 4.0895,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.992357977754964e-05,
+      "loss": 4.0988,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.991519383003912e-05,
+      "loss": 4.0917,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.990682426133233e-05,
+      "loss": 4.0938,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989843831382181e-05,
+      "loss": 4.0911,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989005236631129e-05,
+      "loss": 4.0823,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.988166641880077e-05,
+      "loss": 4.08,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.987329685009398e-05,
+      "loss": 4.0772,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.986491090258346e-05,
+      "loss": 4.0845,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.985652495507294e-05,
+      "loss": 4.0847,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984813900756242e-05,
+      "loss": 4.0828,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983975306005191e-05,
+      "loss": 4.0927,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983136711254139e-05,
+      "loss": 4.1026,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.982298116503087e-05,
+      "loss": 4.0971,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.981459521752035e-05,
+      "loss": 4.0868,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9806242027617286e-05,
+      "loss": 4.0901,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9797856080106766e-05,
+      "loss": 4.091,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9789470132596246e-05,
+      "loss": 4.0863,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9781084185085726e-05,
+      "loss": 4.086,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9772698237575205e-05,
+      "loss": 4.0752,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9764312290064685e-05,
+      "loss": 4.0834,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9755926342554165e-05,
+      "loss": 4.0786,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9747540395043645e-05,
+      "loss": 4.08,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973917082633686e-05,
+      "loss": 4.0918,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973078487882634e-05,
+      "loss": 4.091,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.972239893131582e-05,
+      "loss": 4.0872,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9714012983805294e-05,
+      "loss": 4.0834,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9705627036294774e-05,
+      "loss": 4.0927,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9697241088784254e-05,
+      "loss": 4.0845,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968887152007747e-05,
+      "loss": 4.0758,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968048557256694e-05,
+      "loss": 4.0822,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.967209962505642e-05,
+      "loss": 4.0731,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96637136775459e-05,
+      "loss": 4.079,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.965534410883912e-05,
+      "loss": 4.0884,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96469581613286e-05,
+      "loss": 4.0774,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963857221381808e-05,
+      "loss": 4.0834,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963018626630756e-05,
+      "loss": 4.0928,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.962180031879704e-05,
+      "loss": 4.0858,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.961341437128652e-05,
+      "loss": 4.0904,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9605028423776e-05,
+      "loss": 4.0796,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.959664247626548e-05,
+      "loss": 4.0535,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958825652875496e-05,
+      "loss": 4.096,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957988696004817e-05,
+      "loss": 4.0806,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957150101253765e-05,
+      "loss": 4.0858,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.956311506502713e-05,
+      "loss": 4.0814,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.955472911751661e-05,
+      "loss": 4.0698,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.954635954880982e-05,
+      "loss": 4.0667,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.95379736012993e-05,
+      "loss": 4.0762,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952958765378878e-05,
+      "loss": 4.0701,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9521201706278263e-05,
+      "loss": 4.07,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.95128485163752e-05,
+      "loss": 4.0924,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.950446256886468e-05,
+      "loss": 4.0702,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.949607662135416e-05,
+      "loss": 4.0643,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948769067384364e-05,
+      "loss": 4.0889,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947930472633312e-05,
+      "loss": 4.058,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.94709187788226e-05,
+      "loss": 4.0658,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.946253283131208e-05,
+      "loss": 4.0791,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.945414688380156e-05,
+      "loss": 4.0783,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.944577731509477e-05,
+      "loss": 4.0663,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9437407746387986e-05,
+      "loss": 4.0691,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9429038177681195e-05,
+      "loss": 4.0639,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9420652230170675e-05,
+      "loss": 4.0772,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9412266282660155e-05,
+      "loss": 4.0761,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9403880335149635e-05,
+      "loss": 4.0739,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9395494387639115e-05,
+      "loss": 4.0768,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9387108440128595e-05,
+      "loss": 4.0801,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9378722492618075e-05,
+      "loss": 4.0861,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9370336545107555e-05,
+      "loss": 4.0644,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9361950597597035e-05,
+      "loss": 4.0784,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9353564650086515e-05,
+      "loss": 4.0801,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9345178702575995e-05,
+      "loss": 4.0546,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9336792755065475e-05,
+      "loss": 4.0782,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9328423186358684e-05,
+      "loss": 4.076,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932003723884817e-05,
+      "loss": 4.0847,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.931166767014138e-05,
