diff --git "a/checkpoint-1908000/trainer_state.json" "b/checkpoint-1908000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1908000/trainer_state.json"
@@ -0,0 +1,22581 @@
+{
+  "best_metric": 4.000287055969238,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/superlative-quantifier/lstm/2/checkpoints/checkpoint-1908000",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 1908000,
+  "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": 10.8196,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.563,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0597,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9935,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9463,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8805,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.6971,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.5925,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5066,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4414,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.3894,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3248,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2579,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.1934,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.1367,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0883,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0428,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 6.0064,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9585,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9316,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8963,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8503,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8204,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7864,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7648,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7323,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7024,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6752,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6675,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6271,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6145,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.6054,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5755,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5571,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5518,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 5.5304,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.5192,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4793,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4732,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4448,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 5.4417,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 5.4283,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 5.4231,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.3913,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.3909,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.3769,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.377,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3692,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3412,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.326,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3344,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3277,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.3023,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.2935,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 5.289,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 5.2721,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 5.2796,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 5.2439,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.249,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2383,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2366,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.2118,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2189,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.1924,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.1975,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.1963,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 5.1898,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 5.1693,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.1526,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1448,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1597,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 5.1499,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.134,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1439,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1404,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 5.126,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362880913648985e-05,
+      "loss": 5.1139,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 5.1033,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.1011,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.0854,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 5.0873,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 5.0756,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 5.0885,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 5.0735,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.0516,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 5.0544,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 5.0528,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 5.0511,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 5.0449,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925391273242342e-05,
+      "loss": 5.0279,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92455267849129e-05,
+      "loss": 5.027,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923714083740238e-05,
+      "loss": 5.0211,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922875488989185e-05,
+      "loss": 5.0099,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 4.9991,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921199937367455e-05,
+      "loss": 4.9975,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920361342616403e-05,
+      "loss": 5.0054,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919522747865351e-05,
+      "loss": 4.994,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.982,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.9726,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.9762,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.9786,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.9713,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.9635,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136542224883596e-05,
+      "loss": 4.9532,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128156277373076e-05,
+      "loss": 4.9619,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119770329862556e-05,
+      "loss": 4.9396,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.9427,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103014813645245e-05,
+      "loss": 4.9353,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.9284,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9236,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9283,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90694874024069e-05,
+      "loss": 4.9156,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.9141,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905271550738586e-05,
+      "loss": 4.9097,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904432955987534e-05,
+      "loss": 4.9063,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.9088,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.9032,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901920447495124e-05,
+      "loss": 4.8992,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901081852744072e-05,
+      "loss": 4.8952,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.8888,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.889,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.8765,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.8721,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968888789888125e-05,
+      "loss": 4.8707,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960502842377605e-05,
+      "loss": 4.8696,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.8635,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.8619,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.8591,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.8463,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8567,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.8547,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.8353,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.8244,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8306,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.8383,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886830655617308e-05,
+      "loss": 4.8448,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.8271,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.8268,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884316509244525e-05,
+      "loss": 4.8248,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883477914493473e-05,
+      "loss": 4.8259,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.8114,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.805,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.8124,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.8018,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.7997,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878447983867534e-05,
+      "loss": 4.8037,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.7876,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.7984,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.7939,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7739,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.759090900421143,
+      "eval_runtime": 298.7399,
+      "eval_samples_per_second": 1277.335,
+      "eval_steps_per_second": 39.918,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8742566479926466e-05,
+      "loss": 4.7881,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.777,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8725794584905426e-05,
+      "loss": 4.7879,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.775,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709039068688115e-05,
+      "loss": 4.776,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870066949998133e-05,
+      "loss": 4.7596,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869228355247081e-05,
+      "loss": 4.7601,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.868389760496029e-05,
+      "loss": 4.7426,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867551165744977e-05,
+      "loss": 4.7688,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866712570993925e-05,
+      "loss": 4.7615,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865873976242873e-05,
+      "loss": 4.7556,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865035381491821e-05,
+      "loss": 4.7612,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864196786740769e-05,
+      "loss": 4.7429,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863358191989717e-05,
+      "loss": 4.7418,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862519597238665e-05,
+      "loss": 4.7365,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.7354,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860844045616934e-05,
+      "loss": 4.7368,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860005450865882e-05,
+      "loss": 4.7278,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85916685611483e-05,
+      "loss": 4.7375,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858328261363778e-05,
+      "loss": 4.7396,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857489666612726e-05,
+      "loss": 4.727,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856651071861674e-05,
+      "loss": 4.7186,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.7189,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549771581203164e-05,
+      "loss": 4.7227,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.7149,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532999686182124e-05,
+      "loss": 4.7017,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524613738671604e-05,
+      "loss": 4.701,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8516227791161084e-05,
+      "loss": 4.7174,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.6892,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.6953,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.704,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.7002,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847431443241221e-05,
+      "loss": 4.7,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846592848490169e-05,
+      "loss": 4.6983,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.696,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.6968,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844078702117387e-05,
+      "loss": 4.6721,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.6823,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842403150495656e-05,
+      "loss": 4.6675,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841564555744604e-05,
+      "loss": 4.6733,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840725960993552e-05,
+      "loss": 4.6735,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398873662425e-05,
+      "loss": 4.6737,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839048771491447e-05,
+      "loss": 4.6683,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838210176740395e-05,
+      "loss": 4.6665,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837371581989343e-05,
+      "loss": 4.6668,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365346251186646e-05,
+      "loss": 4.6707,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8356960303676126e-05,
+      "loss": 4.6726,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8348574356165606e-05,
+      "loss": 4.6558,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8340188408655086e-05,
+      "loss": 4.646,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331818839948295e-05,
+      "loss": 4.6613,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323432892437775e-05,
+      "loss": 4.6662,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315046944927255e-05,
+      "loss": 4.652,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306660997416735e-05,
+      "loss": 4.6473,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298275049906215e-05,
+      "loss": 4.6463,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289889102395695e-05,
+      "loss": 4.6392,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281503154885175e-05,
+      "loss": 4.6493,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273133586178384e-05,
+      "loss": 4.6281,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826474763866787e-05,
+      "loss": 4.6407,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825636169115735e-05,
+      "loss": 4.6357,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824797574364683e-05,
+      "loss": 4.6374,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823960617494004e-05,
+      "loss": 4.619,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823122022742952e-05,
+      "loss": 4.6404,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222834279919e-05,
+      "loss": 4.6128,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821444833240848e-05,
+      "loss": 4.6295,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820607876370169e-05,
+      "loss": 4.63,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819769281619117e-05,
+      "loss": 4.628,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818930686868065e-05,
+      "loss": 4.613,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818092092117013e-05,
+      "loss": 4.6089,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817253497365961e-05,
+      "loss": 4.6071,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816414902614909e-05,
+      "loss": 4.617,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815576307863857e-05,
+      "loss": 4.6193,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147393509931785e-05,
+      "loss": 4.612,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8139007562421265e-05,
+      "loss": 4.6198,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8130621614910745e-05,
+      "loss": 4.6233,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8122235667400224e-05,
+      "loss": 4.6189,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113866098693434e-05,
+      "loss": 4.6019,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105480151182914e-05,
+      "loss": 4.6076,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8097094203672393e-05,
+      "loss": 4.6086,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8088708256161873e-05,
+      "loss": 4.5864,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808032230865135e-05,
+      "loss": 4.6072,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8071936361140827e-05,
+      "loss": 4.5897,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063550413630307e-05,
+      "loss": 4.6114,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055164466119786e-05,
+      "loss": 4.6003,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8046794897413e-05,
+      "loss": 4.578,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803842532870622e-05,
+      "loss": 4.5957,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80300393811957e-05,
+      "loss": 4.5847,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802165343368518e-05,
+      "loss": 4.5978,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801326748617466e-05,
+      "loss": 4.5907,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800488153866413e-05,
+      "loss": 4.5865,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799649559115361e-05,
+      "loss": 4.5837,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798810964364309e-05,
+      "loss": 4.5799,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797972369613257e-05,
+      "loss": 4.5733,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797135412742578e-05,
+      "loss": 4.568,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796296817991526e-05,
+      "loss": 4.5704,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795458223240474e-05,
+      "loss": 4.5805,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794619628489423e-05,
+      "loss": 4.5724,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937826716187436e-05,
+      "loss": 4.565,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929440768676916e-05,
+      "loss": 4.5604,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7921054821166396e-05,
+      "loss": 4.569,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912668873655876e-05,
+      "loss": 4.5742,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7904282926145356e-05,
+      "loss": 4.5665,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895896978634836e-05,
+      "loss": 4.5619,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887511031124316e-05,
+      "loss": 4.5601,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879125083613796e-05,
+      "loss": 4.5661,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870755514907005e-05,
+      "loss": 4.5468,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862369567396485e-05,
+      "loss": 4.5565,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853983619885965e-05,
+      "loss": 4.555,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845597672375445e-05,
+      "loss": 4.5481,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837211724864925e-05,
+      "loss": 4.5496,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782882577735441e-05,
+      "loss": 4.5515,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782043982984389e-05,
+      "loss": 4.5455,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781205388233337e-05,
+      "loss": 4.5435,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780370069243031e-05,
+      "loss": 4.545,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779531474491979e-05,
+      "loss": 4.5415,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778692879740927e-05,
+      "loss": 4.5529,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777854284989875e-05,
+      "loss": 4.5478,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777017328119196e-05,
+      "loss": 4.5448,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776178733368144e-05,
+      "loss": 4.5425,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775340138617092e-05,
+      "loss": 4.544,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77450154386604e-05,
+      "loss": 4.5397,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773662949114988e-05,
+      "loss": 4.5332,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772824354363936e-05,
+      "loss": 4.5305,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719857596128845e-05,
+      "loss": 4.5348,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771147164861832e-05,
+      "loss": 4.5336,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703102079911534e-05,
+      "loss": 4.5314,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694716132401014e-05,
+      "loss": 4.5325,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686330184890494e-05,
+      "loss": 4.5252,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76779606161837e-05,
+      "loss": 4.5276,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766957466867318e-05,
+      "loss": 4.5277,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766118872116266e-05,
+      "loss": 4.537,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7652802773652136e-05,
+      "loss": 4.5168,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7644416826141616e-05,
+      "loss": 4.5072,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763604725743483e-05,
+      "loss": 4.5142,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762766130992431e-05,
+      "loss": 4.5263,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761927536241379e-05,
+      "loss": 4.5303,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761088941490327e-05,
+      "loss": 4.5198,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760251984619649e-05,
+      "loss": 4.519,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759413389868596e-05,
+      "loss": 4.5219,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758574795117544e-05,
+      "loss": 4.5201,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757736200366492e-05,
+      "loss": 4.5137,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75689760561544e-05,
+      "loss": 4.5107,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756059010864388e-05,
+      "loss": 4.5147,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755220416113336e-05,
+      "loss": 4.5111,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754381821362284e-05,
+      "loss": 4.5023,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753544864491605e-05,
+      "loss": 4.5162,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527062697405536e-05,
+      "loss": 4.4989,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518676749895016e-05,
+      "loss": 4.5099,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510290802384496e-05,
+      "loss": 4.5068,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7501921233677705e-05,
+      "loss": 4.4948,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.481330394744873,
+      "eval_runtime": 298.0471,
+      "eval_samples_per_second": 1280.304,
+      "eval_steps_per_second": 40.01,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493535286167185e-05,
+      "loss": 4.506,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485149338656665e-05,
+      "loss": 4.4964,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476763391146145e-05,
+      "loss": 4.5061,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468377443635625e-05,
+      "loss": 4.4983,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7459991496125105e-05,
+      "loss": 4.5036,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451605548614585e-05,
+      "loss": 4.489,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443219601104065e-05,
+      "loss": 4.4876,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434833653593545e-05,
+      "loss": 4.4743,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426447706083025e-05,
+      "loss": 4.5051,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418061758572505e-05,
+      "loss": 4.4956,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409675811061985e-05,
+      "loss": 4.4876,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401289863551465e-05,
+      "loss": 4.5027,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739292029484468e-05,
+      "loss": 4.4791,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384534347334154e-05,
+      "loss": 4.4868,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376148399823634e-05,
+      "loss": 4.4795,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7367762452313114e-05,
+      "loss": 4.4765,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359376504802594e-05,
+      "loss": 4.4854,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350990557292074e-05,
+      "loss": 4.4737,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342604609781554e-05,
+      "loss": 4.4851,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7334218662271034e-05,
+      "loss": 4.4885,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7325832714760514e-05,
+      "loss": 4.4783,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7317446767249994e-05,
+      "loss": 4.4745,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7309060819739474e-05,
+      "loss": 4.4744,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730069125103269e-05,
+      "loss": 4.4782,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729230530352217e-05,
+      "loss": 4.4669,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728391935601165e-05,
+      "loss": 4.4674,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727553340850113e-05,
+      "loss": 4.4579,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726714746099061e-05,
+      "loss": 4.4788,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725877789228382e-05,
+      "loss": 4.4497,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725040832357703e-05,
+      "loss": 4.4626,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724202237606651e-05,
+      "loss": 4.4675,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723363642855599e-05,
+      "loss": 4.469,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722525048104547e-05,
+      "loss": 4.4654,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721686453353495e-05,
+      "loss": 4.4668,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720847858602443e-05,
+      "loss": 4.4646,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720010901731764e-05,
+      "loss": 4.4693,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719172306980712e-05,
+      "loss": 4.4502,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71833371222966e-05,
+      "loss": 4.4557,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717495117478608e-05,
+      "loss": 4.4417,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716656522727556e-05,
+      "loss": 4.4503,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715819565856877e-05,
+      "loss": 4.4473,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714980971105825e-05,
+      "loss": 4.4565,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714142376354773e-05,
+      "loss": 4.4466,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713303781603721e-05,
+      "loss": 4.448,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124651868526685e-05,
+      "loss": 4.4509,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71162822998199e-05,
+      "loss": 4.4519,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710789635230938e-05,
+      "loss": 4.4549,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709951040479886e-05,
+      "loss": 4.4401,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709112445728834e-05,
+      "loss": 4.4299,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708275488858156e-05,
+      "loss": 4.4523,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074368941071037e-05,
+      "loss": 4.4506,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7065982993560517e-05,
+      "loss": 4.4423,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705759704604999e-05,
+      "loss": 4.4409,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704921109853947e-05,
+      "loss": 4.4328,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704082515102895e-05,
+      "loss": 4.4359,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703245558232216e-05,
+      "loss": 4.4424,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702406963481164e-05,
+      "loss": 4.4238,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701568368730112e-05,
+      "loss": 4.4364,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70072977397906e-05,
+      "loss": 4.4327,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699891179228008e-05,
+      "loss": 4.4314,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699052584476956e-05,
+      "loss": 4.4183,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982139897259045e-05,
+      "loss": 4.4412,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973753949748525e-05,
+      "loss": 4.4174,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965384381041734e-05,
+      "loss": 4.43,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956998433531214e-05,
+      "loss": 4.4268,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6948612486020694e-05,
+      "loss": 4.4332,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6940226538510174e-05,
+      "loss": 4.4163,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693185696980338e-05,
+      "loss": 4.4182,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692347102229286e-05,
+      "loss": 4.406,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691508507478234e-05,
+      "loss": 4.4256,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690669912727182e-05,
+      "loss": 4.4273,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689832955856503e-05,
+      "loss": 4.4231,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688994361105451e-05,
+      "loss": 4.4254,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688157404234773e-05,
+      "loss": 4.4321,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687318809483721e-05,
+      "loss": 4.4329,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686480214732669e-05,
+      "loss": 4.4107,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685641619981617e-05,
+      "loss": 4.4198,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684803025230565e-05,
+      "loss": 4.4264,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683964430479513e-05,
+      "loss": 4.4014,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683125835728461e-05,
+      "loss": 4.4196,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682288878857782e-05,
+      "loss": 4.4041,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68145028410673e-05,
+      "loss": 4.4284,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680611689355678e-05,
+      "loss": 4.4183,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679773094604626e-05,
+      "loss": 4.3975,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678934499853574e-05,
+      "loss": 4.4103,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678095905102522e-05,
+      "loss": 4.4088,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772573103514697e-05,
+      "loss": 4.4177,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764187156004177e-05,
+      "loss": 4.4119,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675581758729739e-05,
+      "loss": 4.409,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674743163978687e-05,
+      "loss": 4.4055,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739045692276346e-05,
+      "loss": 4.4022,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730659744765826e-05,
+      "loss": 4.4002,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672229017605904e-05,
+      "loss": 4.3886,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671390422854852e-05,
+      "loss": 4.4001,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705518281037995e-05,
+      "loss": 4.4085,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697132333527474e-05,
+      "loss": 4.4021,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668876276482069e-05,
+      "loss": 4.3933,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668037681731017e-05,
+      "loss": 4.391,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667199086979965e-05,
+      "loss": 4.3959,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666360492228913e-05,
+      "loss": 4.4069,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665521897477861e-05,
+      "loss": 4.3948,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664683302726809e-05,
+      "loss": 4.3959,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663844707975757e-05,
+      "loss": 4.3897,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663006113224705e-05,
+      "loss": 4.4003,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662169156354026e-05,
+      "loss": 4.3792,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661330561602974e-05,
+      "loss": 4.3943,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660493604732295e-05,
+      "loss": 4.3856,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659655009981243e-05,
+      "loss": 4.3884,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658816415230191e-05,
+      "loss": 4.3849,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657977820479139e-05,
+      "loss": 4.3866,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657139225728087e-05,
+      "loss": 4.3859,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563006309770355e-05,
+      "loss": 4.381,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554620362259835e-05,
+      "loss": 4.3866,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546250793553044e-05,
+      "loss": 4.382,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6537864846042524e-05,
+      "loss": 4.3903,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652949527733573e-05,
+      "loss": 4.391,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652110932982521e-05,
+      "loss": 4.3874,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651272338231469e-05,
+      "loss": 4.3871,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650433743480417e-05,
+      "loss": 4.3833,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649595148729365e-05,
+      "loss": 4.379,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648756553978313e-05,
+      "loss": 4.3792,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647917959227261e-05,
+      "loss": 4.3795,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647079364476209e-05,
+      "loss": 4.3751,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646242407605531e-05,
+      "loss": 4.3849,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645403812854479e-05,
+      "loss": 4.3759,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644565218103427e-05,
+      "loss": 4.3812,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643726623352375e-05,
+      "loss": 4.373,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6428913043620687e-05,
+      "loss": 4.3744,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420527096110166e-05,
+      "loss": 4.3722,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412157527403376e-05,
+      "loss": 4.3879,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403771579892856e-05,
+      "loss": 4.366,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395385632382335e-05,
+      "loss": 4.3631,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6386999684871815e-05,
+      "loss": 4.3591,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378613737361295e-05,
+      "loss": 4.3752,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370227789850775e-05,
+      "loss": 4.3809,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636184184234026e-05,
+      "loss": 4.3771,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635345589482974e-05,
+      "loss": 4.3701,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634506994731922e-05,
+      "loss": 4.3778,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63366839998087e-05,
+      "loss": 4.3695,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632829805229818e-05,
+      "loss": 4.3732,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631992848359139e-05,
+      "loss": 4.3628,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631154253608087e-05,
+      "loss": 4.3722,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630315658857035e-05,