+      "loss": 4.0783,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.930328172263086e-05,
+      "loss": 4.0615,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.929489577512034e-05,
+      "loss": 4.0576,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.928650982760982e-05,
+      "loss": 4.0713,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.92781238800993e-05,
+      "loss": 4.08,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926975431139251e-05,
+      "loss": 4.0703,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926136836388199e-05,
+      "loss": 4.0658,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.925298241637147e-05,
+      "loss": 4.0694,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.924459646886095e-05,
+      "loss": 4.0778,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.923621052135042e-05,
+      "loss": 4.0655,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.922782457383991e-05,
+      "loss": 4.0687,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921943862632939e-05,
+      "loss": 4.0735,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921105267881887e-05,
+      "loss": 4.0797,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.920268311011208e-05,
+      "loss": 4.0711,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.919429716260156e-05,
+      "loss": 4.062,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.918591121509104e-05,
+      "loss": 4.0647,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.917754164638425e-05,
+      "loss": 4.0666,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9169155698873727e-05,
+      "loss": 4.0792,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9160769751363206e-05,
+      "loss": 4.0775,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9152383803852686e-05,
+      "loss": 4.0692,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9143997856342166e-05,
+      "loss": 4.0656,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9135611908831646e-05,
+      "loss": 4.061,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9127225961321126e-05,
+      "loss": 4.0659,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911885639261434e-05,
+      "loss": 4.0663,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911047044510382e-05,
+      "loss": 4.0537,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.91020844975933e-05,
+      "loss": 4.0753,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.909369855008278e-05,
+      "loss": 4.0616,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.908531260257226e-05,
+      "loss": 4.0636,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.907694303386547e-05,
+      "loss": 4.0623,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906855708635495e-05,
+      "loss": 4.0582,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906017113884443e-05,
+      "loss": 4.0748,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.905178519133391e-05,
+      "loss": 4.0591,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.904339924382339e-05,
+      "loss": 4.0847,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.903501329631287e-05,
+      "loss": 4.0748,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.902662734880235e-05,
+      "loss": 4.0673,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.901825778009556e-05,
+      "loss": 4.071,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.900987183258504e-05,
+      "loss": 4.0641,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9001485885074527e-05,
+      "loss": 4.0662,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8993099937564007e-05,
+      "loss": 4.0729,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8984713990053487e-05,
+      "loss": 4.0644,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8976344421346696e-05,
+      "loss": 4.0704,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8967958473836176e-05,
+      "loss": 4.0602,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8959572526325656e-05,
+      "loss": 4.068,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8951186578815135e-05,
+      "loss": 4.0665,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8942800631304615e-05,
+      "loss": 4.0552,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.893441468379409e-05,
+      "loss": 4.0711,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8926045115087304e-05,
+      "loss": 4.0577,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8917659167576784e-05,
+      "loss": 4.0791,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8909273220066264e-05,
+      "loss": 4.0658,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8900887272555744e-05,
+      "loss": 4.0689,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8892501325045224e-05,
+      "loss": 4.0553,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8884115377534704e-05,
+      "loss": 4.0728,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8875729430024184e-05,
+      "loss": 4.0628,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8867343482513664e-05,
+      "loss": 4.0685,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885897391380687e-05,
+      "loss": 4.0644,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885058796629635e-05,
+      "loss": 4.0719,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.884220201878583e-05,
+      "loss": 4.0658,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.883381607127531e-05,
+      "loss": 4.0823,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.882543012376479e-05,
+      "loss": 4.0557,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8817060555058e-05,
+      "loss": 4.0742,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880867460754748e-05,
+      "loss": 4.0677,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.88003050388407e-05,
+      "loss": 4.0683,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.879191909133018e-05,
+      "loss": 4.0594,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.878353314381966e-05,
+      "loss": 4.0686,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.877514719630914e-05,
+      "loss": 4.0655,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.876676124879862e-05,
+      "loss": 4.0597,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.87583753012881e-05,
+      "loss": 4.0683,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.091145992279053,
+      "eval_runtime": 296.5858,
+      "eval_samples_per_second": 1286.612,
+      "eval_steps_per_second": 40.208,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.874998935377758e-05,
+      "loss": 4.0588,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.874160340626706e-05,
+      "loss": 4.0519,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.873321745875654e-05,
+      "loss": 4.0654,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.872483151124602e-05,
+      "loss": 4.0749,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.87164455637355e-05,