+      "loss": 4.3668,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629477064105983e-05,
+      "loss": 4.3572,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286384693549304e-05,
+      "loss": 4.3724,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277998746038784e-05,
+      "loss": 4.3628,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269612798528264e-05,
+      "loss": 4.3635,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6261226851017744e-05,
+      "loss": 4.3637,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625285728231096e-05,
+      "loss": 4.3521,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.347398281097412,
+      "eval_runtime": 291.2821,
+      "eval_samples_per_second": 1310.039,
+      "eval_steps_per_second": 40.94,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244487713604176e-05,
+      "loss": 4.3667,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6236101766093656e-05,
+      "loss": 4.357,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622771581858313e-05,
+      "loss": 4.365,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621932987107261e-05,
+      "loss": 4.3576,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621094392356209e-05,
+      "loss": 4.3697,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620255797605157e-05,
+      "loss": 4.3511,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619417202854105e-05,
+      "loss": 4.3473,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618578608103053e-05,
+      "loss": 4.34,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617741651232374e-05,
+      "loss": 4.3652,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616903056481322e-05,
+      "loss": 4.3632,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61606446173027e-05,
+      "loss": 4.3478,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152258669792184e-05,
+      "loss": 4.3661,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6143889101085393e-05,
+      "loss": 4.3467,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613550315357487e-05,
+      "loss": 4.351,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612711720606435e-05,
+      "loss": 4.3498,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611873125855383e-05,
+      "loss": 4.3424,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611034531104331e-05,
+      "loss": 4.3494,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610195936353279e-05,
+      "loss": 4.3436,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609357341602227e-05,
+      "loss": 4.3526,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608518746851175e-05,
+      "loss": 4.3549,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607681789980496e-05,
+      "loss": 4.353,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606843195229444e-05,
+      "loss": 4.3449,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606004600478392e-05,
+      "loss": 4.339,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60516600572734e-05,
+      "loss": 4.3534,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604329048856662e-05,
+      "loss": 4.3351,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60349045410561e-05,
+      "loss": 4.3404,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602651859354558e-05,
+      "loss": 4.3305,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601813264603506e-05,
+      "loss": 4.35,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600976307732827e-05,
+      "loss": 4.3247,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600137712981775e-05,
+      "loss": 4.3353,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599299118230723e-05,
+      "loss": 4.3432,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598460523479671e-05,
+      "loss": 4.3388,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597621928728619e-05,
+      "loss": 4.3394,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967849718579396e-05,
+      "loss": 4.3418,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959463771068876e-05,
+      "loss": 4.3407,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5951077823558356e-05,
+      "loss": 4.3433,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594270825485157e-05,
+      "loss": 4.3311,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593432230734105e-05,
+      "loss": 4.324,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592593635983053e-05,
+      "loss": 4.3203,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591755041232001e-05,
+      "loss": 4.3274,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590916446480949e-05,
+      "loss": 4.3239,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5900778517298965e-05,
+      "loss": 4.3364,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5892392569788445e-05,
+      "loss": 4.3246,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5884006622277925e-05,
+      "loss": 4.3271,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587563705357114e-05,
+      "loss": 4.3322,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586726748486435e-05,
+      "loss": 4.3258,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585888153735383e-05,
+      "loss": 4.3372,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585049558984331e-05,
+      "loss": 4.3212,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584210964233279e-05,
+      "loss": 4.3057,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5833740073626005e-05,
+      "loss": 4.3377,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5825354126115485e-05,
+      "loss": 4.3303,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5816968178604965e-05,
+      "loss": 4.3285,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580858223109444e-05,
+      "loss": 4.3199,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580019628358392e-05,
+      "loss": 4.3167,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57918103360734e-05,
+      "loss": 4.3182,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578342438856288e-05,
+      "loss": 4.3222,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577503844105236e-05,
+      "loss": 4.3092,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576666887234557e-05,
+      "loss": 4.3147,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575828292483505e-05,
+      "loss": 4.323,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574989697732453e-05,
+      "loss": 4.3128,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574151102981401e-05,
+      "loss": 4.3026,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733125082303494e-05,
+      "loss": 4.3286,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57247555135967e-05,
+      "loss": 4.302,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571636956608618e-05,
+      "loss": 4.3098,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570798361857566e-05,
+      "loss": 4.3174,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569959767106514e-05,
+      "loss": 4.3208,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569122810235835e-05,
+      "loss": 4.3037,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568284215484783e-05,
+      "loss": 4.307,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567445620733731e-05,
+      "loss": 4.2906,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566607025982679e-05,
+      "loss": 4.3178,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565768431231627e-05,
+      "loss": 4.3129,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564929836480575e-05,
+      "loss": 4.3128,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564091241729523e-05,
+      "loss": 4.3103,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563252646978471e-05,
+      "loss": 4.3222,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562415690107793e-05,
+      "loss": 4.325,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615787332371137e-05,
+      "loss": 4.2986,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5607401384860617e-05,
+      "loss": 4.3104,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5599015437350096e-05,
+      "loss": 4.3181,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5590629489839576e-05,
+      "loss": 4.2917,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582243542329056e-05,
+      "loss": 4.3128,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5573857594818536e-05,
+      "loss": 4.2993,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5565471647308016e-05,
+      "loss": 4.316,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557085699797496e-05,
+      "loss": 4.313,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548716131090705e-05,
+      "loss": 4.2912,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5540330183580185e-05,
+      "loss": 4.3037,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5531944236069665e-05,
+      "loss": 4.3008,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523558288559145e-05,
+      "loss": 4.3126,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551518871985236e-05,
+      "loss": 4.3012,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550681915114557e-05,
+      "loss": 4.3013,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549843320363505e-05,
+      "loss": 4.2992,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549004725612453e-05,
+      "loss": 4.2999,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548166130861401e-05,
+      "loss": 4.2952,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547327536110349e-05,
+      "loss": 4.284,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546488941359297e-05,
+      "loss": 4.2931,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545651984488618e-05,
+      "loss": 4.3021,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544813389737566e-05,
+      "loss": 4.3037,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543976432866887e-05,
+      "loss": 4.2848,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431378381158355e-05,
+      "loss": 4.2915,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422992433647835e-05,
+      "loss": 4.2922,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414606486137315e-05,
+      "loss": 4.3039,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406220538626795e-05,
+      "loss": 4.2919,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397834591116275e-05,
+      "loss": 4.294,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538944864360575e-05,
+      "loss": 4.2862,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538106269609523e-05,
+      "loss": 4.3013,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537267674858471e-05,
+      "loss": 4.2751,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536429080107419e-05,
+      "loss": 4.2926,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53559212323674e-05,
+      "loss": 4.2831,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534753528485688e-05,
+      "loss": 4.2884,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533914933734636e-05,
+      "loss": 4.283,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533076338983584e-05,
+      "loss": 4.293,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532237744232532e-05,
+      "loss": 4.282,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53139914948148e-05,
+      "loss": 4.2823,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530560554730428e-05,
+      "loss": 4.2853,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529721959979376e-05,
+      "loss": 4.2851,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528885003108697e-05,
+      "loss": 4.2921,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528046408357645e-05,
+      "loss": 4.2936,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527209451486966e-05,
+      "loss": 4.2881,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526370856735914e-05,
+      "loss": 4.2914,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525532261984862e-05,
+      "loss": 4.2867,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52469366723381e-05,
+      "loss": 4.279,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523855072482758e-05,
+      "loss": 4.2809,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523018115612079e-05,
+      "loss": 4.2817,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522179520861027e-05,
+      "loss": 4.2811,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521340926109976e-05,
+      "loss": 4.2871,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520502331358924e-05,
+      "loss": 4.2799,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5196653744882446e-05,
+      "loss": 4.2877,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5188284176175655e-05,
+      "loss": 4.276,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5179898228665135e-05,
+      "loss": 4.2794,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5171512281154615e-05,
+      "loss": 4.2769,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5163126333644095e-05,
+      "loss": 4.2923,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154740386133575e-05,
+      "loss": 4.2741,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146354438623055e-05,
+      "loss": 4.2679,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5137968491112535e-05,
+      "loss": 4.2638,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129582543602015e-05,
+      "loss": 4.2815,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121212974895224e-05,
+      "loss": 4.2878,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511282702738471e-05,
+      "loss": 4.2869,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510444107987419e-05,
+      "loss": 4.2777,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509605513236367e-05,
+      "loss": 4.2838,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508766918485315e-05,
+      "loss": 4.2774,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507928323734263e-05,
+      "loss": 4.2783,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070897289832104e-05,
+      "loss": 4.271,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062511342321584e-05,
+      "loss": 4.2804,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50541417736148e-05,
+      "loss": 4.2774,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504575582610428e-05,
+      "loss": 4.2644,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503736987859375e-05,
+      "loss": 4.2814,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502898393108323e-05,
+      "loss": 4.2742,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502059798357271e-05,
+      "loss": 4.2673,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501221203606219e-05,
+      "loss": 4.2727,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500382608855168e-05,
+      "loss": 4.2673,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.265713214874268,
+      "eval_runtime": 288.8086,
+      "eval_samples_per_second": 1321.259,
+      "eval_steps_per_second": 41.29,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499545651984489e-05,
+      "loss": 4.273,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498707057233437e-05,
+      "loss": 4.2684,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497868462482385e-05,
+      "loss": 4.276,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497029867731333e-05,
+      "loss": 4.2647,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496191272980281e-05,
+      "loss": 4.2778,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495352678229229e-05,
+      "loss": 4.2664,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494514083478177e-05,
+      "loss": 4.2566,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493675488727125e-05,
+      "loss": 4.2573,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492836893976073e-05,
+      "loss": 4.2723,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491998299225021e-05,
+      "loss": 4.2736,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491159704473969e-05,
+      "loss": 4.2633,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490321109722917e-05,
+      "loss": 4.2738,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489484152852238e-05,
+      "loss": 4.2603,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886455581011864e-05,
+      "loss": 4.2644,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878069633501344e-05,
+      "loss": 4.2606,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869683685990824e-05,
+      "loss": 4.2557,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48612977384803e-05,
+      "loss": 4.2601,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485291179096978e-05,
+      "loss": 4.2562,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484452584345926e-05,
+      "loss": 4.2679,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483613989594874e-05,
+      "loss": 4.2681,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4827753948438217e-05,
+      "loss": 4.2684,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4819368000927697e-05,
+      "loss": 4.2599,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4810982053417177e-05,
+      "loss": 4.2542,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4802596105906656e-05,
+      "loss": 4.2644,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4794226537199866e-05,
+      "loss": 4.2544,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4785840589689346e-05,
+      "loss": 4.2551,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4777454642178825e-05,
+      "loss": 4.245,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476906869466831e-05,
+      "loss": 4.2609,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476069912596152e-05,
+      "loss": 4.2472,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752313178451e-05,
+      "loss": 4.2464,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474392723094048e-05,
+      "loss": 4.2608,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473554128342996e-05,
+      "loss": 4.2541,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472715533591944e-05,
+      "loss": 4.2599,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471878576721265e-05,
+      "loss": 4.2613,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471041619850586e-05,
+      "loss": 4.2525,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470203025099534e-05,
+      "loss": 4.2611,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469364430348482e-05,
+      "loss": 4.2515,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46852583559743e-05,
+      "loss": 4.241,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467687240846378e-05,
+      "loss": 4.2377,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668486460953266e-05,
+      "loss": 4.2432,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660100513442746e-05,
+      "loss": 4.2432,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651714565932226e-05,
+      "loss": 4.255,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643344997225435e-05,
+      "loss": 4.2426,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4634959049714915e-05,
+      "loss": 4.2448,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626573102204395e-05,
+      "loss": 4.2535,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618187154693875e-05,
+      "loss": 4.2481,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460983396479081e-05,
+      "loss": 4.2543,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460144801728029e-05,
+      "loss": 4.2401,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459306206976977e-05,
+      "loss": 4.2234,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458467612225925e-05,
+      "loss": 4.2563,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457629017474873e-05,
+      "loss": 4.2557,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456790422723822e-05,
+      "loss": 4.2485,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45595182797277e-05,
+      "loss": 4.2412,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455113233221718e-05,
+      "loss": 4.2367,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454274638470665e-05,
+      "loss": 4.2382,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453436043719613e-05,
+      "loss": 4.2416,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452597448968561e-05,
+      "loss": 4.233,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451760492097883e-05,
+      "loss": 4.2372,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45092189734683e-05,
+      "loss": 4.2469,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450083302595778e-05,
+      "loss": 4.2381,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449244707844726e-05,
+      "loss": 4.2254,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448406113093674e-05,
+      "loss": 4.2517,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447569156222996e-05,
+      "loss": 4.2262,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446730561471944e-05,
+      "loss": 4.2281,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445891966720892e-05,
+      "loss": 4.2431,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44505337196984e-05,
+      "loss": 4.2471,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444214777218788e-05,
+      "loss": 4.2236,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443376182467736e-05,
+      "loss": 4.2331,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442537587716684e-05,
+      "loss": 4.2158,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441698992965632e-05,
+      "loss": 4.2412,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4408620360949526e-05,
+      "loss": 4.2407,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400234413439006e-05,
+      "loss": 4.2321,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4391848465928486e-05,
+      "loss": 4.2374,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4383462518417966e-05,
+      "loss": 4.2428,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4375092949711175e-05,
+      "loss": 4.2533,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436672338100439e-05,
+      "loss": 4.2235,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435833743349387e-05,
+      "loss": 4.2364,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434995148598335e-05,
+      "loss": 4.2417,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434156553847283e-05,
+      "loss": 4.2167,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433317959096231e-05,
+      "loss": 4.2401,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432481002225552e-05,
+      "loss": 4.2269,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316424074745e-05,
+      "loss": 4.2413,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430803812723448e-05,
+      "loss": 4.2392,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429965217972396e-05,
+      "loss": 4.2207,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429128261101717e-05,
+      "loss": 4.2264,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428289666350665e-05,
+      "loss": 4.2264,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427451071599613e-05,
+      "loss": 4.2426,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426612476848561e-05,
+      "loss": 4.2293,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257738820975095e-05,
+      "loss": 4.2259,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249352873464575e-05,
+      "loss": 4.2234,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4240983304757784e-05,
+      "loss": 4.2303,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232597357247264e-05,
+      "loss": 4.2227,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224211409736744e-05,
+      "loss": 4.2091,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4215825462226224e-05,
+      "loss": 4.2241,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207439514715704e-05,
+      "loss": 4.2291,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4199053567205184e-05,
+      "loss": 4.2312,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190667619694664e-05,
+      "loss": 4.2127,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418228167218414e-05,
+      "loss": 4.2186,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417391210347735e-05,
+      "loss": 4.2212,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416552615596683e-05,
+      "loss": 4.2298,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415714020845631e-05,
+      "loss": 4.2243,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414875426094579e-05,
+      "loss": 4.2228,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414036831343527e-05,
+      "loss": 4.2151,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413199874472849e-05,
+      "loss": 4.2292,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412361279721796e-05,
+      "loss": 4.2039,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411522684970744e-05,
+      "loss": 4.2223,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410685728100066e-05,
+      "loss": 4.211,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409847133349014e-05,
+      "loss": 4.2207,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409008538597961e-05,
+      "loss": 4.2084,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408169943846909e-05,
+      "loss": 4.2232,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407331349095857e-05,
+      "loss": 4.2141,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406492754344805e-05,
+      "loss": 4.2115,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405654159593753e-05,
+      "loss": 4.2156,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404817202723075e-05,
+      "loss": 4.216,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403978607972023e-05,
+      "loss": 4.2214,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403140013220971e-05,
+      "loss": 4.228,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402301418469919e-05,
+      "loss": 4.2222,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401462823718867e-05,
+      "loss": 4.2213,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4006242289678147e-05,
+      "loss": 4.2171,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3997856342167627e-05,
+      "loss": 4.2119,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3989470394657106e-05,
+      "loss": 4.2086,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981100825950316e-05,
+      "loss": 4.2194,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3972714878439796e-05,
+      "loss": 4.2102,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3964328930929276e-05,
+      "loss": 4.2174,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3955942983418755e-05,
+      "loss": 4.2113,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947573414711965e-05,
+      "loss": 4.2201,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393918746720145e-05,
+      "loss": 4.2095,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393081789849466e-05,
+      "loss": 4.2125,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392243195098414e-05,
+      "loss": 4.2084,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391404600347362e-05,
+      "loss": 4.2206,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39056600559631e-05,
+      "loss": 4.2099,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389727410845258e-05,
+      "loss": 4.2021,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388888816094206e-05,
+      "loss": 4.1948,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388051859223527e-05,
+      "loss": 4.2158,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387213264472475e-05,
+      "loss": 4.2203,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386374669721423e-05,
+      "loss": 4.2185,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385536074970371e-05,
+      "loss": 4.2141,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384697480219319e-05,
+      "loss": 4.2157,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383858885468267e-05,
+      "loss": 4.2096,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383020290717215e-05,
+      "loss": 4.2173,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382181695966163e-05,
+      "loss": 4.2014,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813463769758574e-05,
+      "loss": 4.2116,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805077822248054e-05,
+      "loss": 4.209,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796691874737534e-05,
+      "loss": 4.2031,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788305927227014e-05,
+      "loss": 4.214,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377993635852022e-05,
+      "loss": 4.2079,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37715504110097e-05,
+      "loss": 4.2051,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376316446349918e-05,
+      "loss": 4.2062,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375477851598866e-05,
+      "loss": 4.2044,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.2105817794799805,
+      "eval_runtime": 292.2977,
+      "eval_samples_per_second": 1305.488,
+      "eval_steps_per_second": 40.797,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374639256847814e-05,
+      "loss": 4.2057,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373802299977136e-05,
+      "loss": 4.2043,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372963705226084e-05,
+      "loss": 4.2104,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372125110475032e-05,
+      "loss": 4.2014,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37128651572398e-05,
+      "loss": 4.2134,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370447920972927e-05,
+      "loss": 4.2027,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369609326221875e-05,
+      "loss": 4.1932,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368770731470823e-05,
+      "loss": 4.1907,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367933774600145e-05,
+      "loss": 4.2087,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367095179849092e-05,
+      "loss": 4.2099,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36625658509804e-05,
+      "loss": 4.2028,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365417990346988e-05,
+      "loss": 4.2075,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3645810334763096e-05,
+      "loss": 4.2012,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3637424387252576e-05,
+      "loss": 4.2041,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3629038439742056e-05,
+      "loss": 4.1985,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3620652492231536e-05,
+      "loss": 4.1892,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3612266544721016e-05,
+      "loss": 4.1992,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3603880597210496e-05,
+      "loss": 4.1932,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3595494649699976e-05,
+      "loss": 4.205,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3587108702189456e-05,
+      "loss": 4.206,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3578722754678936e-05,
+      "loss": 4.2104,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3570336807168416e-05,
+      "loss": 4.1987,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3561950859657896e-05,
+      "loss": 4.195,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3553564912147376e-05,
+      "loss": 4.1994,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3545195343440585e-05,
+      "loss": 4.1918,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353684215353753e-05,
+      "loss": 4.1958,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352845620602701e-05,
+      "loss": 4.18,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352007025851649e-05,
+      "loss": 4.2016,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351168431100597e-05,
+      "loss": 4.1893,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350329836349545e-05,
+      "loss": 4.1851,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349491241598493e-05,
+      "loss": 4.1944,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348652646847441e-05,
+      "loss": 4.199,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347814052096389e-05,
+      "loss": 4.2006,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346975457345337e-05,
+      "loss": 4.1972,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346138500474658e-05,
+      "loss": 4.1918,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345299905723606e-05,
+      "loss": 4.1989,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344461310972554e-05,
+      "loss": 4.1966,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343622716221502e-05,
+      "loss": 4.1788,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34278412147045e-05,
+      "loss": 4.1802,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341945526719398e-05,
+      "loss": 4.181,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341106931968346e-05,
+      "loss": 4.1874,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402699750976674e-05,
+      "loss": 4.1953,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3394313803466154e-05,
+      "loss": 4.1851,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385927855955634e-05,
+      "loss": 4.1819,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337754190844511e-05,
+      "loss": 4.1949,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336915596093459e-05,
+      "loss": 4.1865,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33607863922278e-05,
+      "loss": 4.1995,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3352400444717276e-05,
+      "loss": 4.1816,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3344014497206756e-05,
+      "loss": 4.1613,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3335628549696236e-05,
+      "loss": 4.1992,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3327242602185716e-05,
+      "loss": 4.1919,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331887303347893e-05,
+      "loss": 4.1924,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331048708596841e-05,
+      "loss": 4.185,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330210113845789e-05,
+      "loss": 4.1803,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329371519094737e-05,
+      "loss": 4.1783,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328532924343685e-05,
+      "loss": 4.1814,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327694329592633e-05,
+      "loss": 4.1759,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326855734841581e-05,
+      "loss": 4.1791,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326017140090529e-05,
+      "loss": 4.1871,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32518018321985e-05,
+      "loss": 4.1803,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324341588468798e-05,
+      "loss": 4.1694,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323502993717746e-05,
+      "loss": 4.1911,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322664398966694e-05,
+      "loss": 4.171,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321827442096015e-05,
+      "loss": 4.1716,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3209904852253366e-05,
+      "loss": 4.185,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201518904742846e-05,
+      "loss": 4.1878,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193132957232326e-05,
+      "loss": 4.166,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3184747009721806e-05,
+      "loss": 4.1793,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176361062211286e-05,
+      "loss": 4.1593,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3167991493504495e-05,
+      "loss": 4.1814,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3159605545993975e-05,
+      "loss": 4.1821,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151219598483455e-05,
+      "loss": 4.1798,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3142833650972935e-05,
+      "loss": 4.1795,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3134447703462415e-05,
+      "loss": 4.1865,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3126061755951895e-05,
+      "loss": 4.1929,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3117692187245104e-05,
+      "loss": 4.1685,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310930623973459e-05,
+      "loss": 4.1833,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310092029222407e-05,
+      "loss": 4.1857,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309253434471355e-05,
+      "loss": 4.1608,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308414839720303e-05,
+      "loss": 4.1854,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307576244969251e-05,
+      "loss": 4.1713,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306739288098572e-05,
+      "loss": 4.1845,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30590069334752e-05,
+      "loss": 4.1831,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305062098596468e-05,
+      "loss": 4.168,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304223503845416e-05,
+      "loss": 4.1663,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303384909094364e-05,
+      "loss": 4.1717,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302546314343311e-05,