+      "loss": 4.0704,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870805961622497e-05,
+      "loss": 4.0567,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869967366871445e-05,
+      "loss": 4.0735,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869128772120393e-05,
+      "loss": 4.0532,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.868290177369342e-05,
+      "loss": 4.0624,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.86745158261829e-05,
+      "loss": 4.0739,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.866612987867238e-05,
+      "loss": 4.0631,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865774393116186e-05,
+      "loss": 4.0678,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8649374362455066e-05,
+      "loss": 4.0673,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8640988414944546e-05,
+      "loss": 4.0565,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8632602467434026e-05,
+      "loss": 4.0534,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8624216519923506e-05,
+      "loss": 4.0525,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8615830572412986e-05,
+      "loss": 4.0558,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8607444624902466e-05,
+      "loss": 4.0596,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8599058677391946e-05,
+      "loss": 4.0539,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8590689108685155e-05,
+      "loss": 4.0663,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8582303161174635e-05,
+      "loss": 4.073,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8573917213664115e-05,
+      "loss": 4.0727,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8565531266153595e-05,
+      "loss": 4.062,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.855714531864308e-05,
+      "loss": 4.0592,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854875937113256e-05,
+      "loss": 4.067,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854037342362204e-05,
+      "loss": 4.0583,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8531987476111515e-05,
+      "loss": 4.0562,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.852361790740473e-05,
+      "loss": 4.0543,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.851523195989421e-05,
+      "loss": 4.055,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.850684601238369e-05,
+      "loss": 4.0527,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8498460064873164e-05,
+      "loss": 4.0545,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8490074117362644e-05,
+      "loss": 4.0642,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8481688169852124e-05,
+      "loss": 4.0693,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8473302222341604e-05,
+      "loss": 4.0602,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8464916274831084e-05,
+      "loss": 4.061,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8456530327320564e-05,
+      "loss": 4.0577,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.844816075861378e-05,
+      "loss": 4.0653,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843977481110326e-05,
+      "loss": 4.0488,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843138886359274e-05,
+      "loss": 4.0566,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842300291608222e-05,
+      "loss": 4.0454,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84146169685717e-05,
+      "loss": 4.057,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.840623102106118e-05,
+      "loss": 4.0583,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839784507355066e-05,
+      "loss": 4.0532,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838945912604014e-05,
+      "loss": 4.058,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838108955733335e-05,
+      "loss": 4.064,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837270360982283e-05,
+      "loss": 4.0599,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836431766231231e-05,
+      "loss": 4.0645,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835593171480179e-05,
+      "loss": 4.0572,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834754576729127e-05,
+      "loss": 4.0266,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833915981978075e-05,
+      "loss": 4.0685,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8330790251073964e-05,
+      "loss": 4.0555,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8322404303563444e-05,
+      "loss": 4.0629,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8314018356052924e-05,
+      "loss": 4.0563,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8305632408542404e-05,
+      "loss": 4.0449,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.829729559744307e-05,
+      "loss": 4.0413,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828890964993255e-05,
+      "loss": 4.0476,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828052370242203e-05,
+      "loss": 4.0484,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.827213775491151e-05,
+      "loss": 4.0416,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.826375180740099e-05,
+      "loss": 4.0663,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.825536585989047e-05,
+      "loss": 4.0483,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.824697991237996e-05,
+      "loss": 4.034,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823859396486944e-05,
+      "loss": 4.0664,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823020801735892e-05,
+      "loss": 4.0366,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82218220698484e-05,
+      "loss": 4.0395,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821343612233788e-05,
+      "loss": 4.0551,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.820505017482735e-05,
+      "loss": 4.0496,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8196680606120567e-05,
+      "loss": 4.0445,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8188311037413776e-05,
+      "loss": 4.0442,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8179925089903256e-05,
+      "loss": 4.0392,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8171539142392736e-05,
+      "loss": 4.0515,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8163153194882216e-05,
+      "loss": 4.0516,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8154767247371695e-05,
+      "loss": 4.0522,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8146381299861175e-05,
+      "loss": 4.05,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8137995352350655e-05,
+      "loss": 4.0539,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8129609404840135e-05,
+      "loss": 4.0641,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812123983613335e-05,
+      "loss": 4.0376,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8112853888622824e-05,
+      "loss": 4.05,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810448431991604e-05,
+      "loss": 4.0589,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.809609837240552e-05,