+      "loss": 4.1883,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301707719592259e-05,
+      "loss": 4.1722,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300869124841207e-05,
+      "loss": 4.1724,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300032167970529e-05,
+      "loss": 4.1699,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299193573219477e-05,
+      "loss": 4.1761,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.298354978468425e-05,
+      "loss": 4.1683,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297516383717373e-05,
+      "loss": 4.1546,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296677788966321e-05,
+      "loss": 4.1683,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295839194215269e-05,
+      "loss": 4.1772,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295000599464217e-05,
+      "loss": 4.178,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294162004713165e-05,
+      "loss": 4.1608,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293325047842486e-05,
+      "loss": 4.1631,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2924880909718066e-05,
+      "loss": 4.1681,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2916494962207546e-05,
+      "loss": 4.1733,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2908109014697026e-05,
+      "loss": 4.1764,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2899723067186506e-05,
+      "loss": 4.1624,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289133711967599e-05,
+      "loss": 4.1668,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288295117216547e-05,
+      "loss": 4.1746,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287456522465495e-05,
+      "loss": 4.1524,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286617927714443e-05,
+      "loss": 4.1661,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285782608724137e-05,
+      "loss": 4.1574,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284944013973085e-05,
+      "loss": 4.1702,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284105419222033e-05,
+      "loss": 4.1563,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283266824470981e-05,
+      "loss": 4.1709,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282428229719929e-05,
+      "loss": 4.1608,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281589634968877e-05,
+      "loss": 4.1611,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280751040217825e-05,
+      "loss": 4.1619,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279912445466773e-05,
+      "loss": 4.1663,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2790754885960946e-05,
+      "loss": 4.1631,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782368938450426e-05,
+      "loss": 4.1801,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2773999369743635e-05,
+      "loss": 4.169,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2765613422233115e-05,
+      "loss": 4.1715,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757243853526324e-05,
+      "loss": 4.1611,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2748857906015804e-05,
+      "loss": 4.1653,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2740471958505284e-05,
+      "loss": 4.1574,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2732086010994764e-05,
+      "loss": 4.1643,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2723700063484244e-05,
+      "loss": 4.1609,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2715314115973724e-05,
+      "loss": 4.164,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2706928168463204e-05,
+      "loss": 4.1631,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2698542220952684e-05,
+      "loss": 4.164,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26901726522459e-05,
+      "loss": 4.1585,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268180308353911e-05,
+      "loss": 4.1654,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267341713602859e-05,
+      "loss": 4.1576,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266503118851807e-05,
+      "loss": 4.1666,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265664524100755e-05,
+      "loss": 4.1616,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264825929349703e-05,
+      "loss": 4.1551,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263987334598651e-05,
+      "loss": 4.1408,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263148739847599e-05,
+      "loss": 4.1676,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262310145096547e-05,
+      "loss": 4.1692,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261473188225868e-05,
+      "loss": 4.1685,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260634593474816e-05,
+      "loss": 4.163,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259795998723764e-05,
+      "loss": 4.1647,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258957403972712e-05,
+      "loss": 4.1594,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25811880922166e-05,
+      "loss": 4.1717,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2572818523509813e-05,
+      "loss": 4.1521,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2564432575999293e-05,
+      "loss": 4.1602,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255604662848877e-05,
+      "loss": 4.1582,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2547660680978247e-05,
+      "loss": 4.1561,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253929111227146e-05,
+      "loss": 4.167,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253090516476094e-05,
+      "loss": 4.154,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252251921725042e-05,
+      "loss": 4.1616,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2514133269739896e-05,
+      "loss": 4.1532,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2505747322229375e-05,
+      "loss": 4.1584,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.16970682144165,
+      "eval_runtime": 289.0674,
+      "eval_samples_per_second": 1320.076,
+      "eval_steps_per_second": 41.253,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249737775352259e-05,
+      "loss": 4.1653,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248899180601207e-05,
+      "loss": 4.1556,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248060585850155e-05,
+      "loss": 4.1614,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247221991099103e-05,
+      "loss": 4.1526,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246383396348051e-05,
+      "loss": 4.1665,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245544801596999e-05,
+      "loss": 4.1532,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244706206845947e-05,
+      "loss": 4.1472,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243867612094895e-05,
+      "loss": 4.1414,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243029017343843e-05,
+      "loss": 4.161,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242190422592791e-05,
+      "loss": 4.1602,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241351827841739e-05,
+      "loss": 4.1542,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240513233090687e-05,
+      "loss": 4.1603,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239676276220008e-05,
+      "loss": 4.1532,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238837681468956e-05,
+      "loss": 4.1546,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237999086717904e-05,
+      "loss": 4.1476,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237160491966852e-05,
+      "loss": 4.1439,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363218972158e-05,
+      "loss": 4.1506,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2354833024647487e-05,
+      "loss": 4.1497,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346447077136966e-05,
+      "loss": 4.1547,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233806112962644e-05,
+      "loss": 4.1584,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232967518211592e-05,
+      "loss": 4.165,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23212892346054e-05,
+      "loss": 4.1522,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231290328709488e-05,
+      "loss": 4.1476,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230453371838809e-05,
+      "loss": 4.1514,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229614777087757e-05,
+      "loss": 4.1487,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228776182336705e-05,
+      "loss": 4.1489,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227937587585653e-05,
+      "loss": 4.1301,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227098992834601e-05,
+      "loss": 4.1592,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226260398083549e-05,
+      "loss": 4.1456,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225421803332497e-05,
+      "loss": 4.1389,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245832085814455e-05,
+      "loss": 4.1471,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237462517107664e-05,
+      "loss": 4.1531,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2229076569597144e-05,
+      "loss": 4.1492,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2220690622086624e-05,
+      "loss": 4.1535,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2212304674576104e-05,
+      "loss": 4.1448,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220393510586931e-05,
+      "loss": 4.1554,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219554915835879e-05,
+      "loss": 4.1492,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218716321084827e-05,
+      "loss": 4.1327,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217877726333775e-05,
+      "loss": 4.1343,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217039131582723e-05,
+      "loss": 4.1357,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216200536831671e-05,
+      "loss": 4.1417,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215361942080619e-05,
+      "loss": 4.1507,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214524985209941e-05,
+      "loss": 4.138,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213688028339262e-05,
+      "loss": 4.1384,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21284943358821e-05,
+      "loss": 4.1458,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212010838837158e-05,
+      "loss": 4.1388,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211172244086106e-05,
+      "loss": 4.155,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210333649335054e-05,
+      "loss": 4.1389,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209495054584002e-05,
+      "loss": 4.1124,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20865645983295e-05,
+      "loss": 4.1546,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207817865081897e-05,
+      "loss": 4.1495,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206980908211219e-05,
+      "loss": 4.1459,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061423134601667e-05,
+      "loss": 4.1395,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2053037187091147e-05,
+      "loss": 4.1356,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2044651239580626e-05,
+      "loss": 4.1326,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036265292070106e-05,
+      "loss": 4.1394,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.202789572336332e-05,
+      "loss": 4.1321,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2019509775852796e-05,
+      "loss": 4.1331,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011123828342275e-05,
+      "loss": 4.1424,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2002737880831755e-05,
+      "loss": 4.1383,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1994351933321235e-05,
+      "loss": 4.125,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1985965985810715e-05,
+      "loss": 4.1469,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1977580038300195e-05,
+      "loss": 4.1251,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1969194090789675e-05,
+      "loss": 4.13,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1960824522082884e-05,
+      "loss": 4.1415,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1952438574572364e-05,
+      "loss": 4.1424,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1944052627061844e-05,
+      "loss": 4.1228,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1935666679551324e-05,
+      "loss": 4.1328,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192729711084454e-05,
+      "loss": 4.1164,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191891116333402e-05,
+      "loss": 4.1375,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19105252158235e-05,
+      "loss": 4.1397,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190213926831298e-05,
+      "loss": 4.1393,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189375332080246e-05,
+      "loss": 4.1332,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188538375209567e-05,
+      "loss": 4.1429,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187699780458515e-05,
+      "loss": 4.1514,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186862823587836e-05,
+      "loss": 4.1279,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186024228836784e-05,
+      "loss": 4.139,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185185634085732e-05,
+      "loss": 4.1412,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18434703933468e-05,
+      "loss": 4.1202,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183508444583628e-05,
+      "loss": 4.1392,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1826698498325765e-05,
+      "loss": 4.1342,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818312550815245e-05,
+      "loss": 4.1381,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1809926603304725e-05,
+      "loss": 4.1414,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1801557034597934e-05,
+      "loss": 4.1275,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1793171087087414e-05,
+      "loss": 4.1257,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1784785139576894e-05,
+      "loss": 4.1307,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776399192066373e-05,
+      "loss": 4.1392,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1768013244555853e-05,
+      "loss": 4.133,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.175964367584906e-05,
+      "loss": 4.1296,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175127410714227e-05,
+      "loss": 4.1303,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174288815963175e-05,
+      "loss": 4.1351,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173450221212123e-05,
+      "loss": 4.1265,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172611626461072e-05,
+      "loss": 4.1119,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17177303171002e-05,
+      "loss": 4.129,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170934436958968e-05,
+      "loss": 4.1314,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170095842207916e-05,
+      "loss": 4.1344,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169257247456863e-05,
+      "loss": 4.1212,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168420290586185e-05,
+      "loss": 4.1209,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167581695835133e-05,
+      "loss": 4.1284,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166743101084081e-05,
+      "loss": 4.1284,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1659061442134016e-05,
+      "loss": 4.1389,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1650675494623496e-05,
+      "loss": 4.1217,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642289547112976e-05,
+      "loss": 4.1255,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1633903599602456e-05,
+      "loss": 4.1322,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1625517652091936e-05,
+      "loss": 4.1107,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1617131704581416e-05,
+      "loss": 4.128,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160876213587463e-05,
+      "loss": 4.1117,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1600376188364105e-05,
+      "loss": 4.1323,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159200661965732e-05,
+      "loss": 4.1137,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15836206721468e-05,
+      "loss": 4.1294,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157523472463628e-05,
+      "loss": 4.1197,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1566848777125754e-05,
+      "loss": 4.118,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558462829615234e-05,
+      "loss": 4.1222,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1550076882104714e-05,
+      "loss": 4.1269,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1541690934594194e-05,
+      "loss": 4.1242,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1533304987083674e-05,
+      "loss": 4.1385,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1524919039573154e-05,
+      "loss": 4.1257,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151653309206264e-05,
+      "loss": 4.1383,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150814714455212e-05,
+      "loss": 4.1216,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149977757584533e-05,
+      "loss": 4.1213,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149139162833481e-05,
+      "loss": 4.1224,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148300568082429e-05,
+      "loss": 4.1202,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147461973331377e-05,
+      "loss": 4.1201,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146623378580325e-05,
+      "loss": 4.125,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145784783829273e-05,
+      "loss": 4.122,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144946189078221e-05,
+      "loss": 4.1261,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144107594327169e-05,
+      "loss": 4.1204,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14327063745649e-05,
+      "loss": 4.1253,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142432042705438e-05,
+      "loss": 4.1209,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141593447954386e-05,
+      "loss": 4.1218,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140754853203334e-05,
+      "loss": 4.1191,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139916258452282e-05,
+      "loss": 4.1183,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13907766370123e-05,
+      "loss": 4.1015,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138239068950178e-05,
+      "loss": 4.128,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137400474199126e-05,
+      "loss": 4.129,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136563517328447e-05,
+      "loss": 4.1312,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135724922577395e-05,
+      "loss": 4.1201,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134886327826343e-05,
+      "loss": 4.1293,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134047733075291e-05,
+      "loss": 4.1217,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332107762046116e-05,
+      "loss": 4.1304,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1323721814535596e-05,
+      "loss": 4.1111,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315335867025076e-05,
+      "loss": 4.1189,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130696629831829e-05,
+      "loss": 4.1224,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129858035080777e-05,
+      "loss": 4.1193,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129019440329725e-05,
+      "loss": 4.123,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128180845578673e-05,
+      "loss": 4.1165,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127342250827621e-05,
+      "loss": 4.1209,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126505293956942e-05,
+      "loss": 4.1141,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12566669920589e-05,
+      "loss": 4.1237,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.139695644378662,
+      "eval_runtime": 298.1743,
+      "eval_samples_per_second": 1279.758,
+      "eval_steps_per_second": 39.993,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124828104454838e-05,
+      "loss": 4.1158,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123989509703786e-05,
+      "loss": 4.1191,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123150914952734e-05,
+      "loss": 4.1183,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122312320201682e-05,
+      "loss": 4.1129,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12147372545063e-05,
+      "loss": 4.1251,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120635130699578e-05,
+      "loss": 4.1189,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1197981738288996e-05,
+      "loss": 4.1105,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1189595790778476e-05,
+      "loss": 4.1059,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1181226222071685e-05,
+      "loss": 4.1172,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1172840274561165e-05,
+      "loss": 4.1245,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1164454327050645e-05,
+      "loss": 4.1133,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1156068379540125e-05,
+      "loss": 4.1208,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1147698810833334e-05,
+      "loss": 4.1228,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1139312863322814e-05,
+      "loss": 4.1115,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1130926915812294e-05,
+      "loss": 4.1109,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1122540968301774e-05,
+      "loss": 4.1066,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1114155020791254e-05,
+      "loss": 4.1138,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1105769073280734e-05,
+      "loss": 4.1096,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109739950457395e-05,
+      "loss": 4.1125,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108901355706343e-05,
+      "loss": 4.1238,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108062760955291e-05,
+      "loss": 4.1259,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107224166204239e-05,
+      "loss": 4.1153,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106385571453187e-05,
+      "loss": 4.1099,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105546976702135e-05,
+      "loss": 4.1117,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104710019831456e-05,
+      "loss": 4.1115,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103871425080404e-05,
+      "loss": 4.1103,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103032830329352e-05,
+      "loss": 4.0963,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1021942355783e-05,
+      "loss": 4.1202,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101355640827247e-05,
+      "loss": 4.1098,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.100517046076195e-05,
+      "loss": 4.1028,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.099678451325143e-05,
+      "loss": 4.103,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098841494454465e-05,
+      "loss": 4.1177,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098002899703413e-05,
+      "loss": 4.1158,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097164304952361e-05,
+      "loss": 4.114,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096325710201309e-05,
+      "loss": 4.1072,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095487115450257e-05,
+      "loss": 4.1138,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094648520699205e-05,
+      "loss": 4.1148,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093809925948153e-05,
+      "loss": 4.097,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092971331197101e-05,
+      "loss": 4.097,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0921360122067946e-05,
+      "loss": 4.0983,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0912974174557426e-05,
+      "loss": 4.1044,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0904588227046906e-05,
+      "loss": 4.1106,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0896202279536386e-05,
+      "loss": 4.107,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08878327108296e-05,
+      "loss": 4.1011,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087944676331908e-05,
+      "loss": 4.1071,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087106081580856e-05,
+      "loss": 4.1021,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086267486829804e-05,
+      "loss": 4.1154,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085428892078752e-05,
+      "loss": 4.1071,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0845902973277e-05,
+      "loss": 4.078,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0837549783373946e-05,
+      "loss": 4.1153,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082916383586342e-05,
+      "loss": 4.1112,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08207778883529e-05,
+      "loss": 4.1151,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081239194084238e-05,
+      "loss": 4.1011,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080400599333186e-05,
+      "loss": 4.0985,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079562004582134e-05,
+      "loss": 4.1005,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0787234098310826e-05,
+      "loss": 4.1028,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0778848150800306e-05,
+      "loss": 4.1013,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0770462203289786e-05,
+      "loss": 4.0922,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0762076255779266e-05,
+      "loss": 4.1077,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0753690308268746e-05,
+      "loss": 4.1041,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0745304360758226e-05,
+      "loss": 4.0883,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0736918413247706e-05,
+      "loss": 4.1122,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0728548844540915e-05,
+      "loss": 4.0911,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0720162897030395e-05,
+      "loss": 4.0911,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0711776949519875e-05,
+      "loss": 4.1058,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0703391002009355e-05,
+      "loss": 4.1065,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0695021433302564e-05,
+      "loss": 4.0889,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0686635485792044e-05,
+      "loss": 4.0976,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067826591708526e-05,
+      "loss": 4.0809,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066987996957474e-05,
+      "loss": 4.1018,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066149402206422e-05,
+      "loss": 4.1035,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06531080745537e-05,
+      "loss": 4.1043,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064472212704318e-05,
+      "loss": 4.0975,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.063633617953266e-05,
+      "loss": 4.104,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062795023202214e-05,
+      "loss": 4.1191,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061958066331535e-05,
+      "loss": 4.088,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061119471580483e-05,
+      "loss": 4.1053,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060280876829431e-05,
+      "loss": 4.1061,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059442282078378e-05,
+      "loss": 4.0869,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058603687327326e-05,
+      "loss": 4.1027,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057765092576274e-05,
+      "loss": 4.102,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056926497825223e-05,
+      "loss": 4.1034,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056089540954544e-05,
+      "loss": 4.1116,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055250946203492e-05,
+      "loss": 4.0876,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05441235145244e-05,
+      "loss": 4.0935,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053573756701388e-05,
+      "loss": 4.0948,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0527367998307086e-05,
+      "loss": 4.1046,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0518982050796566e-05,
+      "loss": 4.096,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0510596103286046e-05,
+      "loss": 4.1006,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0502210155775526e-05,
+      "loss": 4.0983,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0493824208265006e-05,
+      "loss": 4.0946,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0485438260754486e-05,
+      "loss": 4.0935,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0477052313243966e-05,
+      "loss": 4.0779,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0468666365733446e-05,
+      "loss": 4.0926,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046029679702666e-05,
+      "loss": 4.0965,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045191084951614e-05,
+      "loss": 4.1025,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044352490200562e-05,
+      "loss": 4.0875,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04351389544951e-05,
+      "loss": 4.0838,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042676938578831e-05,
+      "loss": 4.093,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041839981708152e-05,
+      "loss": 4.0957,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0410013869571e-05,
+      "loss": 4.1003,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040162792206048e-05,
+      "loss": 4.0928,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039324197454996e-05,
+      "loss": 4.0941,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038487240584317e-05,
+      "loss": 4.0931,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037648645833265e-05,
+      "loss": 4.0834,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0368100510822135e-05,
+      "loss": 4.0882,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0359730942115345e-05,
+      "loss": 4.0798,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0351344994604824e-05,
+      "loss": 4.0997,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0342959047094304e-05,
+      "loss": 4.0806,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0334573099583784e-05,
+      "loss": 4.0949,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0326187152073264e-05,
+      "loss": 4.0859,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0317801204562744e-05,
+      "loss": 4.0922,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0309415257052224e-05,
+      "loss": 4.0866,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0301029309541704e-05,
+      "loss": 4.0915,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0292643362031184e-05,
+      "loss": 4.0907,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284257414520664e-05,
+      "loss": 4.1061,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027588784581387e-05,
+      "loss": 4.0888,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026750189830335e-05,
+      "loss": 4.1022,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025911595079283e-05,
+      "loss": 4.0856,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025073000328231e-05,
+      "loss": 4.0884,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02423440557718e-05,
+      "loss": 4.0885,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023397448706501e-05,
+      "loss": 4.0912,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022558853955449e-05,
+      "loss": 4.0837,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021720259204397e-05,
+      "loss": 4.0925,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020881664453344e-05,
+      "loss": 4.0846,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020043069702292e-05,
+      "loss": 4.0925,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01920447495124e-05,
+      "loss": 4.0905,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018367518080562e-05,
+      "loss": 4.0867,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017528923329509e-05,
+      "loss": 4.0895,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016690328578457e-05,
+      "loss": 4.0944,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015851733827406e-05,
+      "loss": 4.0837,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015013139076354e-05,
+      "loss": 4.0858,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014174544325302e-05,
+      "loss": 4.0664,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01333594957425e-05,
+      "loss": 4.0948,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012497354823198e-05,
+      "loss": 4.097,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011658760072146e-05,
+      "loss": 4.0977,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0108218032014667e-05,
+      "loss": 4.0878,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0099832084504146e-05,
+      "loss": 4.0938,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0091446136993626e-05,
+      "loss": 4.0903,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0083060189483106e-05,
+      "loss": 4.096,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0074690620776315e-05,
+      "loss": 4.08,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0066304673265795e-05,
+      "loss": 4.0871,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0057918725755275e-05,
+      "loss": 4.0871,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0049532778244755e-05,
+      "loss": 4.088,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0041146830734235e-05,
+      "loss": 4.0894,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003277726202745e-05,
+      "loss": 4.0813,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002439131451693e-05,
+      "loss": 4.0923,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001600536700641e-05,
+      "loss": 4.0794,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000761941949589e-05,
+      "loss": 4.0937,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.115260601043701,
+      "eval_runtime": 297.6648,
+      "eval_samples_per_second": 1281.949,
+      "eval_steps_per_second": 40.062,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999923347198537e-05,
+      "loss": 4.0875,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999084752447485e-05,
+      "loss": 4.086,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998246157696433e-05,
+      "loss": 4.087,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9974075629453804e-05,
+      "loss": 4.0846,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9965689681943284e-05,
+      "loss": 4.0953,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9957303734432764e-05,
+      "loss": 4.0861,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9948917786922244e-05,
+      "loss": 4.0764,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9940531839411724e-05,
+      "loss": 4.074,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9932145891901204e-05,
+      "loss": 4.0847,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992375994439069e-05,
+      "loss": 4.0927,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991537399688017e-05,
+      "loss": 4.0846,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990698804936965e-05,
+      "loss": 4.0869,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989861848066286e-05,
+      "loss": 4.0885,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989023253315234e-05,
+      "loss": 4.0793,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988184658564182e-05,
+      "loss": 4.079,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98734606381313e-05,
+      "loss": 4.0742,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986507469062078e-05,
+      "loss": 4.0831,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985668874311026e-05,
+      "loss": 4.0784,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984830279559974e-05,
+      "loss": 4.0823,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983991684808922e-05,
+      "loss": 4.0903,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983153090057869e-05,
+      "loss": 4.0942,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982314495306817e-05,
+      "loss": 4.0828,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981475900555766e-05,
+      "loss": 4.081,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980637305804714e-05,
+      "loss": 4.0803,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9798019868144084e-05,
+      "loss": 4.0794,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9789633920633564e-05,
+      "loss": 4.0789,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781247973123044e-05,
+      "loss": 4.0663,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977286202561252e-05,
+      "loss": 4.09,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9764476078102e-05,
+      "loss": 4.0796,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975609013059148e-05,
+      "loss": 4.0683,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974770418308096e-05,
+      "loss": 4.0727,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973931823557044e-05,
+      "loss": 4.0876,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973093228805992e-05,
+      "loss": 4.0862,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.97225463405494e-05,
+      "loss": 4.0845,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971416039303888e-05,
+      "loss": 4.0779,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970580720313582e-05,
+      "loss": 4.0773,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96974212556253e-05,
+      "loss": 4.0889,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968903530811478e-05,
+      "loss": 4.0665,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968064936060426e-05,
+      "loss": 4.067,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967226341309374e-05,
+      "loss": 4.0707,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966387746558322e-05,
+      "loss": 4.0691,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96554915180727e-05,
+      "loss": 4.0801,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964710557056218e-05,
+      "loss": 4.0699,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963871962305166e-05,
+      "loss": 4.0789,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963033367554114e-05,
+      "loss": 4.0736,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962196410683435e-05,
+      "loss": 4.0706,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.961357815932383e-05,
+      "loss": 4.0881,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960519221181331e-05,
+      "loss": 4.0741,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959680626430279e-05,
+      "loss": 4.0528,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9588436695596006e-05,
+      "loss": 4.0823,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9580067126889215e-05,
+      "loss": 4.0773,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9571681179378695e-05,
+      "loss": 4.0847,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9563295231868175e-05,
+      "loss": 4.072,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9554909284357655e-05,
+      "loss": 4.067,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9546523336847135e-05,
+      "loss": 4.0698,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9538137389336615e-05,
+      "loss": 4.0707,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9529751441826095e-05,
+      "loss": 4.0752,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9521365494315575e-05,
+      "loss": 4.0591,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9512995925608784e-05,
+      "loss": 4.0806,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9504609978098264e-05,
+      "loss": 4.0725,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9496224030587744e-05,
+      "loss": 4.0605,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948783808307723e-05,
+      "loss": 4.082,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947946851437044e-05,
+      "loss": 4.0604,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947108256685992e-05,
+      "loss": 4.0633,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94626966193494e-05,
+      "loss": 4.0741,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945431067183888e-05,
+      "loss": 4.0797,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944594110313209e-05,
+      "loss": 4.0544,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943755515562157e-05,
+      "loss": 4.0746,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942916920811105e-05,
+      "loss": 4.0525,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942078326060053e-05,
+      "loss": 4.0714,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941239731309e-05,
+      "loss": 4.0758,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940401136557948e-05,
+      "loss": 4.0698,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939562541806897e-05,
+      "loss": 4.0747,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938723947055845e-05,
+      "loss": 4.0729,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937885352304793e-05,
+      "loss": 4.0909,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9370500333144874e-05,
+      "loss": 4.0581,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9362114385634354e-05,
+      "loss": 4.0765,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935372843812383e-05,
+      "loss": 4.0732,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934534249061331e-05,
+      "loss": 4.0649,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933695654310279e-05,
+      "loss": 4.0722,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932857059559227e-05,
+      "loss": 4.0697,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9320184648081747e-05,
+      "loss": 4.0739,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9311798700571227e-05,
+      "loss": 4.0819,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9303429131864436e-05,
+      "loss": 4.0576,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929504318435392e-05,
+      "loss": 4.0654,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92866572368434e-05,
+      "loss": 4.0663,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927827128933288e-05,
+      "loss": 4.0787,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926990172062609e-05,
+      "loss": 4.0627,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926151577311557e-05,
+      "loss": 4.0703,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925312982560505e-05,
+      "loss": 4.0707,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924474387809453e-05,
+      "loss": 4.0662,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923635793058401e-05,
+      "loss": 4.0633,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922797198307349e-05,
+      "loss": 4.0493,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.92196024143667e-05,
+      "loss": 4.0624,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921121646685618e-05,
+      "loss": 4.0662,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920283051934566e-05,
+      "loss": 4.0732,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919444457183514e-05,
+      "loss": 4.0592,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918605862432462e-05,
+      "loss": 4.0561,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91776726768141e-05,
+      "loss": 4.0614,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916928672930359e-05,
+      "loss": 4.0706,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916090078179307e-05,
+      "loss": 4.07,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915251483428254e-05,
+      "loss": 4.0643,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9144145265575756e-05,
+      "loss": 4.0685,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9135759318065236e-05,
+      "loss": 4.0603,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9127373370554716e-05,
+      "loss": 4.0591,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911898742304419e-05,
+      "loss": 4.0567,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9110634233141134e-05,
+      "loss": 4.0563,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9102248285630614e-05,
+      "loss": 4.0717,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9093862338120094e-05,
+      "loss": 4.0533,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9085476390609574e-05,
+      "loss": 4.0644,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907710682190279e-05,
+      "loss": 4.0546,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906872087439227e-05,
+      "loss": 4.0661,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906033492688175e-05,
+      "loss": 4.0596,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905194897937123e-05,
+      "loss": 4.0609,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904356303186071e-05,
+      "loss": 4.0639,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903517708435019e-05,
+      "loss": 4.0744,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902679113683966e-05,
+      "loss": 4.0632,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901842156813288e-05,
+      "loss": 4.0724,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901003562062236e-05,
+      "loss": 4.062,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900164967311184e-05,
+      "loss": 4.0615,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899326372560131e-05,
+      "loss": 4.0634,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898487777809079e-05,
+      "loss": 4.0627,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897649183058028e-05,
+      "loss": 4.0546,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896810588306976e-05,
+      "loss": 4.0638,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895973631436297e-05,
+      "loss": 4.0561,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895135036685245e-05,
+      "loss": 4.0659,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894296441934193e-05,
+      "loss": 4.0587,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893457847183141e-05,
+      "loss": 4.0653,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8926208903124616e-05,
+      "loss": 4.0597,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8917822955614096e-05,
+      "loss": 4.0692,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8909437008103576e-05,
+      "loss": 4.0542,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8901051060593056e-05,
+      "loss": 4.0597,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8892665113082536e-05,
+      "loss": 4.0375,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8884279165572016e-05,
+      "loss": 4.0691,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8875893218061496e-05,
+      "loss": 4.0673,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8867507270550976e-05,
+      "loss": 4.074,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885913770184419e-05,
+      "loss": 4.0578,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885075175433367e-05,
+      "loss": 4.0678,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884236580682315e-05,
+      "loss": 4.0614,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883397985931263e-05,
+      "loss": 4.0731,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882561029060584e-05,
+      "loss": 4.0509,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881722434309532e-05,
+      "loss": 4.0578,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88088383955848e-05,
+      "loss": 4.0592,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880046882687801e-05,
+      "loss": 4.0673,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879208287936749e-05,
+      "loss": 4.0553,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878369693185697e-05,
+      "loss": 4.055,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877531098434645e-05,
+      "loss": 4.0627,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876692503683593e-05,
+      "loss": 4.059,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8758539089325416e-05,
+      "loss": 4.0646,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.096133232116699,
+      "eval_runtime": 308.4387,
+      "eval_samples_per_second": 1237.17,
+      "eval_steps_per_second": 38.662,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8750153141814896e-05,
+      "loss": 4.0516,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8741767194304376e-05,
+      "loss": 4.063,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8733397625597585e-05,
+      "loss": 4.056,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8725028056890794e-05,
+      "loss": 4.0589,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8716642109380274e-05,
+      "loss": 4.0699,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8708256161869754e-05,
+      "loss": 4.0582,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8699886593162963e-05,
+      "loss": 4.0501,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869150064565244e-05,
+      "loss": 4.0483,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.868311469814192e-05,
+      "loss": 4.0579,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86747287506314e-05,
+      "loss": 4.0676,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.866634280312088e-05,
+      "loss": 4.0573,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86579732344141e-05,
+      "loss": 4.0595,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864958728690358e-05,
+      "loss": 4.0639,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864120133939306e-05,
+      "loss": 4.0539,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.863281539188254e-05,
+      "loss": 4.0515,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862442944437202e-05,
+      "loss": 4.0468,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86160434968615e-05,
+      "loss": 4.0588,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.860765754935097e-05,
+      "loss": 4.0543,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859927160184045e-05,
+      "loss": 4.0504,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859088565432993e-05,
+      "loss": 4.0627,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858249970681941e-05,
+      "loss": 4.0682,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857411375930889e-05,
+      "loss": 4.0567,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.856574419060211e-05,
+      "loss": 4.0551,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855735824309159e-05,
+      "loss": 4.0568,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854897229558107e-05,
+      "loss": 4.0559,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854058634807055e-05,
+      "loss": 4.0535,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853220040056003e-05,
+      "loss": 4.0398,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852381445304951e-05,
+      "loss": 4.0616,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851544488434272e-05,
+      "loss": 4.0587,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8507058936832197e-05,
+      "loss": 4.0406,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8498672989321677e-05,
+      "loss": 4.0485,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8490287041811157e-05,
+      "loss": 4.0556,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8481901094300636e-05,
+      "loss": 4.0584,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8473531525593846e-05,
+      "loss": 4.0596,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8465145578083326e-05,
+      "loss": 4.0561,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8456759630572805e-05,
+      "loss": 4.0466,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8448373683062285e-05,
+      "loss": 4.0663,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843998773555177e-05,
+      "loss": 4.0384,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843160178804125e-05,
+      "loss": 4.04,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.842321584053073e-05,
+      "loss": 4.0434,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.841484627182394e-05,
+      "loss": 4.0471,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840646032431342e-05,
+      "loss": 4.0526,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83980743768029e-05,
+      "loss": 4.0457,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838968842929238e-05,
+      "loss": 4.0519,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838130248178186e-05,
+      "loss": 4.0491,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8372916534271334e-05,
+      "loss": 4.0443,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8364530586760814e-05,
+      "loss": 4.0567,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8356144639250294e-05,
+      "loss": 4.0557,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834777507054351e-05,
+      "loss": 4.0277,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833938912303299e-05,
+      "loss": 4.054,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8331019554326206e-05,
+      "loss": 4.0563,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8322633606815686e-05,
+      "loss": 4.061,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8314264038108895e-05,
+      "loss": 4.045,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8305878090598375e-05,
+      "loss": 4.0431,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8297492143087855e-05,
+      "loss": 4.0402,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8289106195577335e-05,
+      "loss": 4.0448,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828072024806681e-05,
+      "loss": 4.0505,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.827233430055629e-05,
+      "loss": 4.0343,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826394835304577e-05,
+      "loss": 4.0523,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.825556240553525e-05,
+      "loss": 4.0485,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824717645802473e-05,
+      "loss": 4.0383,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823879051051421e-05,
+      "loss": 4.051,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8230404563003694e-05,
+      "loss": 4.037,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8222018615493174e-05,
+      "loss": 4.0396,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8213632667982654e-05,
+      "loss": 4.0444,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8205263099275863e-05,
+      "loss": 4.0551,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819689353056907e-05,
+      "loss": 4.0356,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818850758305855e-05,
+      "loss": 4.0484,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818012163554803e-05,
+      "loss": 4.0244,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817173568803751e-05,
+      "loss": 4.0468,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816334974052699e-05,
+      "loss": 4.0484,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815496379301647e-05,
+      "loss": 4.0502,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814657784550595e-05,
+      "loss": 4.0436,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813819189799543e-05,
+      "loss": 4.0489,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812980595048491e-05,
+      "loss": 4.0634,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812145276058186e-05,
+      "loss": 4.0367,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811306681307134e-05,
+      "loss": 4.0515,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810468086556082e-05,
+      "loss": 4.042,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80962949180503e-05,
+      "loss": 4.0448,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808790897053978e-05,
+      "loss": 4.0456,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807952302302926e-05,
+      "loss": 4.0423,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807113707551874e-05,
+      "loss": 4.0509,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806275112800822e-05,
+      "loss": 4.06,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805436518049769e-05,
+      "loss": 4.0321,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8045995611790906e-05,
+      "loss": 4.0422,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8037609664280386e-05,
+      "loss": 4.0397,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80292400955736e-05,
+      "loss": 4.0545,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802085414806308e-05,
+      "loss": 4.0373,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.801246820055256e-05,
+      "loss": 4.0503,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.800408225304204e-05,
+      "loss": 4.0447,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.799569630553152e-05,
+      "loss": 4.0414,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7987310358020995e-05,
+      "loss": 4.0407,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7978924410510475e-05,
+      "loss": 4.0281,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7970538462999955e-05,
+      "loss": 4.037,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7962168894293164e-05,
+      "loss": 4.0439,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7953782946782644e-05,
+      "loss": 4.0508,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7945396999272124e-05,
+      "loss": 4.0333,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793702743056534e-05,
+      "loss": 4.0344,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792864148305482e-05,
+      "loss": 4.0379,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79202555355443e-05,
+      "loss": 4.0441,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791186958803378e-05,
+      "loss": 4.0457,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.790348364052326e-05,
+      "loss": 4.0384,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789509769301274e-05,
+      "loss": 4.0434,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788671174550222e-05,
+      "loss": 4.0389,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78783257979917e-05,
+      "loss": 4.0389,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786993985048118e-05,
+      "loss": 4.0295,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786158666057812e-05,
+      "loss": 4.0329,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78532007130676e-05,
+      "loss": 4.0438,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.784481476555708e-05,
+      "loss": 4.0332,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.783642881804656e-05,
+      "loss": 4.0387,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782804287053604e-05,
+      "loss": 4.0292,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781965692302552e-05,
+      "loss": 4.0404,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7811270975515004e-05,
+      "loss": 4.034,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7802885028004484e-05,
+      "loss": 4.0376,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.779451545929769e-05,
+      "loss": 4.0427,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778612951178717e-05,
+      "loss": 4.0536,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777775994308038e-05,
+      "loss": 4.0402,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776937399556986e-05,
+      "loss": 4.0448,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776098804805934e-05,
+      "loss": 4.0417,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.775260210054882e-05,
+      "loss": 4.0337,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77442161530383e-05,
+      "loss": 4.0432,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.773584658433151e-05,
+      "loss": 4.0323,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772746063682099e-05,
+      "loss": 4.0327,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771907468931047e-05,
+      "loss": 4.0419,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771068874179996e-05,
+      "loss": 4.0369,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770230279428944e-05,
+      "loss": 4.0418,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769393322558265e-05,
+      "loss": 4.0339,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7685547278072127e-05,
+      "loss": 4.0432,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7677161330561607e-05,
+      "loss": 4.0317,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7668775383051087e-05,
+      "loss": 4.0467,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7660389435540566e-05,
+      "loss": 4.0332,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7652019866833776e-05,
+      "loss": 4.0344,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7643633919323256e-05,
+      "loss": 4.0119,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7635247971812735e-05,
+      "loss": 4.0468,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7626862024302215e-05,
+      "loss": 4.042,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7618476076791695e-05,
+      "loss": 4.0516,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7610090129281175e-05,
+      "loss": 4.0355,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7601704181770655e-05,
+      "loss": 4.0446,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7593318234260135e-05,
+      "loss": 4.0368,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758494866555335e-05,
+      "loss": 4.052,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.757656271804283e-05,
+      "loss": 4.03,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7568176770532304e-05,
+      "loss": 4.0335,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7559790823021784e-05,
+      "loss": 4.0364,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7551421254315e-05,
+      "loss": 4.0436,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754305168560821e-05,
+      "loss": 4.0343,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753466573809769e-05,
+      "loss": 4.0314,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752627979058717e-05,
+      "loss": 4.0365,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751789384307665e-05,
+      "loss": 4.0388,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.750950789556613e-05,
+      "loss": 4.0397,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.080237865447998,
+      "eval_runtime": 301.6388,
+      "eval_samples_per_second": 1265.059,
+      "eval_steps_per_second": 39.534,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750112194805561e-05,
+      "loss": 4.0274,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749273600054509e-05,
+      "loss": 4.0413,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748435005303457e-05,
+      "loss": 4.0358,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747596410552405e-05,
+      "loss": 4.036,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746757815801353e-05,
+      "loss": 4.0463,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745919221050301e-05,
+      "loss": 4.0348,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745080626299249e-05,
+      "loss": 4.0324,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744242031548197e-05,
+      "loss": 4.0241,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.743403436797145e-05,
+      "loss": 4.0327,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.742564842046093e-05,
+      "loss": 4.0401,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.741726247295041e-05,
+      "loss": 4.0411,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740887652543989e-05,
+      "loss": 4.0364,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.74005069567331e-05,
+      "loss": 4.0413,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.739212100922258e-05,
+      "loss": 4.0292,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738373506171206e-05,
+      "loss": 4.0303,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7375349114201544e-05,
+      "loss": 4.0219,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736696316669102e-05,
+      "loss": 4.039,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73585772191805e-05,
+      "loss": 4.0316,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735019127166998e-05,
+      "loss": 4.0295,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734180532415946e-05,
+      "loss": 4.0392,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733341937664894e-05,
+      "loss": 4.0473,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732503342913842e-05,
+      "loss": 4.0341,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7316663860431626e-05,
+      "loss": 4.0306,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7308277912921106e-05,
+      "loss": 4.0331,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7299891965410586e-05,
+      "loss": 4.0332,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7291506017900066e-05,
+      "loss": 4.0317,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7283120070389546e-05,
+      "loss": 4.0157,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7274734122879026e-05,
+      "loss": 4.0359,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726636455417224e-05,
+      "loss": 4.0357,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725797860666172e-05,
+      "loss": 4.0201,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.72495926591512e-05,
+      "loss": 4.0262,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724120671164068e-05,
+      "loss": 4.0348,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723283714293389e-05,
+      "loss": 4.032,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722445119542337e-05,
+      "loss": 4.042,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.721606524791285e-05,
+      "loss": 4.0329,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720767930040233e-05,
+      "loss": 4.0257,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719932611049927e-05,
+      "loss": 4.0471,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719094016298875e-05,
+      "loss": 4.0116,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7182554215478236e-05,
+      "loss": 4.0191,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7174168267967716e-05,
+      "loss": 4.0245,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7165782320457196e-05,
+      "loss": 4.0214,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7157396372946676e-05,
+      "loss": 4.0277,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7149010425436155e-05,
+      "loss": 4.0265,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7140624477925635e-05,
+      "loss": 4.0299,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7132238530415115e-05,
+      "loss": 4.0263,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7123852582904595e-05,
+      "loss": 4.0216,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7115466635394075e-05,
+      "loss": 4.0322,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.710708068788355e-05,
+      "loss": 4.0322,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7098711119176764e-05,
+      "loss": 4.0073,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7090325171666244e-05,
+      "loss": 4.0325,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7081939224155724e-05,
+      "loss": 4.0318,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7073553276645204e-05,
+      "loss": 4.0412,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7065167329134684e-05,
+      "loss": 4.0226,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7056781381624164e-05,
+      "loss": 4.023,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7048395434113644e-05,
+      "loss": 4.0181,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7040009486603124e-05,
+      "loss": 4.0187,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703163991789633e-05,
+      "loss": 4.0324,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702327034918955e-05,
+      "loss": 4.0139,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.701488440167902e-05,
+      "loss": 4.0288,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70064984541685e-05,
+      "loss": 4.0259,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699812888546172e-05,
+      "loss": 4.0193,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69897429379512e-05,
+      "loss": 4.0268,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698135699044067e-05,
+      "loss": 4.0151,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697297104293016e-05,
+      "loss": 4.0179,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696458509541964e-05,
+      "loss": 4.0223,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695619914790912e-05,
+      "loss": 4.0329,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69478132003986e-05,
+      "loss": 4.0129,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693942725288808e-05,
+      "loss": 4.0275,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693104130537756e-05,
+      "loss": 4.005,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.692265535786704e-05,
+      "loss": 4.0239,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691426941035652e-05,
+      "loss": 4.025,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.690589984164973e-05,
+      "loss": 4.0317,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689751389413921e-05,
+      "loss": 4.0232,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6889127946628687e-05,
+      "loss": 4.0238,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6880741999118167e-05,
+      "loss": 4.0432,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6872356051607646e-05,
+      "loss": 4.0158,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6863986482900856e-05,
+      "loss": 4.0345,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685561691419407e-05,
+      "loss": 4.0176,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684724734548728e-05,
+      "loss": 4.0213,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683886139797676e-05,
+      "loss": 4.0258,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683047545046624e-05,
+      "loss": 4.0182,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682208950295572e-05,
+      "loss": 4.0313,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68137035554452e-05,
+      "loss": 4.0358,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680531760793468e-05,
+      "loss": 4.0183,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679693166042416e-05,
+      "loss": 4.0164,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678854571291364e-05,
+      "loss": 4.02,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678015976540312e-05,
+      "loss": 4.0293,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.67717738178926e-05,
+      "loss": 4.0177,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676338787038208e-05,
+      "loss": 4.0292,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675500192287156e-05,
+      "loss": 4.0219,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6746632354164776e-05,
+      "loss": 4.0225,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6738246406654256e-05,
+      "loss": 4.0181,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6729860459143736e-05,
+      "loss": 4.0071,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6721490890436945e-05,
+      "loss": 4.0122,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6713104942926425e-05,
+      "loss": 4.0253,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6704718995415905e-05,
+      "loss": 4.0279,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6696333047905385e-05,
+      "loss": 4.0126,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6687963479198594e-05,
+      "loss": 4.0109,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6679577531688074e-05,
+      "loss": 4.0174,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6671191584177554e-05,
+      "loss": 4.0259,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6662805636667034e-05,
+      "loss": 4.0274,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6654419689156514e-05,
+      "loss": 4.0129,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6646033741645994e-05,
+      "loss": 4.0231,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6637647794135474e-05,
+      "loss": 4.0163,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6629261846624954e-05,
+      "loss": 4.0153,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6620875899114434e-05,
+      "loss": 4.0106,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.661250633040764e-05,
+      "loss": 4.0164,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.660412038289712e-05,
+      "loss": 4.0188,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65957344353866e-05,
+      "loss": 4.0118,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.658734848787608e-05,
+      "loss": 4.0176,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657896254036556e-05,
+      "loss": 4.0116,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657057659285504e-05,
+      "loss": 4.0153,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656219064534452e-05,
+      "loss": 4.0164,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6553804697834e-05,
+      "loss": 4.0147,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654543512912721e-05,
+      "loss": 4.0225,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65370491816167e-05,
+      "loss": 4.0334,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652866323410618e-05,
+      "loss": 4.0227,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652029366539939e-05,
+      "loss": 4.0255,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651190771788887e-05,
+      "loss": 4.0211,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650352177037835e-05,
+      "loss": 4.0129,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649513582286783e-05,
+      "loss": 4.0264,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6486766254161036e-05,