+      "loss": 4.0332,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8087712424894993e-05,
+      "loss": 4.0482,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.807932647738447e-05,
+      "loss": 4.0552,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.807095690867769e-05,
+      "loss": 4.0572,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.806257096116717e-05,
+      "loss": 4.0584,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.805418501365665e-05,
+      "loss": 4.0343,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.804579906614613e-05,
+      "loss": 4.032,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.803741311863561e-05,
+      "loss": 4.0487,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802902717112509e-05,
+      "loss": 4.0529,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802064122361457e-05,
+      "loss": 4.0416,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.801227165490778e-05,
+      "loss": 4.046,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.800388570739726e-05,
+      "loss": 4.0429,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799549975988674e-05,
+      "loss": 4.0523,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.798711381237622e-05,
+      "loss": 4.0456,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79787278648657e-05,
+      "loss": 4.0431,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797034191735518e-05,
+      "loss": 4.0458,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796195596984466e-05,
+      "loss": 4.0551,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.795357002233414e-05,
+      "loss": 4.05,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794518407482362e-05,
+      "loss": 4.039,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793681450611683e-05,
+      "loss": 4.0388,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7928428558606314e-05,
+      "loss": 4.0434,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7920042611095793e-05,
+      "loss": 4.0532,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7911656663585273e-05,
+      "loss": 4.0511,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7903270716074753e-05,
+      "loss": 4.0427,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789490114736796e-05,
+      "loss": 4.0446,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.788651519985744e-05,
+      "loss": 4.0355,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787812925234692e-05,
+      "loss": 4.0433,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78697433048364e-05,
+      "loss": 4.0372,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786139011493334e-05,
+      "loss": 4.0327,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.785300416742282e-05,
+      "loss": 4.0502,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78446182199123e-05,
+      "loss": 4.037,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.783623227240178e-05,
+      "loss": 4.0389,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782784632489127e-05,
+      "loss": 4.0381,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781946037738075e-05,
+      "loss": 4.0349,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7811090808673956e-05,
+      "loss": 4.0493,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7802704861163436e-05,
+      "loss": 4.0379,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7794318913652916e-05,
+      "loss": 4.059,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7785932966142396e-05,
+      "loss": 4.0512,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7777547018631876e-05,
+      "loss": 4.0443,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7769177449925085e-05,
+      "loss": 4.0464,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7760791502414565e-05,
+      "loss": 4.0424,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7752405554904045e-05,
+      "loss": 4.0411,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7744019607393525e-05,
+      "loss": 4.052,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7735633659883005e-05,
+      "loss": 4.046,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7727247712372485e-05,
+      "loss": 4.0431,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7718861764861965e-05,
+      "loss": 4.0358,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7710475817351445e-05,
+      "loss": 4.0449,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7702089869840925e-05,
+      "loss": 4.0407,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7693720301134134e-05,
+      "loss": 4.0335,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768535073242735e-05,
+      "loss": 4.0483,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.767696478491683e-05,
+      "loss": 4.0264,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76685788374063e-05,
+      "loss": 4.0597,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766019288989578e-05,
+      "loss": 4.0393,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7651823321189e-05,
+      "loss": 4.0458,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.764343737367848e-05,
+      "loss": 4.0329,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.763505142616796e-05,
+      "loss": 4.0502,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.762666547865744e-05,
+      "loss": 4.0411,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761827953114692e-05,
+      "loss": 4.0456,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76098935836364e-05,
+      "loss": 4.0428,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760150763612588e-05,
+      "loss": 4.0444,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.759313806741909e-05,
+      "loss": 4.0436,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758475211990857e-05,
+      "loss": 4.061,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757636617239805e-05,
+      "loss": 4.0313,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756798022488753e-05,
+      "loss": 4.0517,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755959427737701e-05,
+      "loss": 4.0474,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755120832986649e-05,
+      "loss": 4.0444,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754282238235597e-05,
+      "loss": 4.0347,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753443643484545e-05,
+      "loss": 4.0475,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7526066866138656e-05,
+      "loss": 4.0399,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751769729743187e-05,
+      "loss": 4.0383,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.750931134992135e-05,
+      "loss": 4.0422,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.075286388397217,
+      "eval_runtime": 297.3176,
+      "eval_samples_per_second": 1283.446,
+      "eval_steps_per_second": 40.109,
+      "step": 763200
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 3.153557465701273e+17,
+  "trial_name": null,
+  "trial_params": null
+}