+      "loss": 4.0092,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6478380306650516e-05,
+      "loss": 4.0124,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6469994359139996e-05,
+      "loss": 4.0203,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6461608411629476e-05,
+      "loss": 4.0165,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6453222464118956e-05,
+      "loss": 4.0197,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6444852895412165e-05,
+      "loss": 4.018,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643648332670538e-05,
+      "loss": 4.0206,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642809737919486e-05,
+      "loss": 4.016,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641971143168434e-05,
+      "loss": 4.0229,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641132548417382e-05,
+      "loss": 4.0175,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64029395366633e-05,
+      "loss": 4.0094,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.639455358915278e-05,
+      "loss": 3.9939,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.638616764164226e-05,
+      "loss": 4.0272,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637779807293547e-05,
+      "loss": 4.0202,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636941212542495e-05,
+      "loss": 4.0316,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636102617791443e-05,
+      "loss": 4.0174,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635264023040391e-05,
+      "loss": 4.0219,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634425428289339e-05,
+      "loss": 4.0189,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633586833538287e-05,
+      "loss": 4.0331,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632748238787235e-05,
+      "loss": 4.0067,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631909644036183e-05,
+      "loss": 4.0165,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6310726871655045e-05,
+      "loss": 4.0186,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630234092414452e-05,
+      "loss": 4.0229,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6293954976634e-05,
+      "loss": 4.0127,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628556902912348e-05,
+      "loss": 4.0131,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627718308161296e-05,
+      "loss": 4.0146,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626879713410244e-05,
+      "loss": 4.0243,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626041118659192e-05,
+      "loss": 4.0188,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.067676067352295,
+      "eval_runtime": 300.628,
+      "eval_samples_per_second": 1269.313,
+      "eval_steps_per_second": 39.667,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.62520252390814e-05,
+      "loss": 4.0119,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.624365567037461e-05,
+      "loss": 4.0187,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.623526972286409e-05,
+      "loss": 4.0165,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622688377535357e-05,
+      "loss": 4.0157,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6218497827843054e-05,
+      "loss": 4.0234,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6210111880332534e-05,
+      "loss": 4.0148,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620174231162574e-05,
+      "loss": 4.0158,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.619337274291895e-05,
+      "loss": 4.0043,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.618498679540843e-05,
+      "loss": 4.0114,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617660084789791e-05,
+      "loss": 4.0198,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616821490038739e-05,
+      "loss": 4.022,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61598453316806e-05,
+      "loss": 4.0155,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615147576297381e-05,
+      "loss": 4.0271,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.614308981546329e-05,
+      "loss": 4.0051,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.613470386795278e-05,
+      "loss": 4.0107,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612631792044226e-05,
+      "loss": 4.003,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611793197293174e-05,
+      "loss": 4.0184,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610954602542122e-05,
+      "loss": 4.01,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61011600779107e-05,
+      "loss": 4.0114,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609277413040018e-05,
+      "loss": 4.0182,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6084404561693386e-05,
+      "loss": 4.0273,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6076018614182866e-05,
+      "loss": 4.0165,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6067632666672346e-05,
+      "loss": 4.0096,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6059246719161826e-05,
+      "loss": 4.0151,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6050860771651306e-05,
+      "loss": 4.015,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6042491202944515e-05,
+      "loss": 4.0105,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6034105255433995e-05,
+      "loss": 3.9977,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602573568672721e-05,
+      "loss": 4.0134,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601734973921669e-05,
+      "loss": 4.0165,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600896379170617e-05,
+      "loss": 3.9981,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600057784419565e-05,
+      "loss": 4.0101,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599219189668513e-05,
+      "loss": 4.0161,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598380594917461e-05,
+      "loss": 4.0124,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597542000166409e-05,
+      "loss": 4.0227,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596703405415357e-05,
+      "loss": 4.0129,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595864810664305e-05,
+      "loss": 4.0079,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595026215913253e-05,
+      "loss": 4.0308,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5941876211622e-05,
+      "loss": 3.9898,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.593350664291522e-05,
+      "loss": 4.0013,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59251206954047e-05,
+      "loss": 4.0081,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591673474789418e-05,
+      "loss": 4.0017,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590834880038366e-05,
+      "loss": 4.0068,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589996285287314e-05,
+      "loss": 4.0133,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589157690536262e-05,
+      "loss": 4.0054,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58831909578521e-05,
+      "loss": 4.0105,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587482138914531e-05,
+      "loss": 4.0003,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586643544163479e-05,
+      "loss": 4.0118,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585804949412427e-05,
+      "loss": 4.0152,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584966354661375e-05,
+      "loss": 3.9893,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584127759910323e-05,
+      "loss": 4.0121,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583289165159271e-05,
+      "loss": 4.0142,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582452208288592e-05,
+      "loss": 4.0176,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58161361353754e-05,
+      "loss": 4.0055,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5807750187864884e-05,
+      "loss": 4.0024,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5799364240354364e-05,
+      "loss": 4.0046,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5790978292843844e-05,
+      "loss": 3.9987,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.578260872413705e-05,
+      "loss": 4.0147,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577422277662653e-05,
+      "loss": 3.9929,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.576583682911601e-05,
+      "loss": 4.0116,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.575745088160549e-05,
+      "loss": 4.0055,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574906493409497e-05,
+      "loss": 4.0041,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574067898658445e-05,
+      "loss": 4.0033,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573229303907393e-05,
+      "loss": 4.0025,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572392347036714e-05,
+      "loss": 4.0009,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571553752285662e-05,
+      "loss": 3.9942,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57071515753461e-05,
+      "loss": 4.0202,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569876562783558e-05,
+      "loss": 3.9999,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569037968032506e-05,
+      "loss": 3.9998,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568201011161828e-05,
+      "loss": 3.9893,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567362416410776e-05,
+      "loss": 4.0017,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566523821659724e-05,
+      "loss": 4.0068,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565685226908672e-05,
+      "loss": 4.0121,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564846632157619e-05,
+      "loss": 4.0062,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564008037406567e-05,
+      "loss": 4.0063,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563169442655515e-05,
+      "loss": 4.0214,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562330847904463e-05,
+      "loss": 4.0006,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561493891033784e-05,
+      "loss": 4.0159,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.560655296282732e-05,
+      "loss": 3.9983,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55981670153168e-05,
+      "loss": 4.0026,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5589781067806286e-05,
+      "loss": 4.0062,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5581395120295766e-05,
+      "loss": 3.9996,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5573009172785246e-05,
+      "loss": 4.0149,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5564623225274726e-05,
+      "loss": 4.0192,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5556237277764206e-05,
+      "loss": 4.0047,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5547867709057415e-05,
+      "loss": 3.9928,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5539481761546895e-05,
+      "loss": 4.0006,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5531095814036375e-05,
+      "loss": 4.01,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5522709866525855e-05,
+      "loss": 4.0001,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551435667662279e-05,
+      "loss": 4.009,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550597072911227e-05,
+      "loss": 4.0046,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549758478160175e-05,
+      "loss": 4.0042,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548919883409124e-05,
+      "loss": 4.0002,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548081288658072e-05,
+      "loss": 3.9896,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54724269390702e-05,
+      "loss": 3.9959,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.546405737036341e-05,
+      "loss": 4.0016,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.545567142285289e-05,
+      "loss": 4.0159,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.544728547534237e-05,
+      "loss": 3.9917,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543889952783185e-05,
+      "loss": 3.9955,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543051358032133e-05,
+      "loss": 4.0026,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.542212763281081e-05,
+      "loss": 4.0036,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.541375806410402e-05,
+      "loss": 4.0086,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54053721165935e-05,
+      "loss": 3.9967,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539698616908298e-05,
+      "loss": 4.0081,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538860022157246e-05,
+      "loss": 3.997,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538023065286567e-05,
+      "loss": 3.998,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537184470535515e-05,
+      "loss": 3.9938,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536347513664836e-05,
+      "loss": 3.9972,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535508918913784e-05,
+      "loss": 4.0034,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534670324162732e-05,
+      "loss": 3.9955,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53383172941168e-05,
+      "loss": 3.9967,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532993134660628e-05,
+      "loss": 3.9917,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532154539909576e-05,
+      "loss": 3.9963,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531315945158524e-05,
+      "loss": 4.0009,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.530477350407472e-05,
+      "loss": 4.0003,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5296387556564195e-05,
+      "loss": 4.0035,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5288001609053675e-05,
+      "loss": 4.0078,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527961566154316e-05,
+      "loss": 4.0079,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527124609283637e-05,
+      "loss": 4.0077,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526286014532585e-05,
+      "loss": 4.0032,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.525447419781533e-05,
+      "loss": 3.9959,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524608825030481e-05,
+      "loss": 4.0054,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5237718681598027e-05,
+      "loss": 3.9933,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52293327340875e-05,
+      "loss": 3.9972,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522094678657698e-05,
+      "loss": 4.0032,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521256083906646e-05,
+      "loss": 4.0002,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520417489155594e-05,
+      "loss": 4.0021,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519580532284915e-05,
+      "loss": 3.9971,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518741937533863e-05,
+      "loss": 4.003,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517903342782811e-05,
+      "loss": 3.9974,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5170647480317595e-05,
+      "loss": 4.0058,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5162261532807075e-05,
+      "loss": 3.9998,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5153875585296555e-05,
+      "loss": 3.9962,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5145506016589764e-05,
+      "loss": 3.9723,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5137120069079244e-05,
+      "loss": 4.0069,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5128734121568724e-05,
+      "loss": 4.0018,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5120348174058204e-05,
+      "loss": 4.0145,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5111962226547684e-05,
+      "loss": 4.0026,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5103576279037164e-05,
+      "loss": 3.9998,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5095190331526644e-05,
+      "loss": 4.0037,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5086804384016124e-05,
+      "loss": 4.0134,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507843481530933e-05,
+      "loss": 3.9878,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507004886779881e-05,
+      "loss": 4.0005,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506166292028829e-05,
+      "loss": 3.9983,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505327697277778e-05,
+      "loss": 4.0031,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504489102526726e-05,
+      "loss": 3.9994,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.503652145656047e-05,
+      "loss": 3.9986,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502815188785368e-05,
+      "loss": 3.9956,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501976594034316e-05,
+      "loss": 4.007,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501137999283264e-05,
+      "loss": 3.9979,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.057558536529541,
+      "eval_runtime": 302.2208,
+      "eval_samples_per_second": 1262.623,
+      "eval_steps_per_second": 39.458,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.500299404532212e-05,
+      "loss": 4.0096,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.49946080978116e-05,
+      "loss": 4.0005,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498622215030108e-05,
+      "loss": 3.9953,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497783620279056e-05,
+      "loss": 4.0026,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496945025528003e-05,
+      "loss": 4.0016,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496106430776952e-05,
+      "loss": 4.0033,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4952678360259e-05,
+      "loss": 3.997,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494429241274848e-05,
+      "loss": 3.9868,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493590646523796e-05,
+      "loss": 3.9918,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492752051772744e-05,
+      "loss": 4.0096,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491913457021692e-05,
+      "loss": 4.0004,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.49107486227064e-05,
+      "loss": 3.9943,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4902379053999606e-05,
+      "loss": 4.0128,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4893993106489086e-05,
+      "loss": 3.9862,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4885607158978566e-05,
+      "loss": 3.994,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4877221211468046e-05,
+      "loss": 3.9864,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4868835263957526e-05,
+      "loss": 3.9996,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4860449316447006e-05,
+      "loss": 3.9951,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4852063368936486e-05,
+      "loss": 3.9952,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4843677421425966e-05,
+      "loss": 4.0006,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4835291473915446e-05,
+      "loss": 4.0059,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4826905526404926e-05,
+      "loss": 4.0026,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4818519578894406e-05,
+      "loss": 3.9902,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4810133631383886e-05,
+      "loss": 3.9973,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4801747683873366e-05,
+      "loss": 3.9999,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4793361736362846e-05,
+      "loss": 3.9944,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4784975788852326e-05,
+      "loss": 3.983,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4776589841341806e-05,
+      "loss": 3.9934,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4768220272635015e-05,
+      "loss": 4.0009,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4759834325124495e-05,
+      "loss": 3.9817,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4751448377613975e-05,
+      "loss": 3.9926,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4743078808907184e-05,
+      "loss": 3.9948,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4734692861396664e-05,
+      "loss": 3.9977,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.472630691388615e-05,
+      "loss": 4.0003,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.471792096637563e-05,
+      "loss": 3.9958,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470955139766884e-05,
+      "loss": 3.9982,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470116545015832e-05,
+      "loss": 4.0102,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46927795026478e-05,
+      "loss": 3.9722,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468439355513728e-05,
+      "loss": 3.9847,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467600760762676e-05,
+      "loss": 3.9925,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466762166011624e-05,
+      "loss": 3.9807,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465923571260572e-05,
+      "loss": 3.9936,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46508497650952e-05,
+      "loss": 3.9937,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464249657519214e-05,
+      "loss": 3.9899,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463411062768162e-05,
+      "loss": 3.9917,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4625724680171104e-05,
+      "loss": 3.9869,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461735511146431e-05,
+      "loss": 3.9921,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460896916395379e-05,
+      "loss": 4.0035,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460058321644327e-05,
+      "loss": 3.9687,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459219726893275e-05,
+      "loss": 3.9927,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.458381132142223e-05,
+      "loss": 3.9999,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.457542537391171e-05,
+      "loss": 4.0007,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.456703942640119e-05,
+      "loss": 3.991,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455865347889067e-05,
+      "loss": 3.9829,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455026753138015e-05,
+      "loss": 3.9886,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.454188158386963e-05,
+      "loss": 3.9811,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4533495636359106e-05,
+      "loss": 3.9959,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4525109688848586e-05,
+      "loss": 3.978,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.451672374133807e-05,
+      "loss": 3.9971,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450833779382755e-05,
+      "loss": 3.9855,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449996822512076e-05,
+      "loss": 3.989,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449158227761024e-05,
+      "loss": 3.9829,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448319633009972e-05,
+      "loss": 3.9895,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44748103825892e-05,
+      "loss": 3.9836,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.446644081388241e-05,
+      "loss": 3.979,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445805486637189e-05,
+      "loss": 4.0011,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444966891886137e-05,
+      "loss": 3.9879,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444129935015458e-05,
+      "loss": 3.9818,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443291340264406e-05,
+      "loss": 3.973,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442452745513354e-05,
+      "loss": 3.9894,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4416141507623026e-05,
+      "loss": 3.9859,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4407771938916235e-05,
+      "loss": 4.0001,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4399385991405715e-05,
+      "loss": 3.9839,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4391000043895195e-05,
+      "loss": 3.9941,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4382614096384675e-05,
+      "loss": 4.0001,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4374228148874155e-05,
+      "loss": 3.9846,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4365842201363635e-05,
+      "loss": 3.998,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4357456253853115e-05,
+      "loss": 3.9852,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4349070306342595e-05,
+      "loss": 3.986,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4340700737635804e-05,
+      "loss": 3.9916,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4332314790125284e-05,
+      "loss": 3.9863,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4323928842614764e-05,
+      "loss": 3.9897,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4315542895104244e-05,
+      "loss": 4.0073,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4307156947593724e-05,
+      "loss": 3.9898,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429877100008321e-05,
+      "loss": 3.9762,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429040143137642e-05,
+      "loss": 3.9852,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.42820154838659e-05,
+      "loss": 3.9904,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427362953635538e-05,
+      "loss": 3.9895,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.426524358884486e-05,
+      "loss": 3.9892,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425685764133434e-05,
+      "loss": 3.9897,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424848807262755e-05,
+      "loss": 3.9888,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424010212511703e-05,
+      "loss": 3.9824,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423171617760651e-05,
+      "loss": 3.9711,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.422333023009599e-05,
+      "loss": 3.979,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421494428258547e-05,
+      "loss": 3.9834,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420657471387868e-05,
+      "loss": 4.0021,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419818876636816e-05,
+      "loss": 3.9751,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4189819197661374e-05,
+      "loss": 3.9775,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4181433250150854e-05,
+      "loss": 3.9862,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4173047302640333e-05,
+      "loss": 3.9862,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4164661355129813e-05,
+      "loss": 3.9948,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4156275407619293e-05,
+      "loss": 3.9812,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4147889460108767e-05,
+      "loss": 3.986,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4139503512598247e-05,
+      "loss": 3.9825,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413113394389146e-05,
+      "loss": 3.9805,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.412274799638094e-05,
+      "loss": 3.9784,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.411437842767415e-05,
+      "loss": 3.9841,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.410599248016363e-05,
+      "loss": 3.9861,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409760653265311e-05,
+      "loss": 3.9801,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40892205851426e-05,
+      "loss": 3.9792,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408083463763207e-05,
+      "loss": 3.9776,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407244869012155e-05,
+      "loss": 3.9811,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406407912141477e-05,
+      "loss": 3.9828,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405569317390424e-05,
+      "loss": 3.982,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404730722639372e-05,
+      "loss": 3.987,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40389212788832e-05,
+      "loss": 3.9882,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403053533137268e-05,
+      "loss": 3.9947,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402216576266589e-05,
+      "loss": 3.99,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401377981515537e-05,
+      "loss": 3.9881,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400539386764485e-05,
+      "loss": 3.9855,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3997007920134336e-05,
+      "loss": 3.986,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3988621972623816e-05,
+      "loss": 3.9773,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3980236025113296e-05,
+      "loss": 3.983,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3971850077602776e-05,
+      "loss": 3.9862,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3963464130092256e-05,
+      "loss": 3.9848,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3955078182581736e-05,
+      "loss": 3.9881,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3946692235071216e-05,
+      "loss": 3.982,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3938322666364425e-05,
+      "loss": 3.9843,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3929936718853905e-05,
+      "loss": 3.9787,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3921550771343385e-05,
+      "loss": 3.9932,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3913164823832865e-05,
+      "loss": 3.9827,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3904778876322345e-05,
+      "loss": 3.9801,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3896392928811825e-05,
+      "loss": 3.9578,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3888006981301304e-05,
+      "loss": 3.987,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3879621033790784e-05,
+      "loss": 3.9912,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3871251465084e-05,
+      "loss": 3.9979,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.386286551757348e-05,
+      "loss": 3.9862,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.385449594886669e-05,
+      "loss": 3.9809,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.384611000135617e-05,
+      "loss": 3.9883,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383772405384565e-05,
+      "loss": 3.9978,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382933810633513e-05,
+      "loss": 3.9731,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38209521588246e-05,
+      "loss": 3.9877,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.381256621131408e-05,
+      "loss": 3.9822,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380418026380356e-05,
+      "loss": 3.9877,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379579431629304e-05,
+      "loss": 3.9812,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378742474758626e-05,
+      "loss": 3.9807,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377903880007574e-05,
+      "loss": 3.98,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377065285256522e-05,
+      "loss": 3.9899,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.37622669050547e-05,
+      "loss": 3.9875,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0478081703186035,
+      "eval_runtime": 293.2447,
+      "eval_samples_per_second": 1301.272,
+      "eval_steps_per_second": 40.666,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.375388095754418e-05,
+      "loss": 3.9752,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.374549501003366e-05,
+      "loss": 3.9873,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.373710906252314e-05,
+      "loss": 3.9779,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372872311501262e-05,
+      "loss": 3.9897,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372035354630583e-05,
+      "loss": 3.9891,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.371196759879531e-05,
+      "loss": 3.9834,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3703598030088516e-05,
+      "loss": 3.9821,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3695212082577996e-05,
+      "loss": 3.9692,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3686826135067476e-05,
+      "loss": 3.9811,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3678440187556956e-05,
+      "loss": 3.9895,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3670054240046436e-05,
+      "loss": 3.9834,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366168467133965e-05,
+      "loss": 3.9848,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365329872382913e-05,
+      "loss": 3.9933,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.364491277631861e-05,
+      "loss": 3.9752,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363652682880809e-05,
+      "loss": 3.978,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362814088129757e-05,
+      "loss": 3.973,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361975493378705e-05,
+      "loss": 3.9824,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361138536508026e-05,
+      "loss": 3.9792,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.360299941756974e-05,
+      "loss": 3.977,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.359461347005922e-05,
+      "loss": 3.9847,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35862275225487e-05,
+      "loss": 3.9925,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357784157503818e-05,
+      "loss": 3.9845,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356945562752766e-05,
+      "loss": 3.9777,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356106968001714e-05,
+      "loss": 3.9776,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.355268373250662e-05,
+      "loss": 3.9886,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3544314163799836e-05,
+      "loss": 3.9775,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3535928216289316e-05,
+      "loss": 3.9664,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.352754226877879e-05,
+      "loss": 3.9777,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351915632126827e-05,
+      "loss": 3.9779,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351077037375775e-05,
+      "loss": 3.9687,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3502400805050965e-05,
+      "loss": 3.978,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.349401485754044e-05,
+      "loss": 3.9786,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.348562891002992e-05,
+      "loss": 3.9838,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34772429625194e-05,
+      "loss": 3.9855,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3468873393812614e-05,
+      "loss": 3.9844,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346050382510583e-05,
+      "loss": 3.9825,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.345211787759531e-05,
+      "loss": 3.9931,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.344373193008479e-05,
+      "loss": 3.9599,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.343534598257426e-05,
+      "loss": 3.9725,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.342696003506374e-05,
+      "loss": 3.974,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341857408755322e-05,
+      "loss": 3.9609,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34101881400427e-05,
+      "loss": 3.9823,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.340180219253218e-05,
+      "loss": 3.9785,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339341624502166e-05,
+      "loss": 3.9755,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.338504667631487e-05,
+      "loss": 3.9736,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.337667710760808e-05,
+      "loss": 3.9758,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336829116009757e-05,
+      "loss": 3.9767,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3359921591390784e-05,
+      "loss": 3.9865,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3351535643880263e-05,
+      "loss": 3.9567,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334314969636974e-05,
+      "loss": 3.9709,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3334763748859217e-05,
+      "loss": 3.9903,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3326377801348697e-05,
+      "loss": 3.9859,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3317991853838177e-05,
+      "loss": 3.9754,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3309605906327656e-05,
+      "loss": 3.967,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3301219958817136e-05,
+      "loss": 3.9747,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3292834011306616e-05,
+      "loss": 3.9644,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3284448063796096e-05,
+      "loss": 3.9835,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3276062116285576e-05,
+      "loss": 3.9638,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3267676168775056e-05,
+      "loss": 3.9777,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3259322978872e-05,
+      "loss": 3.9735,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325093703136148e-05,
+      "loss": 3.9763,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.324255108385096e-05,
+      "loss": 3.9661,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.323416513634044e-05,
+      "loss": 3.9768,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.322577918882992e-05,
+      "loss": 3.9653,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32173932413194e-05,
+      "loss": 3.9627,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320900729380888e-05,
+      "loss": 3.9849,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320063772510209e-05,
+      "loss": 3.9751,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319225177759157e-05,
+      "loss": 3.9663,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318386583008105e-05,
+      "loss": 3.9583,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317547988257053e-05,
+      "loss": 3.9704,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316709393506001e-05,
+      "loss": 3.9727,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315870798754949e-05,
+      "loss": 3.9897,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315032204003897e-05,
+      "loss": 3.9679,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.314193609252845e-05,
+      "loss": 3.9806,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3133566523821666e-05,
+      "loss": 3.9861,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3125180576311146e-05,
+      "loss": 3.9699,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3116794628800626e-05,
+      "loss": 3.9785,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.31084086812901e-05,
+      "loss": 3.9761,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3100039112583315e-05,
+      "loss": 3.9744,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3091653165072795e-05,
+      "loss": 3.9719,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3083267217562275e-05,
+      "loss": 3.9703,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.307488127005175e-05,
+      "loss": 3.9788,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.306649532254123e-05,
+      "loss": 3.9919,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3058125753834444e-05,
+      "loss": 3.9762,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3049739806323923e-05,
+      "loss": 3.9598,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304137023761714e-05,
+      "loss": 3.9712,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.303298429010662e-05,
+      "loss": 3.977,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30245983425961e-05,
+      "loss": 3.9758,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.301621239508557e-05,
+      "loss": 3.9776,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.300784282637879e-05,
+      "loss": 3.971,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299945687886827e-05,
+      "loss": 3.978,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299107093135775e-05,
+      "loss": 3.9652,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.298268498384722e-05,
+      "loss": 3.959,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29742990363367e-05,
+      "loss": 3.9662,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.296591308882618e-05,
+      "loss": 3.9652,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295752714131566e-05,
+      "loss": 3.9835,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294914119380514e-05,
+      "loss": 3.9686,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294075524629462e-05,
+      "loss": 3.9627,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.293238567758784e-05,
+      "loss": 3.9752,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.292399973007732e-05,
+      "loss": 3.9703,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29156137825668e-05,
+      "loss": 3.9784,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2907244213860006e-05,
+      "loss": 3.9706,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2898858266349486e-05,
+      "loss": 3.9708,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2890472318838966e-05,
+      "loss": 3.9677,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2882086371328446e-05,
+      "loss": 3.9731,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2873700423817926e-05,
+      "loss": 3.9635,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2865347233914864e-05,
+      "loss": 3.9674,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2856961286404344e-05,
+      "loss": 3.9736,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284857533889383e-05,
+      "loss": 3.9665,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284018939138331e-05,
+      "loss": 3.9638,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.283180344387279e-05,
+      "loss": 3.9622,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2823433875166e-05,
+      "loss": 3.9676,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.281504792765548e-05,
+      "loss": 3.9692,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.280666198014496e-05,
+      "loss": 3.9686,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279827603263444e-05,
+      "loss": 3.9704,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278989008512392e-05,
+      "loss": 3.9775,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.27815041376134e-05,
+      "loss": 3.9787,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.277313456890661e-05,
+      "loss": 3.9756,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.276474862139609e-05,
+      "loss": 3.9754,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.275636267388557e-05,
+      "loss": 3.9701,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.274797672637505e-05,
+      "loss": 3.9725,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273959077886453e-05,
+      "loss": 3.9632,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2731204831354015e-05,
+      "loss": 3.9714,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2722818883843495e-05,
+      "loss": 3.9642,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2714432936332975e-05,
+      "loss": 3.9717,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2706063367626184e-05,
+      "loss": 3.9773,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2697677420115664e-05,
+      "loss": 3.9728,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2689291472605144e-05,
+      "loss": 3.9696,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2680905525094624e-05,
+      "loss": 3.9635,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2672519577584104e-05,
+      "loss": 3.9799,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.266413363007358e-05,
+      "loss": 3.9706,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.265574768256306e-05,
+      "loss": 3.9662,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264736173505254e-05,
+      "loss": 3.9449,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263899216634575e-05,
+      "loss": 3.9701,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263060621883523e-05,
+      "loss": 3.9771,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.262222027132471e-05,
+      "loss": 3.9832,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.261385070261793e-05,
+      "loss": 3.9722,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.260546475510741e-05,
+      "loss": 3.9658,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.259707880759688e-05,
+      "loss": 3.9712,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258869286008636e-05,
+      "loss": 3.9872,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258030691257584e-05,
+      "loss": 3.9624,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.257192096506532e-05,
+      "loss": 3.9732,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25635350175548e-05,
+      "loss": 3.9701,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.255514907004428e-05,
+      "loss": 3.9725,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.254677950133749e-05,
+      "loss": 3.9713,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253839355382697e-05,
+      "loss": 3.9677,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253000760631645e-05,
+      "loss": 3.9647,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252162165880594e-05,
+      "loss": 3.9766,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.251323571129542e-05,
+      "loss": 3.9694,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.040585994720459,
+      "eval_runtime": 301.3219,
+      "eval_samples_per_second": 1266.39,
+      "eval_steps_per_second": 39.576,
+      "step": 1068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2504866142588627e-05,
+      "loss": 3.9685,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2496480195078107e-05,
+      "loss": 3.9753,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2488094247567586e-05,
+      "loss": 3.9637,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2479708300057066e-05,
+      "loss": 3.9756,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2471322352546546e-05,
+      "loss": 3.973,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2462936405036026e-05,
+      "loss": 3.9753,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2454550457525506e-05,
+      "loss": 3.9659,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2446164510014986e-05,
+      "loss": 3.9564,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2437778562504466e-05,
+      "loss": 3.9646,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242939261499394e-05,
+      "loss": 3.9805,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242100666748342e-05,
+      "loss": 3.9688,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.24126207199729e-05,
+      "loss": 3.9706,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2404234772462386e-05,
+      "loss": 3.9807,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2395848824951866e-05,
+      "loss": 3.9642,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2387462877441346e-05,
+      "loss": 3.9644,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2379076929930826e-05,
+      "loss": 3.9596,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2370690982420306e-05,
+      "loss": 3.9684,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2362305034909786e-05,
+      "loss": 3.965,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2353919087399266e-05,
+      "loss": 3.964,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2345533139888746e-05,
+      "loss": 3.9699,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2337147192378226e-05,
+      "loss": 3.9835,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2328777623671435e-05,
+      "loss": 3.9706,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2320391676160915e-05,
+      "loss": 3.9662,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2312005728650395e-05,
+      "loss": 3.9667,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2303636159943604e-05,
+      "loss": 3.969,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2295250212433084e-05,
+      "loss": 3.9689,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.22868806437263e-05,
+      "loss": 3.9554,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227849469621578e-05,
+      "loss": 3.9636,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227010874870526e-05,
+      "loss": 3.966,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.226172280119474e-05,
+      "loss": 3.9573,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.225333685368422e-05,
+      "loss": 3.9606,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.22449509061737e-05,
+      "loss": 3.9724,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223656495866318e-05,
+      "loss": 3.9656,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222817901115265e-05,
+      "loss": 3.9734,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221979306364213e-05,
+      "loss": 3.9697,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221142349493535e-05,
+      "loss": 3.9676,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220303754742483e-05,
+      "loss": 3.9766,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.219465159991431e-05,
+      "loss": 3.9517,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.218626565240379e-05,
+      "loss": 3.9616,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2177896083697004e-05,
+      "loss": 3.959,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2169510136186484e-05,
+      "loss": 3.9506,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216114056747969e-05,
+      "loss": 3.9687,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.215275461996917e-05,
+      "loss": 3.9656,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.214436867245865e-05,
+      "loss": 3.9604,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2135982724948126e-05,
+      "loss": 3.9644,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2127596777437606e-05,
+      "loss": 3.9612,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2119210829927086e-05,
+      "loss": 3.9649,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2110824882416566e-05,
+      "loss": 3.9694,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2102438934906046e-05,
+      "loss": 3.9462,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2094052987395526e-05,
+      "loss": 3.9578,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2085667039885006e-05,
+      "loss": 3.9745,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207729747117822e-05,
+      "loss": 3.9718,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20689115236677e-05,
+      "loss": 3.9651,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206052557615718e-05,
+      "loss": 3.9556,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.205213962864666e-05,
+      "loss": 3.9612,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.204375368113614e-05,
+      "loss": 3.9538,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.203536773362562e-05,
+      "loss": 3.9657,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20269817861151e-05,
+      "loss": 3.9525,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201859583860458e-05,
+      "loss": 3.967,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201022626989779e-05,
+      "loss": 3.9639,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.200184032238727e-05,
+      "loss": 3.9607,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.199347075368048e-05,
+      "loss": 3.9535,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.198508480616996e-05,
+      "loss": 3.9652,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.197669885865944e-05,
+      "loss": 3.9515,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1968312911148926e-05,
+      "loss": 3.9541,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1959926963638406e-05,
+      "loss": 3.9687,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1951541016127886e-05,
+      "loss": 3.9627,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1943155068617366e-05,
+      "loss": 3.9572,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1934769121106846e-05,
+      "loss": 3.9435,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.192638317359632e-05,
+      "loss": 3.9599,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1918013604889535e-05,
+      "loss": 3.96,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1909644036182744e-05,
+      "loss": 3.972,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1901258088672224e-05,
+      "loss": 3.9566,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1892872141161704e-05,
+      "loss": 3.9702,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1884486193651184e-05,
+      "loss": 3.9677,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1876100246140664e-05,
+      "loss": 3.962,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.186773067743388e-05,
+      "loss": 3.9651,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185934472992336e-05,
+      "loss": 3.9598,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185095878241284e-05,
+      "loss": 3.9668,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.184257283490232e-05,
+      "loss": 3.9577,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.183418688739179e-05,
+      "loss": 3.9591,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.182580093988127e-05,
+      "loss": 3.9649,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181741499237075e-05,
+      "loss": 3.9759,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180902904486023e-05,
+      "loss": 3.9632,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180064309734971e-05,
+      "loss": 3.9488,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.179227352864292e-05,
+      "loss": 3.9614,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.17838875811324e-05,
+      "loss": 3.9616,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.177550163362188e-05,
+      "loss": 3.9638,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.176711568611136e-05,
+      "loss": 3.9658,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.175874611740458e-05,
+      "loss": 3.9595,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.175036016989406e-05,
+      "loss": 3.9622,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.174197422238354e-05,
+      "loss": 3.9544,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.173358827487302e-05,
+      "loss": 3.9479,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.172521870616623e-05,
+      "loss": 3.9506,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1716849137459436e-05,
+      "loss": 3.9505,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1708463189948916e-05,
+      "loss": 3.9733,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1700077242438396e-05,
+      "loss": 3.9551,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1691691294927876e-05,
+      "loss": 3.9489,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1683305347417356e-05,
+      "loss": 3.9656,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1674919399906836e-05,
+      "loss": 3.9577,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1666533452396315e-05,
+      "loss": 3.9673,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16581475048858e-05,
+      "loss": 3.9571,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164977793617901e-05,
+      "loss": 3.9547,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164140836747222e-05,
+      "loss": 3.9575,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16330224199617e-05,
+      "loss": 3.9614,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.162463647245118e-05,
+      "loss": 3.9478,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.161626690374439e-05,
+      "loss": 3.9583,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160788095623387e-05,
+      "loss": 3.9589,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159949500872335e-05,
+      "loss": 3.9561,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159110906121283e-05,
+      "loss": 3.9528,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.158272311370231e-05,
+      "loss": 3.9532,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157433716619179e-05,
+      "loss": 3.9518,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.156595121868127e-05,
+      "loss": 3.9562,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1557565271170756e-05,
+      "loss": 3.9583,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1549179323660236e-05,
+      "loss": 3.9577,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1540793376149716e-05,
+      "loss": 3.9643,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1532423807442925e-05,
+      "loss": 3.9694,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1524037859932405e-05,
+      "loss": 3.9622,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1515651912421885e-05,
+      "loss": 3.9609,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1507265964911365e-05,
+      "loss": 3.9612,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1498880017400845e-05,
+      "loss": 3.9589,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1490494069890325e-05,
+      "loss": 3.9477,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1482124501183534e-05,
+      "loss": 3.9608,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1473738553673014e-05,
+      "loss": 3.9549,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1465352606162494e-05,
+      "loss": 3.9585,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1456966658651974e-05,
+      "loss": 3.9605,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1448580711141454e-05,
+      "loss": 3.964,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1440194763630934e-05,
+      "loss": 3.9544,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.143182519492415e-05,
+      "loss": 3.9501,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.142343924741362e-05,
+      "loss": 3.9678,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.14150532999031e-05,
+      "loss": 3.9619,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.140666735239258e-05,
+      "loss": 3.9523,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.139828140488206e-05,
+      "loss": 3.935,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138989545737154e-05,
+      "loss": 3.9554,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138150950986102e-05,
+      "loss": 3.9675,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.13731235623505e-05,
+      "loss": 3.9713,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.136475399364371e-05,
+      "loss": 3.96,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.135636804613319e-05,
+      "loss": 3.9535,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.134799847742641e-05,
+      "loss": 3.9587,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133961252991589e-05,
+      "loss": 3.9735,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133122658240537e-05,
+      "loss": 3.952,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.132284063489485e-05,
+      "loss": 3.9571,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.131445468738433e-05,
+      "loss": 3.9603,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1306085118677536e-05,
+      "loss": 3.9601,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1297699171167016e-05,
+      "loss": 3.9568,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1289313223656496e-05,
+      "loss": 3.9566,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1280927276145976e-05,
+      "loss": 3.9486,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1272541328635456e-05,
+      "loss": 3.9694,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1264155381124936e-05,
+      "loss": 3.9548,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.034160137176514,
+      "eval_runtime": 325.6355,
+      "eval_samples_per_second": 1171.835,
+      "eval_steps_per_second": 36.621,
+      "step": 1144800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1255769433614416e-05,
+      "loss": 3.9588,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1247383486103896e-05,
+      "loss": 3.9629,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1238997538593376e-05,
+      "loss": 3.9491,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1230611591082856e-05,
+      "loss": 3.9654,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.122222564357234e-05,
+      "loss": 3.9616,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1213839696061816e-05,
+      "loss": 3.9671,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1205453748551296e-05,
+      "loss": 3.9529,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1197067801040776e-05,
+      "loss": 3.9442,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1188681853530256e-05,
+      "loss": 3.9468,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1180295906019735e-05,
+      "loss": 3.9728,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1171909958509215e-05,
+      "loss": 3.9544,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1163524010998695e-05,
+      "loss": 3.9592,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1155138063488175e-05,
+      "loss": 3.9646,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1146752115977655e-05,
+      "loss": 3.9581,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1138366168467135e-05,
+      "loss": 3.95,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1129996599760344e-05,
+      "loss": 3.9518,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1121610652249824e-05,
+      "loss": 3.9501,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.111322470473931e-05,
+      "loss": 3.9555,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.110483875722879e-05,
+      "loss": 3.9525,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.109645280971827e-05,
+      "loss": 3.9604,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.108806686220775e-05,
+      "loss": 3.9748,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1079680914697224e-05,
+      "loss": 3.9578,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1071294967186704e-05,
+      "loss": 3.9543,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1062909019676184e-05,
+      "loss": 3.9529,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1054523072165664e-05,
+      "loss": 3.9607,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.104615350345887e-05,
+      "loss": 3.9537,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.103776755594835e-05,
+      "loss": 3.9437,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.102938160843783e-05,
+      "loss": 3.9472,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.102099566092731e-05,
+      "loss": 3.955,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.101260971341679e-05,
+      "loss": 3.9459,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.100422376590627e-05,
+      "loss": 3.9462,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.099583781839576e-05,
+      "loss": 3.9607,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.098745187088524e-05,
+      "loss": 3.9548,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.097908230217845e-05,
+      "loss": 3.9579,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.097069635466793e-05,
+      "loss": 3.9634,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.096232678596114e-05,
+      "loss": 3.9561,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.095394083845062e-05,
+      "loss": 3.9596,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.09455548909401e-05,
+      "loss": 3.9443,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.093716894342958e-05,
+      "loss": 3.9489,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092878299591906e-05,
+      "loss": 3.9447,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092039704840854e-05,
+      "loss": 3.9432,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.091201110089802e-05,
+      "loss": 3.949,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.09036251533875e-05,
+      "loss": 3.9577,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.089523920587698e-05,
+      "loss": 3.9431,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.088686963717019e-05,
+      "loss": 3.9535,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087848368965967e-05,
+      "loss": 3.9478,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087009774214915e-05,
+      "loss": 3.9533,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086172817344236e-05,
+      "loss": 3.957,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.085334222593184e-05,
+      "loss": 3.9372,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.084495627842132e-05,
+      "loss": 3.946,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.08365703309108e-05,
+      "loss": 3.9594,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.082818438340028e-05,
+      "loss": 3.9613,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.081979843588976e-05,
+      "loss": 3.9478,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.081142886718297e-05,
+      "loss": 3.9468,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.080304291967245e-05,
+      "loss": 3.9531,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.079465697216193e-05,
+      "loss": 3.9408,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.078627102465141e-05,
+      "loss": 3.9528,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.077788507714089e-05,
+      "loss": 3.9397,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076949912963037e-05,
+      "loss": 3.9566,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076111318211985e-05,
+      "loss": 3.9488,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.075272723460933e-05,
+      "loss": 3.951,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.074434128709881e-05,
+      "loss": 3.9378,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0735988097195756e-05,
+      "loss": 3.955,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0727602149685236e-05,
+      "loss": 3.9346,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071921620217471e-05,
+      "loss": 3.9461,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071083025466419e-05,
+      "loss": 3.9563,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.070244430715367e-05,
+      "loss": 3.9539,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.069405835964315e-05,
+      "loss": 3.9418,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0685672412132635e-05,
+      "loss": 3.9363,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0677286464622115e-05,
+      "loss": 3.9432,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0668916895915325e-05,
+      "loss": 3.9475,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0660530948404804e-05,
+      "loss": 3.9626,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0652145000894284e-05,
+      "loss": 3.9424,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0643759053383764e-05,
+      "loss": 3.9593,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0635389484676973e-05,
+      "loss": 3.9538,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0627003537166453e-05,
+      "loss": 3.9526,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.061863396845966e-05,
+      "loss": 3.9521,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.061024802094914e-05,
+      "loss": 3.9495,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.060186207343862e-05,
+      "loss": 3.9526,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.05934761259281e-05,
+      "loss": 3.9456,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.058509017841759e-05,
+      "loss": 3.9473,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.057670423090707e-05,
+      "loss": 3.9514,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056831828339655e-05,
+      "loss": 3.9641,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055993233588603e-05,
+      "loss": 3.9498,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055156276717924e-05,
+      "loss": 3.9439,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.054317681966872e-05,
+      "loss": 3.9431,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.05347908721582e-05,
+      "loss": 3.9519,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.052642130345141e-05,
+      "loss": 3.9495,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051803535594089e-05,
+      "loss": 3.9563,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0509649408430367e-05,
+      "loss": 3.9471,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.050126346091985e-05,
+      "loss": 3.9502,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.049289389221306e-05,
+      "loss": 3.9426,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.048450794470254e-05,
+      "loss": 3.9376,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.047612199719202e-05,
+      "loss": 3.9363,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.04677360496815e-05,
+      "loss": 3.9401,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.045935010217098e-05,
+      "loss": 3.9561,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.045096415466046e-05,
+      "loss": 3.9505,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0442578207149943e-05,
+      "loss": 3.9363,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0434192259639423e-05,
+      "loss": 3.9482,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0425806312128896e-05,
+      "loss": 3.9514,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.041743674342211e-05,
+      "loss": 3.9542,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.040905079591159e-05,
+      "loss": 3.9454,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.040066484840107e-05,
+      "loss": 3.9428,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0392278900890548e-05,
+      "loss": 3.9452,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0383909332183764e-05,
+      "loss": 3.953,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0375523384673244e-05,
+      "loss": 3.9337,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0367153815966453e-05,
+      "loss": 3.9446,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0358767868455933e-05,
+      "loss": 3.949,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0350381920945413e-05,
+      "loss": 3.9399,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0341995973434896e-05,
+      "loss": 3.9429,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.033361002592437e-05,
+      "loss": 3.9446,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.032522407841385e-05,
+      "loss": 3.939,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.031683813090333e-05,
+      "loss": 3.9465,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.030845218339281e-05,
+      "loss": 3.9457,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.030006623588229e-05,
+      "loss": 3.9422,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0291696667175502e-05,
+      "loss": 3.9559,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.028331071966498e-05,
+      "loss": 3.9552,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.027494115095819e-05,
+      "loss": 3.9502,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.026655520344767e-05,
+      "loss": 3.9522,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.025816925593715e-05,
+      "loss": 3.9459,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0249799687230367e-05,
+      "loss": 3.9477,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0241413739719843e-05,
+      "loss": 3.9417,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0233027792209323e-05,
+      "loss": 3.9425,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0224641844698803e-05,
+      "loss": 3.9463,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0216255897188283e-05,
+      "loss": 3.9458,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0207869949677763e-05,
+      "loss": 3.946,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0199484002167243e-05,
+      "loss": 3.9547,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0191098054656726e-05,
+      "loss": 3.9447,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0182712107146206e-05,
+      "loss": 3.939,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0174326159635686e-05,
+      "loss": 3.9552,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0165940212125166e-05,
+      "loss": 3.9506,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0157570643418375e-05,
+      "loss": 3.9424,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0149184695907855e-05,
+      "loss": 3.928,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0140798748397335e-05,
+      "loss": 3.9423,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0132412800886815e-05,
+      "loss": 3.9578,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.01240268533763e-05,
+      "loss": 3.9575,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.011564090586578e-05,
+      "loss": 3.9462,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.010725495835526e-05,
+      "loss": 3.942,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.009886901084473e-05,
+      "loss": 3.9489,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.009048306333421e-05,
+      "loss": 3.9603,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0082113494627427e-05,
+      "loss": 3.9421,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0073727547116907e-05,
+      "loss": 3.9463,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0065341599606384e-05,
+      "loss": 3.9461,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0056955652095864e-05,
+      "loss": 3.9475,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.004860246219281e-05,
+      "loss": 3.9447,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.004021651468229e-05,
+      "loss": 3.9497,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.003183056717177e-05,
+      "loss": 3.9381,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0023444619661252e-05,
+      "loss": 3.9548,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0015058672150732e-05,
+      "loss": 3.945,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.0006672724640205e-05,
+      "loss": 3.9407,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.02894401550293,
+      "eval_runtime": 288.1013,
+      "eval_samples_per_second": 1324.503,
+      "eval_steps_per_second": 41.392,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9998286777129685e-05,
+      "loss": 3.9514,
+      "step": 1221632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9989900829619165e-05,
+      "loss": 3.9383,
+      "step": 1222144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.998153126091238e-05,
+      "loss": 3.9535,
+      "step": 1222656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9973145313401858e-05,
+      "loss": 3.9468,
+      "step": 1223168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9964775744695073e-05,
+      "loss": 3.9556,
+      "step": 1223680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9956389797184553e-05,
+      "loss": 3.9421,
+      "step": 1224192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9948003849674027e-05,
+      "loss": 3.9328,
+      "step": 1224704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9939617902163507e-05,
+      "loss": 3.935,
+      "step": 1225216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.993123195465299e-05,
+      "loss": 3.9622,
+      "step": 1225728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.992284600714247e-05,
+      "loss": 3.9462,
+      "step": 1226240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.991446005963195e-05,
+      "loss": 3.9444,
+      "step": 1226752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.990607411212143e-05,
+      "loss": 3.9538,
+      "step": 1227264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.989770454341464e-05,
+      "loss": 3.9459,
+      "step": 1227776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.988931859590412e-05,
+      "loss": 3.9391,
+      "step": 1228288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.98809326483936e-05,
+      "loss": 3.9426,
+      "step": 1228800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9872546700883082e-05,
+      "loss": 3.9369,
+      "step": 1229312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9864160753372562e-05,
+      "loss": 3.9454,
+      "step": 1229824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9855774805862042e-05,
+      "loss": 3.9385,
+      "step": 1230336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.984742161595898e-05,
+      "loss": 3.9508,
+      "step": 1230848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.983903566844846e-05,
+      "loss": 3.9633,
+      "step": 1231360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9830649720937944e-05,
+      "loss": 3.9464,
+      "step": 1231872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9822263773427424e-05,
+      "loss": 3.939,
+      "step": 1232384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9813877825916903e-05,
+      "loss": 3.9442,
+      "step": 1232896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9805491878406383e-05,
+      "loss": 3.9523,
+      "step": 1233408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9797105930895863e-05,
+      "loss": 3.9384,
+      "step": 1233920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9788719983385343e-05,
+      "loss": 3.9343,
+      "step": 1234432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9780334035874823e-05,
+      "loss": 3.9368,
+      "step": 1234944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9771948088364303e-05,
+      "loss": 3.9506,
+      "step": 1235456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9763562140853783e-05,
+      "loss": 3.9299,
+      "step": 1235968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9755176193343267e-05,
+      "loss": 3.9359,
+      "step": 1236480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9746806624636476e-05,
+      "loss": 3.9492,
+      "step": 1236992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9738420677125956e-05,
+      "loss": 3.9465,
+      "step": 1237504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9730034729615436e-05,
+      "loss": 3.946,
+      "step": 1238016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9721648782104916e-05,
+      "loss": 3.9525,
+      "step": 1238528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9713279213398128e-05,
+      "loss": 3.9429,
+      "step": 1239040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9704893265887608e-05,
+      "loss": 3.9497,
+      "step": 1239552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9696507318377088e-05,
+      "loss": 3.9336,
+      "step": 1240064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9688121370866568e-05,
+      "loss": 3.9418,
+      "step": 1240576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.967973542335604e-05,
+      "loss": 3.9339,
+      "step": 1241088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.967134947584552e-05,
+      "loss": 3.9331,
+      "step": 1241600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9662979907138737e-05,
+      "loss": 3.9401,
+      "step": 1242112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9654593959628213e-05,
+      "loss": 3.9457,
+      "step": 1242624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9646208012117693e-05,
+      "loss": 3.9344,
+      "step": 1243136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9637822064607173e-05,
+      "loss": 3.9401,
+      "step": 1243648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.962945249590039e-05,
+      "loss": 3.938,
+      "step": 1244160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9621066548389866e-05,
+      "loss": 3.942,
+      "step": 1244672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9612680600879346e-05,
+      "loss": 3.9484,
+      "step": 1245184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9604294653368826e-05,
+      "loss": 3.9309,
+      "step": 1245696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9595908705858306e-05,
+      "loss": 3.9314,
+      "step": 1246208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9587522758347786e-05,
+      "loss": 3.9478,
+      "step": 1246720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9579136810837266e-05,
+      "loss": 3.952,
+      "step": 1247232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9570750863326746e-05,
+      "loss": 3.9371,
+      "step": 1247744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9562381294619958e-05,
+      "loss": 3.9393,
+      "step": 1248256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9553995347109438e-05,
+      "loss": 3.9427,
+      "step": 1248768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9545625778402647e-05,
+      "loss": 3.9293,
+      "step": 1249280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9537256209695863e-05,
+      "loss": 3.9413,
+      "step": 1249792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9528870262185336e-05,
+      "loss": 3.9272,
+      "step": 1250304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.952048431467482e-05,
+      "loss": 3.944,
+      "step": 1250816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.95120983671643e-05,
+      "loss": 3.9419,
+      "step": 1251328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.950371241965378e-05,
+      "loss": 3.9402,
+      "step": 1251840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.949532647214326e-05,
+      "loss": 3.9257,
+      "step": 1252352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.948694052463274e-05,
+      "loss": 3.9463,
+      "step": 1252864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.947855457712222e-05,
+      "loss": 3.9236,
+      "step": 1253376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.94701686296117e-05,
+      "loss": 3.9383,
+      "step": 1253888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.946178268210118e-05,
+      "loss": 3.9423,
+      "step": 1254400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.945339673459066e-05,
+      "loss": 3.9442,
+      "step": 1254912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.944501078708014e-05,
+      "loss": 3.9284,
+      "step": 1255424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.943664121837335e-05,
+      "loss": 3.9278,
+      "step": 1255936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.942825527086283e-05,
+      "loss": 3.9306,
+      "step": 1256448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.941986932335231e-05,
+      "loss": 3.9412,
+      "step": 1256960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.941148337584179e-05,
+      "loss": 3.9417,
+      "step": 1257472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.940309742833127e-05,
+      "loss": 3.9357,
+      "step": 1257984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9394727859624484e-05,
+      "loss": 3.9458,
+      "step": 1258496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9386341912113964e-05,
+      "loss": 3.9491,
+      "step": 1259008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9377955964603444e-05,
+      "loss": 3.9394,
+      "step": 1259520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9369586395896653e-05,
+      "loss": 3.9412,
+      "step": 1260032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9361200448386133e-05,
+      "loss": 3.9403,
+      "step": 1260544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9352814500875613e-05,
+      "loss": 3.9455,
+      "step": 1261056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9344428553365093e-05,
+      "loss": 3.932,
+      "step": 1261568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9336058984658305e-05,
+      "loss": 3.9383,
+      "step": 1262080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9327673037147785e-05,
+      "loss": 3.9396,
+      "step": 1262592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9319287089637265e-05,
+      "loss": 3.9541,
+      "step": 1263104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9310901142126745e-05,
+      "loss": 3.9426,
+      "step": 1263616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9302515194616225e-05,
+      "loss": 3.9287,
+      "step": 1264128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9294145625909438e-05,
+      "loss": 3.9344,
+      "step": 1264640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9285759678398917e-05,
+      "loss": 3.936,
+      "step": 1265152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9277373730888397e-05,
+      "loss": 3.9443,
+      "step": 1265664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.926898778337787e-05,
+      "loss": 3.9444,
+      "step": 1266176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.926060183586735e-05,
+      "loss": 3.9351,
+      "step": 1266688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.925221588835683e-05,
+      "loss": 3.9395,
+      "step": 1267200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9243829940846314e-05,
+      "loss": 3.9337,
+      "step": 1267712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9235443993335794e-05,
+      "loss": 3.9265,
+      "step": 1268224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9227058045825274e-05,
+      "loss": 3.924,
+      "step": 1268736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.921870485592222e-05,
+      "loss": 3.9344,
+      "step": 1269248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.92103189084117e-05,
+      "loss": 3.9461,
+      "step": 1269760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9201932960901175e-05,
+      "loss": 3.9383,
+      "step": 1270272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9193547013390655e-05,
+      "loss": 3.9285,
+      "step": 1270784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9185161065880135e-05,
+      "loss": 3.9359,
+      "step": 1271296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9176775118369615e-05,
+      "loss": 3.94,
+      "step": 1271808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9168389170859095e-05,
+      "loss": 3.9446,
+      "step": 1272320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9160003223348575e-05,
+      "loss": 3.9372,
+      "step": 1272832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9151633654641784e-05,
+      "loss": 3.9345,
+      "step": 1273344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9143247707131268e-05,
+      "loss": 3.9313,
+      "step": 1273856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9134861759620747e-05,
+      "loss": 3.9451,
+      "step": 1274368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9126475812110227e-05,
+      "loss": 3.9242,
+      "step": 1274880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.91181390010109e-05,
+      "loss": 3.9325,
+      "step": 1275392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.910975305350038e-05,
+      "loss": 3.9399,
+      "step": 1275904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.910136710598986e-05,
+      "loss": 3.9331,
+      "step": 1276416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9092981158479345e-05,
+      "loss": 3.93,
+      "step": 1276928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9084595210968818e-05,
+      "loss": 3.9368,
+      "step": 1277440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9076209263458298e-05,
+      "loss": 3.9302,
+      "step": 1277952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9067823315947778e-05,
+      "loss": 3.9319,
+      "step": 1278464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9059437368437258e-05,
+      "loss": 3.9349,
+      "step": 1278976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9051051420926738e-05,
+      "loss": 3.9339,
+      "step": 1279488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.904266547341622e-05,
+      "loss": 3.9446,
+      "step": 1280000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.90342795259057e-05,
+      "loss": 3.945,
+      "step": 1280512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.902589357839518e-05,
+      "loss": 3.9387,
+      "step": 1281024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.901752400968839e-05,
+      "loss": 3.9456,
+      "step": 1281536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.900913806217787e-05,
+      "loss": 3.9361,
+      "step": 1282048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.900075211466735e-05,
+      "loss": 3.9355,
+      "step": 1282560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.899236616715683e-05,
+      "loss": 3.9296,
+      "step": 1283072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8983980219646313e-05,
+      "loss": 3.9344,
+      "step": 1283584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8975594272135793e-05,
+      "loss": 3.936,
+      "step": 1284096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8967208324625273e-05,
+      "loss": 3.9364,
+      "step": 1284608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8958822377114753e-05,
+      "loss": 3.9355,
+      "step": 1285120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8950452808407962e-05,
+      "loss": 3.9397,
+      "step": 1285632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8942066860897442e-05,
+      "loss": 3.9371,
+      "step": 1286144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8933697292190655e-05,
+      "loss": 3.9282,
+      "step": 1286656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8925311344680135e-05,
+      "loss": 3.9421,
+      "step": 1287168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8916925397169615e-05,
+      "loss": 3.9408,
+      "step": 1287680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8908539449659095e-05,
+      "loss": 3.9295,
+      "step": 1288192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8900153502148575e-05,
+      "loss": 3.92,
+      "step": 1288704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8891767554638055e-05,
+      "loss": 3.9326,
+      "step": 1289216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8883381607127535e-05,
+      "loss": 3.9445,
+      "step": 1289728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8875012038420747e-05,
+      "loss": 3.9491,
+      "step": 1290240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8866626090910227e-05,
+      "loss": 3.9325,
+      "step": 1290752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8858256522203436e-05,
+      "loss": 3.9347,
+      "step": 1291264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8849870574692916e-05,
+      "loss": 3.9446,
+      "step": 1291776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8841484627182396e-05,
+      "loss": 3.9464,
+      "step": 1292288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8833098679671876e-05,
+      "loss": 3.9383,
+      "step": 1292800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8824712732161356e-05,
+      "loss": 3.9337,
+      "step": 1293312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8816326784650832e-05,
+      "loss": 3.9374,
+      "step": 1293824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8807940837140312e-05,
+      "loss": 3.9371,
+      "step": 1294336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8799554889629792e-05,
+      "loss": 3.9313,
+      "step": 1294848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8791185320923008e-05,
+      "loss": 3.9432,
+      "step": 1295360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8782799373412485e-05,
+      "loss": 3.9299,
+      "step": 1295872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8774413425901965e-05,
+      "loss": 3.9437,
+      "step": 1296384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8766027478391445e-05,
+      "loss": 3.9373,
+      "step": 1296896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8757641530880925e-05,
+      "loss": 3.9259,
+      "step": 1297408
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0236029624938965,
+      "eval_runtime": 303.947,
+      "eval_samples_per_second": 1255.452,
+      "eval_steps_per_second": 39.234,
+      "step": 1297440
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8749255583370405e-05,
+      "loss": 3.9411,
+      "step": 1297920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8740869635859885e-05,
+      "loss": 3.9307,
+      "step": 1298432
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8732483688349365e-05,
+      "loss": 3.9423,
+      "step": 1298944
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8724097740838845e-05,
+      "loss": 3.9387,
+      "step": 1299456
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8715711793328324e-05,
+      "loss": 3.9449,
+      "step": 1299968
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8707325845817808e-05,
+      "loss": 3.9331,
+      "step": 1300480
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8698939898307288e-05,
+      "loss": 3.926,
+      "step": 1300992
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8690553950796768e-05,
+      "loss": 3.9187,
+      "step": 1301504
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8682184382089977e-05,
+      "loss": 3.952,
+      "step": 1302016
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8673798434579457e-05,
+      "loss": 3.9397,
+      "step": 1302528
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8665412487068937e-05,
+      "loss": 3.9333,
+      "step": 1303040
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8657026539558417e-05,
+      "loss": 3.9444,
+      "step": 1303552
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.864865697085163e-05,
+      "loss": 3.9333,
+      "step": 1304064
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.864027102334111e-05,
+      "loss": 3.9293,
+      "step": 1304576
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.863188507583059e-05,
+      "loss": 3.9319,
+      "step": 1305088
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.862349912832007e-05,
+      "loss": 3.9305,
+      "step": 1305600
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8615129559613278e-05,
+      "loss": 3.9316,
+      "step": 1306112
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.860674361210276e-05,
+      "loss": 3.928,
+      "step": 1306624
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.859835766459224e-05,
+      "loss": 3.94,
+      "step": 1307136
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.858997171708172e-05,
+      "loss": 3.953,
+      "step": 1307648
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8581585769571195e-05,
+      "loss": 3.9359,
+      "step": 1308160
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.857321620086441e-05,
+      "loss": 3.9294,
+      "step": 1308672
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.856483025335389e-05,
+      "loss": 3.9351,
+      "step": 1309184
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8556444305843367e-05,
+      "loss": 3.9399,
+      "step": 1309696
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8548058358332847e-05,
+      "loss": 3.932,
+      "step": 1310208
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8539672410822327e-05,
+      "loss": 3.9232,
+      "step": 1310720
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8531302842115543e-05,
+      "loss": 3.928,
+      "step": 1311232
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8522916894605016e-05,
+      "loss": 3.9416,
+      "step": 1311744
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.85145309470945e-05,
+      "loss": 3.9169,
+      "step": 1312256
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.850614499958398e-05,
+      "loss": 3.926,
+      "step": 1312768
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8497775430877195e-05,
+      "loss": 3.9377,
+      "step": 1313280
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8489389483366668e-05,
+      "loss": 3.9353,
+      "step": 1313792
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8481003535856148e-05,
+      "loss": 3.9409,
+      "step": 1314304
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8472633967149364e-05,
+      "loss": 3.9358,
+      "step": 1314816
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8464264398442577e-05,
+      "loss": 3.9356,
+      "step": 1315328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8455878450932057e-05,
+      "loss": 3.939,
+      "step": 1315840
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8447492503421537e-05,
+      "loss": 3.9212,
+      "step": 1316352
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8439106555911016e-05,
+      "loss": 3.9348,
+      "step": 1316864
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8430736987204226e-05,
+      "loss": 3.9208,
+      "step": 1317376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8422351039693706e-05,
+      "loss": 3.9253,
+      "step": 1317888
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8413965092183185e-05,
+      "loss": 3.9303,
+      "step": 1318400
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.840557914467267e-05,
+      "loss": 3.9347,
+      "step": 1318912
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8397193197162142e-05,
+      "loss": 3.9311,
+      "step": 1319424
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8388807249651622e-05,
+      "loss": 3.9263,
+      "step": 1319936
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8380421302141102e-05,
+      "loss": 3.9313,
+      "step": 1320448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8372035354630582e-05,
+      "loss": 3.933,
+      "step": 1320960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8363649407120062e-05,
+      "loss": 3.9352,
+      "step": 1321472
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8355263459609542e-05,
+      "loss": 3.9234,
+      "step": 1321984
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8346893890902754e-05,
+      "loss": 3.9197,
+      "step": 1322496
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8338507943392234e-05,
+      "loss": 3.9361,
+      "step": 1323008
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8330121995881714e-05,
+      "loss": 3.9418,
+      "step": 1323520
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8321736048371194e-05,
+      "loss": 3.9282,
+      "step": 1324032
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8313350100860674e-05,
+      "loss": 3.9281,
+      "step": 1324544
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8304964153350154e-05,
+      "loss": 3.9271,
+      "step": 1325056
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8296594584643367e-05,
+      "loss": 3.9223,
+      "step": 1325568
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8288208637132846e-05,
+      "loss": 3.932,
+      "step": 1326080
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8279822689622326e-05,
+      "loss": 3.9182,
+      "step": 1326592
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8271436742111806e-05,
+      "loss": 3.9342,
+      "step": 1327104
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8263050794601286e-05,
+      "loss": 3.9332,
+      "step": 1327616
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8254664847090766e-05,
+      "loss": 3.9303,
+      "step": 1328128
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8246278899580246e-05,
+      "loss": 3.9148,
+      "step": 1328640
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8237892952069726e-05,
+      "loss": 3.9382,
+      "step": 1329152
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.822952338336294e-05,
+      "loss": 3.913,
+      "step": 1329664
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.822113743585242e-05,
+      "loss": 3.9274,
+      "step": 1330176
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8212767867145628e-05,
+      "loss": 3.9362,
+      "step": 1330688
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8204381919635108e-05,
+      "loss": 3.9345,
+      "step": 1331200
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8195995972124588e-05,
+      "loss": 3.9209,
+      "step": 1331712
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.818761002461407e-05,
+      "loss": 3.919,
+      "step": 1332224
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.817924045590728e-05,
+      "loss": 3.9171,
+      "step": 1332736
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.817085450839676e-05,
+      "loss": 3.928,
+      "step": 1333248
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.816246856088624e-05,
+      "loss": 3.9325,
+      "step": 1333760
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.815408261337572e-05,
+      "loss": 3.9294,
+      "step": 1334272
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.81456966658652e-05,
+      "loss": 3.9369,
+      "step": 1334784
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8137310718354676e-05,
+      "loss": 3.9368,
+      "step": 1335296
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8128924770844156e-05,
+      "loss": 3.9358,
+      "step": 1335808
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8120555202137372e-05,
+      "loss": 3.9231,
+      "step": 1336320
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8112169254626852e-05,
+      "loss": 3.9368,
+      "step": 1336832
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.810379968592006e-05,
+      "loss": 3.9367,
+      "step": 1337344
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.809541373840954e-05,
+      "loss": 3.9181,
+      "step": 1337856
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8087027790899025e-05,
+      "loss": 3.937,
+      "step": 1338368
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8078641843388498e-05,
+      "loss": 3.9253,
+      "step": 1338880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8070255895877978e-05,
+      "loss": 3.9471,
+      "step": 1339392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8061886327171194e-05,
+      "loss": 3.934,
+      "step": 1339904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.8053500379660674e-05,
+      "loss": 3.914,
+      "step": 1340416
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.804511443215015e-05,
+      "loss": 3.9317,
+      "step": 1340928
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.803672848463963e-05,
+      "loss": 3.9218,
+      "step": 1341440
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.802834253712911e-05,
+      "loss": 3.937,
+      "step": 1341952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.801995658961859e-05,
+      "loss": 3.932,
+      "step": 1342464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.801157064210807e-05,
+      "loss": 3.9296,
+      "step": 1342976
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.800320107340128e-05,
+      "loss": 3.9292,
+      "step": 1343488
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7994815125890762e-05,
+      "loss": 3.924,
+      "step": 1344000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7986429178380242e-05,
+      "loss": 3.9184,
+      "step": 1344512
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7978043230869722e-05,
+      "loss": 3.9155,
+      "step": 1345024
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7969657283359202e-05,
+      "loss": 3.9264,
+      "step": 1345536
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7961271335848682e-05,
+      "loss": 3.9356,
+      "step": 1346048
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7952885388338162e-05,
+      "loss": 3.9269,
+      "step": 1346560
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7944499440827642e-05,
+      "loss": 3.9209,
+      "step": 1347072
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7936129872120855e-05,
+      "loss": 3.9233,
+      "step": 1347584
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7927743924610335e-05,
+      "loss": 3.9274,
+      "step": 1348096
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7919357977099815e-05,
+      "loss": 3.936,
+      "step": 1348608
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7910972029589295e-05,
+      "loss": 3.9268,
+      "step": 1349120
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7902586082078774e-05,
+      "loss": 3.9255,
+      "step": 1349632
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7894200134568254e-05,
+      "loss": 3.9258,
+      "step": 1350144
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7885846944665196e-05,
+      "loss": 3.9346,
+      "step": 1350656
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7877460997154676e-05,
+      "loss": 3.9131,
+      "step": 1351168
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7869075049644156e-05,
+      "loss": 3.9257,
+      "step": 1351680
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7860705480937365e-05,
+      "loss": 3.9285,
+      "step": 1352192
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7852319533426845e-05,
+      "loss": 3.9225,
+      "step": 1352704
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7843933585916325e-05,
+      "loss": 3.9182,
+      "step": 1353216
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.783554763840581e-05,
+      "loss": 3.9246,
+      "step": 1353728
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7827161690895288e-05,
+      "loss": 3.9234,
+      "step": 1354240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7818775743384768e-05,
+      "loss": 3.9222,
+      "step": 1354752
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7810389795874248e-05,
+      "loss": 3.9239,
+      "step": 1355264
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7802003848363728e-05,
+      "loss": 3.9252,
+      "step": 1355776
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7793617900853208e-05,
+      "loss": 3.9369,
+      "step": 1356288
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7785231953342688e-05,
+      "loss": 3.9364,
+      "step": 1356800
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7776846005832165e-05,
+      "loss": 3.9309,
+      "step": 1357312
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7768460058321645e-05,
+      "loss": 3.9333,
+      "step": 1357824
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7760074110811125e-05,
+      "loss": 3.9303,
+      "step": 1358336
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.775172092090807e-05,
+      "loss": 3.9261,
+      "step": 1358848
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.774333497339755e-05,
+      "loss": 3.9183,
+      "step": 1359360
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.773494902588703e-05,
+      "loss": 3.923,
+      "step": 1359872
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.772656307837651e-05,
+      "loss": 3.9269,
+      "step": 1360384
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7718193509669722e-05,
+      "loss": 3.9279,
+      "step": 1360896
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7709807562159202e-05,
+      "loss": 3.9274,
+      "step": 1361408
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7701421614648682e-05,
+      "loss": 3.9309,
+      "step": 1361920
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7693035667138162e-05,
+      "loss": 3.9221,
+      "step": 1362432
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.768466609843137e-05,
+      "loss": 3.9253,
+      "step": 1362944
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.767628015092085e-05,
+      "loss": 3.9262,
+      "step": 1363456
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7667894203410334e-05,
+      "loss": 3.9338,
+      "step": 1363968
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7659508255899807e-05,
+      "loss": 3.921,
+      "step": 1364480
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7651122308389287e-05,
+      "loss": 3.9124,
+      "step": 1364992
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7642736360878767e-05,
+      "loss": 3.9192,
+      "step": 1365504
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7634350413368247e-05,
+      "loss": 3.9357,
+      "step": 1366016
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7625964465857727e-05,
+      "loss": 3.9388,
+      "step": 1366528
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.761759489715094e-05,
+      "loss": 3.9269,
+      "step": 1367040
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.760920894964042e-05,
+      "loss": 3.9255,
+      "step": 1367552
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7600839380933635e-05,
+      "loss": 3.9358,
+      "step": 1368064
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.759245343342311e-05,
+      "loss": 3.9356,
+      "step": 1368576
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.758406748591259e-05,
+      "loss": 3.9274,
+      "step": 1369088
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7575681538402072e-05,
+      "loss": 3.9242,
+      "step": 1369600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7567295590891552e-05,
+      "loss": 3.9271,
+      "step": 1370112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7558909643381032e-05,
+      "loss": 3.9296,
+      "step": 1370624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.755054007467424e-05,
+      "loss": 3.9193,
+      "step": 1371136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.754215412716372e-05,
+      "loss": 3.9341,
+      "step": 1371648
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.75337681796532e-05,
+      "loss": 3.9206,
+      "step": 1372160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.752538223214268e-05,
+      "loss": 3.9329,
+      "step": 1372672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7516996284632164e-05,
+      "loss": 3.9271,
+      "step": 1373184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.7508610337121644e-05,
+      "loss": 3.9208,
+      "step": 1373696
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.018845081329346,
+      "eval_runtime": 302.335,
+      "eval_samples_per_second": 1262.146,
+      "eval_steps_per_second": 39.443,
+      "step": 1373760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7500224389611124e-05,
+      "loss": 3.9331,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7491838442100604e-05,
+      "loss": 3.9213,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7483452494590084e-05,
+      "loss": 3.9308,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7475066547079564e-05,
+      "loss": 3.9293,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7466680599569044e-05,
+      "loss": 3.938,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.745829465205852e-05,
+      "loss": 3.9215,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7449908704548e-05,
+      "loss": 3.9182,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.744152275703748e-05,
+      "loss": 3.9102,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.743313680952696e-05,
+      "loss": 3.9445,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.742476724082017e-05,
+      "loss": 3.9288,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.741638129330965e-05,
+      "loss": 3.9212,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7407995345799133e-05,
+      "loss": 3.9396,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7399609398288613e-05,
+      "loss": 3.9208,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7391223450778093e-05,
+      "loss": 3.926,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7382837503267573e-05,
+      "loss": 3.9202,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7374451555757053e-05,
+      "loss": 3.92,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7366065608246533e-05,
+      "loss": 3.9285,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7357679660736012e-05,
+      "loss": 3.9166,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7349293713225492e-05,
+      "loss": 3.9334,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7340907765714972e-05,
+      "loss": 3.9403,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7332521818204452e-05,
+      "loss": 3.927,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7324135870693936e-05,
+      "loss": 3.9231,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.731574992318341e-05,
+      "loss": 3.9264,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.730736397567289e-05,
+      "loss": 3.9298,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.729897802816237e-05,
+      "loss": 3.9203,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7290608459455585e-05,
+      "loss": 3.9203,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7282238890748797e-05,
+      "loss": 3.913,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7273852943238277e-05,
+      "loss": 3.9333,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7265466995727757e-05,
+      "loss": 3.9069,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7257081048217237e-05,
+      "loss": 3.9188,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.724869510070671e-05,
+      "loss": 3.9288,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7240325531999926e-05,
+      "loss": 3.9288,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7231939584489406e-05,
+      "loss": 3.9287,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7223553636978883e-05,
+      "loss": 3.9274,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.72151840682721e-05,
+      "loss": 3.9238,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.720679812076158e-05,
+      "loss": 3.9323,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.719841217325106e-05,
+      "loss": 3.9171,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7190026225740535e-05,
+      "loss": 3.9218,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7181640278230015e-05,
+      "loss": 3.9107,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7173254330719495e-05,
+      "loss": 3.9179,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7164868383208975e-05,
+      "loss": 3.9173,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7156482435698455e-05,
+      "loss": 3.9294,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7148096488187935e-05,
+      "loss": 3.92,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7139710540677415e-05,
+      "loss": 3.9179,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7131340971970627e-05,
+      "loss": 3.9242,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7122955024460107e-05,
+      "loss": 3.9228,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7114569076949587e-05,
+      "loss": 3.9256,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7106183129439067e-05,
+      "loss": 3.9146,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7097797181928547e-05,
+      "loss": 3.9085,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7089411234418027e-05,
+      "loss": 3.9326,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7081025286907507e-05,
+      "loss": 3.93,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7072639339396987e-05,
+      "loss": 3.9205,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.70642697706902e-05,
+      "loss": 3.9213,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.705588382317968e-05,
+      "loss": 3.9153,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.704751425447289e-05,
+      "loss": 3.9176,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.703912830696237e-05,
+      "loss": 3.9235,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7030742359451848e-05,
+      "loss": 3.9089,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7022356411941328e-05,
+      "loss": 3.9253,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.701397046443081e-05,
+      "loss": 3.9218,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.700560089572402e-05,
+      "loss": 3.9185,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.69972149482135e-05,
+      "loss": 3.9084,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698882900070298e-05,
+      "loss": 3.9305,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698044305319246e-05,
+      "loss": 3.9069,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.697205710568194e-05,
+      "loss": 3.92,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.696367115817142e-05,
+      "loss": 3.9228,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6955285210660897e-05,
+      "loss": 3.9271,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6946899263150377e-05,
+      "loss": 3.9105,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6938529694443593e-05,
+      "loss": 3.9152,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6930143746933066e-05,
+      "loss": 3.9043,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6921774178226282e-05,
+      "loss": 3.9205,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6913404609519494e-05,
+      "loss": 3.9242,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6905018662008974e-05,
+      "loss": 3.923,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6896632714498454e-05,
+      "loss": 3.9256,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6888246766987934e-05,
+      "loss": 3.9266,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6879860819477414e-05,
+      "loss": 3.9316,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6871474871966894e-05,
+      "loss": 3.9113,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6863105303260107e-05,
+      "loss": 3.9241,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6854719355749587e-05,
+      "loss": 3.9319,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6846333408239067e-05,
+      "loss": 3.9084,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6837963839532276e-05,
+      "loss": 3.9253,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6829577892021756e-05,
+      "loss": 3.9139,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6821191944511236e-05,
+      "loss": 3.9395,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.681280599700072e-05,
+      "loss": 3.9249,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6804420049490192e-05,
+      "loss": 3.907,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6796034101979672e-05,
+      "loss": 3.9197,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6787648154469152e-05,
+      "loss": 3.9151,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6779262206958632e-05,
+      "loss": 3.9281,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6770876259448112e-05,
+      "loss": 3.9246,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6762490311937595e-05,
+      "loss": 3.9174,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6754104364427075e-05,
+      "loss": 3.9192,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6745718416916555e-05,
+      "loss": 3.916,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6737332469406035e-05,
+      "loss": 3.9127,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6728946521895515e-05,
+      "loss": 3.901,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6720576953188724e-05,
+      "loss": 3.9175,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6712191005678204e-05,
+      "loss": 3.9255,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6703805058167684e-05,
+      "loss": 3.9197,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6695419110657167e-05,
+      "loss": 3.9103,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6687033163146647e-05,
+      "loss": 3.914,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6678647215636127e-05,
+      "loss": 3.9173,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6670277646929336e-05,
+      "loss": 3.9301,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6661891699418816e-05,
+      "loss": 3.9154,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6653505751908296e-05,
+      "loss": 3.918,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6645119804397776e-05,
+      "loss": 3.9133,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.663675023569099e-05,
+      "loss": 3.9274,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.662836428818047e-05,
+      "loss": 3.9024,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.661997834066995e-05,
+      "loss": 3.9207,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.661159239315943e-05,
+      "loss": 3.9162,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6603206445648905e-05,
+      "loss": 3.9195,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6594820498138385e-05,
+      "loss": 3.9084,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6586434550627865e-05,
+      "loss": 3.9182,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.657806498192108e-05,
+      "loss": 3.9144,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6569679034410554e-05,
+      "loss": 3.9139,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6561293086900034e-05,
+      "loss": 3.9183,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6552907139389514e-05,
+      "loss": 3.9165,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6544521191878997e-05,
+      "loss": 3.923,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6536135244368477e-05,
+      "loss": 3.9276,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6527765675661687e-05,
+      "loss": 3.9215,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6519379728151166e-05,
+      "loss": 3.9252,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6510993780640646e-05,
+      "loss": 3.9195,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6502607833130126e-05,
+      "loss": 3.9143,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6494221885619606e-05,
+      "loss": 3.9132,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.648585231691282e-05,
+      "loss": 3.9176,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.64774663694023e-05,
+      "loss": 3.9136,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.646908042189178e-05,
+      "loss": 3.9236,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.646069447438126e-05,
+      "loss": 3.9161,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.645230852687074e-05,
+      "loss": 3.9227,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.644392257936022e-05,
+      "loss": 3.9156,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.643555301065343e-05,
+      "loss": 3.9152,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.642716706314291e-05,
+      "loss": 3.9158,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.641878111563239e-05,
+      "loss": 3.9285,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.641039516812187e-05,
+      "loss": 3.9116,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.640200922061135e-05,
+      "loss": 3.9087,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.639362327310083e-05,
+      "loss": 3.9069,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6385253704394043e-05,
+      "loss": 3.9239,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6376867756883523e-05,
+      "loss": 3.9293,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6368481809373003e-05,
+      "loss": 3.9217,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6360095861862483e-05,
+      "loss": 3.9153,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6351709914351963e-05,
+      "loss": 3.9286,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6343340345645172e-05,
+      "loss": 3.9237,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6334954398134652e-05,
+      "loss": 3.9221,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6326568450624136e-05,
+      "loss": 3.914,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6318182503113616e-05,
+      "loss": 3.9213,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6309812934406825e-05,
+      "loss": 3.9193,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6301426986896305e-05,
+      "loss": 3.909,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6293041039385785e-05,
+      "loss": 3.9261,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6284655091875264e-05,
+      "loss": 3.9194,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.627626914436474e-05,
+      "loss": 3.9195,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.626788319685422e-05,
+      "loss": 3.92,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.62594972493437e-05,
+      "loss": 3.9086,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.015772819519043,
+      "eval_runtime": 295.5556,
+      "eval_samples_per_second": 1291.097,
+      "eval_steps_per_second": 40.348,
+      "step": 1450080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.625111130183318e-05,
+      "loss": 3.9259,
+      "step": 1450496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.624274173312639e-05,
+      "loss": 3.9167,
+      "step": 1451008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6234355785615873e-05,
+      "loss": 3.9217,
+      "step": 1451520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6225986216909086e-05,
+      "loss": 3.9205,
+      "step": 1452032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6217600269398562e-05,
+      "loss": 3.9321,
+      "step": 1452544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6209214321888042e-05,
+      "loss": 3.9157,
+      "step": 1453056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6200828374377522e-05,
+      "loss": 3.907,
+      "step": 1453568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6192442426867002e-05,
+      "loss": 3.9043,
+      "step": 1454080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6184072858160215e-05,
+      "loss": 3.9289,
+      "step": 1454592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6175686910649695e-05,
+      "loss": 3.9255,
+      "step": 1455104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6167300963139175e-05,
+      "loss": 3.9088,
+      "step": 1455616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6158915015628655e-05,
+      "loss": 3.9282,
+      "step": 1456128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6150545446921864e-05,
+      "loss": 3.914,
+      "step": 1456640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6142159499411344e-05,
+      "loss": 3.9179,
+      "step": 1457152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6133773551900824e-05,
+      "loss": 3.9161,
+      "step": 1457664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6125387604390307e-05,
+      "loss": 3.9108,
+      "step": 1458176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6117001656879787e-05,
+      "loss": 3.9174,
+      "step": 1458688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6108615709369267e-05,
+      "loss": 3.9091,
+      "step": 1459200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6100246140662476e-05,
+      "loss": 3.9243,
+      "step": 1459712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6091860193151956e-05,
+      "loss": 3.9309,
+      "step": 1460224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6083474245641436e-05,
+      "loss": 3.9235,
+      "step": 1460736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6075088298130916e-05,
+      "loss": 3.9155,
+      "step": 1461248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.60667023506204e-05,
+      "loss": 3.9126,
+      "step": 1461760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.605831640310988e-05,
+      "loss": 3.9266,
+      "step": 1462272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.604993045559936e-05,
+      "loss": 3.9117,
+      "step": 1462784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.604154450808884e-05,
+      "loss": 3.9123,
+      "step": 1463296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6033174939382048e-05,
+      "loss": 3.9064,
+      "step": 1463808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6024788991871528e-05,
+      "loss": 3.9242,
+      "step": 1464320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.601641942316474e-05,
+      "loss": 3.9,
+      "step": 1464832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.600804985445795e-05,
+      "loss": 3.9117,
+      "step": 1465344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.599966390694743e-05,
+      "loss": 3.9205,
+      "step": 1465856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.599127795943691e-05,
+      "loss": 3.9182,
+      "step": 1466368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.598289201192639e-05,
+      "loss": 3.921,
+      "step": 1466880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.597450606441587e-05,
+      "loss": 3.9211,
+      "step": 1467392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5966136495709082e-05,
+      "loss": 3.9177,
+      "step": 1467904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5957750548198562e-05,
+      "loss": 3.9233,
+      "step": 1468416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5949364600688042e-05,
+      "loss": 3.912,
+      "step": 1468928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5940978653177522e-05,
+      "loss": 3.9078,
+      "step": 1469440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5932592705667002e-05,
+      "loss": 3.9033,
+      "step": 1469952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5924206758156482e-05,
+      "loss": 3.9093,
+      "step": 1470464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5915820810645962e-05,
+      "loss": 3.9077,
+      "step": 1470976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5907434863135445e-05,
+      "loss": 3.9209,
+      "step": 1471488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5899048915624925e-05,
+      "loss": 3.9111,
+      "step": 1472000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5890662968114398e-05,
+      "loss": 3.9111,
+      "step": 1472512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5882293399407614e-05,
+      "loss": 3.9179,
+      "step": 1473024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5873907451897094e-05,
+      "loss": 3.91,
+      "step": 1473536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5865521504386574e-05,
+      "loss": 3.9229,
+      "step": 1474048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.585713555687605e-05,
+      "loss": 3.9089,
+      "step": 1474560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.584874960936553e-05,
+      "loss": 3.8949,
+      "step": 1475072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.584036366185501e-05,
+      "loss": 3.9284,
+      "step": 1475584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5831994093148226e-05,
+      "loss": 3.9197,
+      "step": 1476096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.58236081456377e-05,
+      "loss": 3.9166,
+      "step": 1476608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5815222198127183e-05,
+      "loss": 3.9114,
+      "step": 1477120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5806836250616663e-05,
+      "loss": 3.9085,
+      "step": 1477632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5798450303106143e-05,
+      "loss": 3.9095,
+      "step": 1478144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5790064355595623e-05,
+      "loss": 3.914,
+      "step": 1478656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5781678408085103e-05,
+      "loss": 3.9033,
+      "step": 1479168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5773292460574583e-05,
+      "loss": 3.9112,
+      "step": 1479680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5764906513064063e-05,
+      "loss": 3.919,
+      "step": 1480192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5756553323161004e-05,
+      "loss": 3.9083,
+      "step": 1480704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5748167375650484e-05,
+      "loss": 3.899,
+      "step": 1481216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5739781428139964e-05,
+      "loss": 3.9236,
+      "step": 1481728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5731395480629444e-05,
+      "loss": 3.8992,
+      "step": 1482240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5723009533118924e-05,
+      "loss": 3.9067,
+      "step": 1482752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5714623585608404e-05,
+      "loss": 3.9184,
+      "step": 1483264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5706237638097884e-05,
+      "loss": 3.9196,
+      "step": 1483776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5697851690587367e-05,
+      "loss": 3.9005,
+      "step": 1484288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5689465743076847e-05,
+      "loss": 3.909,
+      "step": 1484800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5681096174370056e-05,
+      "loss": 3.893,
+      "step": 1485312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5672710226859536e-05,
+      "loss": 3.9177,
+      "step": 1485824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5664324279349016e-05,
+      "loss": 3.9161,
+      "step": 1486336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.565595471064223e-05,
+      "loss": 3.9147,
+      "step": 1486848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.564756876313171e-05,
+      "loss": 3.9146,
+      "step": 1487360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.563918281562119e-05,
+      "loss": 3.9191,
+      "step": 1487872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.563079686811067e-05,
+      "loss": 3.926,
+      "step": 1488384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5622427299403878e-05,
+      "loss": 3.9045,
+      "step": 1488896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5614041351893358e-05,
+      "loss": 3.9161,
+      "step": 1489408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5605655404382838e-05,
+      "loss": 3.9235,
+      "step": 1489920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5597269456872318e-05,
+      "loss": 3.8984,
+      "step": 1490432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.55888835093618e-05,
+      "loss": 3.9178,
+      "step": 1490944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.558051394065501e-05,
+      "loss": 3.9089,
+      "step": 1491456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.557212799314449e-05,
+      "loss": 3.9252,
+      "step": 1491968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.556374204563397e-05,
+      "loss": 3.9205,
+      "step": 1492480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.555535609812345e-05,
+      "loss": 3.8987,
+      "step": 1492992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.554697015061293e-05,
+      "loss": 3.9129,
+      "step": 1493504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.553858420310241e-05,
+      "loss": 3.9083,
+      "step": 1494016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5530198255591886e-05,
+      "loss": 3.9213,
+      "step": 1494528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5521812308081366e-05,
+      "loss": 3.9132,
+      "step": 1495040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5513442739374582e-05,
+      "loss": 3.9102,
+      "step": 1495552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.550507317066779e-05,
+      "loss": 3.9117,
+      "step": 1496064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.549668722315727e-05,
+      "loss": 3.9112,
+      "step": 1496576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5488301275646755e-05,
+      "loss": 3.9064,
+      "step": 1497088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5479915328136235e-05,
+      "loss": 3.8956,
+      "step": 1497600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5471529380625708e-05,
+      "loss": 3.9071,
+      "step": 1498112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5463143433115188e-05,
+      "loss": 3.919,
+      "step": 1498624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5454757485604668e-05,
+      "loss": 3.9184,
+      "step": 1499136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5446371538094148e-05,
+      "loss": 3.8989,
+      "step": 1499648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.543800196938736e-05,
+      "loss": 3.9091,
+      "step": 1500160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.542961602187684e-05,
+      "loss": 3.9083,
+      "step": 1500672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.542123007436632e-05,
+      "loss": 3.9208,
+      "step": 1501184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.54128441268558e-05,
+      "loss": 3.9096,
+      "step": 1501696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.540447455814901e-05,
+      "loss": 3.9115,
+      "step": 1502208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5396088610638492e-05,
+      "loss": 3.9043,
+      "step": 1502720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5387719041931708e-05,
+      "loss": 3.9214,
+      "step": 1503232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.537933309442118e-05,
+      "loss": 3.8933,
+      "step": 1503744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.537094714691066e-05,
+      "loss": 3.9128,
+      "step": 1504256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.536256119940014e-05,
+      "loss": 3.9043,
+      "step": 1504768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.535417525188962e-05,
+      "loss": 3.9095,
+      "step": 1505280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5345805683182834e-05,
+      "loss": 3.9004,
+      "step": 1505792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5337419735672314e-05,
+      "loss": 3.915,
+      "step": 1506304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5329033788161794e-05,
+      "loss": 3.9012,
+      "step": 1506816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5320647840651274e-05,
+      "loss": 3.9053,
+      "step": 1507328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5312261893140754e-05,
+      "loss": 3.9092,
+      "step": 1507840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5303875945630234e-05,
+      "loss": 3.909,
+      "step": 1508352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5295489998119714e-05,
+      "loss": 3.9179,
+      "step": 1508864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5287104050609193e-05,
+      "loss": 3.9199,
+      "step": 1509376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5278734481902406e-05,
+      "loss": 3.9134,
+      "step": 1509888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5270348534391886e-05,
+      "loss": 3.9171,
+      "step": 1510400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5261962586881366e-05,
+      "loss": 3.913,
+      "step": 1510912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5253576639370846e-05,
+      "loss": 3.9052,
+      "step": 1511424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5245207070664055e-05,
+      "loss": 3.9057,
+      "step": 1511936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5236821123153538e-05,
+      "loss": 3.9085,
+      "step": 1512448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5228435175643018e-05,
+      "loss": 3.9087,
+      "step": 1512960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5220049228132498e-05,
+      "loss": 3.9155,
+      "step": 1513472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5211663280621978e-05,
+      "loss": 3.907,
+      "step": 1513984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5203277333111458e-05,
+      "loss": 3.9162,
+      "step": 1514496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5194891385600938e-05,
+      "loss": 3.9038,
+      "step": 1515008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5186521816894147e-05,
+      "loss": 3.9087,
+      "step": 1515520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.517813586938363e-05,
+      "loss": 3.9062,
+      "step": 1516032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.516974992187311e-05,
+      "loss": 3.9207,
+      "step": 1516544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.516136397436259e-05,
+      "loss": 3.9056,
+      "step": 1517056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.515297802685207e-05,
+      "loss": 3.8993,
+      "step": 1517568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5144592079341544e-05,
+      "loss": 3.8969,
+      "step": 1518080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5136206131831023e-05,
+      "loss": 3.9135,
+      "step": 1518592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.512783656312424e-05,
+      "loss": 3.9194,
+      "step": 1519104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5119450615613716e-05,
+      "loss": 3.9161,
+      "step": 1519616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5111064668103196e-05,
+      "loss": 3.912,
+      "step": 1520128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5102678720592676e-05,
+      "loss": 3.9212,
+      "step": 1520640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5094292773082156e-05,
+      "loss": 3.9139,
+      "step": 1521152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5085906825571636e-05,
+      "loss": 3.913,
+      "step": 1521664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5077520878061116e-05,
+      "loss": 3.9096,
+      "step": 1522176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.50691349305506e-05,
+      "loss": 3.9127,
+      "step": 1522688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.506074898304008e-05,
+      "loss": 3.9125,
+      "step": 1523200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5052379414333288e-05,
+      "loss": 3.899,
+      "step": 1523712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5043993466822768e-05,
+      "loss": 3.9193,
+      "step": 1524224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5035607519312248e-05,
+      "loss": 3.9133,
+      "step": 1524736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5027221571801728e-05,
+      "loss": 3.9084,
+      "step": 1525248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5018835624291208e-05,
+      "loss": 3.9098,
+      "step": 1525760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.501046605558442e-05,
+      "loss": 3.9049,
+      "step": 1526272
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.012207508087158,
+      "eval_runtime": 297.2743,
+      "eval_samples_per_second": 1283.633,
+      "eval_steps_per_second": 40.114,
+      "step": 1526400
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.50020801080739e-05,
+      "loss": 3.9125,
+      "step": 1526784
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.499369416056338e-05,
+      "loss": 3.9088,
+      "step": 1527296
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.498530821305286e-05,
+      "loss": 3.9148,
+      "step": 1527808
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.497693864434607e-05,
+      "loss": 3.9094,
+      "step": 1528320
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.496855269683555e-05,
+      "loss": 3.923,
+      "step": 1528832
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4960166749325033e-05,
+      "loss": 3.9109,
+      "step": 1529344
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4951797180618242e-05,
+      "loss": 3.898,
+      "step": 1529856
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4943411233107722e-05,
+      "loss": 3.8998,
+      "step": 1530368
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.49350252855972e-05,
+      "loss": 3.9157,
+      "step": 1530880
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.492663933808668e-05,
+      "loss": 3.916,
+      "step": 1531392
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.491825339057616e-05,
+      "loss": 3.9042,
+      "step": 1531904
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.490986744306564e-05,
+      "loss": 3.9178,
+      "step": 1532416
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4901497874358854e-05,
+      "loss": 3.9088,
+      "step": 1532928
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4893111926848334e-05,
+      "loss": 3.9105,
+      "step": 1533440
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4884725979337814e-05,
+      "loss": 3.906,
+      "step": 1533952
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.487634003182729e-05,
+      "loss": 3.9026,
+      "step": 1534464
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4867970463120503e-05,
+      "loss": 3.9056,
+      "step": 1534976
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4859584515609986e-05,
+      "loss": 3.903,
+      "step": 1535488
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4851198568099463e-05,
+      "loss": 3.9176,
+      "step": 1536000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4842812620588943e-05,
+      "loss": 3.9179,
+      "step": 1536512
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4834426673078423e-05,
+      "loss": 3.9155,
+      "step": 1537024
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4826040725567903e-05,
+      "loss": 3.9102,
+      "step": 1537536
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4817654778057383e-05,
+      "loss": 3.9034,
+      "step": 1538048
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4809268830546863e-05,
+      "loss": 3.9153,
+      "step": 1538560
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4800899261840075e-05,
+      "loss": 3.9049,
+      "step": 1539072
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4792529693133288e-05,
+      "loss": 3.9035,
+      "step": 1539584
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4784143745622764e-05,
+      "loss": 3.8962,
+      "step": 1540096
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4775757798112244e-05,
+      "loss": 3.9105,
+      "step": 1540608
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4767371850601724e-05,
+      "loss": 3.8991,
+      "step": 1541120
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.4758985903091207e-05,
+      "loss": 3.8969,
+      "step": 1541632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4750616334384417e-05,
+      "loss": 3.9136,
+      "step": 1542144
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4742230386873897e-05,
+      "loss": 3.9082,
+      "step": 1542656
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4733844439363376e-05,
+      "loss": 3.9151,
+      "step": 1543168
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4725458491852856e-05,
+      "loss": 3.9146,
+      "step": 1543680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.471708892314607e-05,
+      "loss": 3.9039,
+      "step": 1544192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.470870297563555e-05,
+      "loss": 3.9167,
+      "step": 1544704
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.470031702812503e-05,
+      "loss": 3.9075,
+      "step": 1545216
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.469193108061451e-05,
+      "loss": 3.8988,
+      "step": 1545728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4683561511907718e-05,
+      "loss": 3.8951,
+      "step": 1546240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4675175564397198e-05,
+      "loss": 3.8975,
+      "step": 1546752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4666789616886678e-05,
+      "loss": 3.9019,
+      "step": 1547264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4658403669376158e-05,
+      "loss": 3.9135,
+      "step": 1547776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.465001772186564e-05,
+      "loss": 3.9016,
+      "step": 1548288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.464164815315885e-05,
+      "loss": 3.9014,
+      "step": 1548800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.463326220564833e-05,
+      "loss": 3.9133,
+      "step": 1549312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.462487625813781e-05,
+      "loss": 3.9048,
+      "step": 1549824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.461649031062729e-05,
+      "loss": 3.9124,
+      "step": 1550336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.460810436311677e-05,
+      "loss": 3.8985,
+      "step": 1550848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4599734794409983e-05,
+      "loss": 3.8835,
+      "step": 1551360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4591348846899462e-05,
+      "loss": 3.9185,
+      "step": 1551872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4582962899388942e-05,
+      "loss": 3.9149,
+      "step": 1552384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.457457695187842e-05,
+      "loss": 3.907,
+      "step": 1552896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.45661910043679e-05,
+      "loss": 3.904,
+      "step": 1553408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.455780505685738e-05,
+      "loss": 3.8988,
+      "step": 1553920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4549419109346862e-05,
+      "loss": 3.8993,
+      "step": 1554432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4541033161836342e-05,
+      "loss": 3.9041,
+      "step": 1554944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.453266359312955e-05,
+      "loss": 3.8975,
+      "step": 1555456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4524294024422764e-05,
+      "loss": 3.9026,
+      "step": 1555968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.451590807691224e-05,
+      "loss": 3.9109,
+      "step": 1556480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4507538508205456e-05,
+      "loss": 3.9052,
+      "step": 1556992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4499152560694936e-05,
+      "loss": 3.8902,
+      "step": 1557504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4490766613184413e-05,
+      "loss": 3.9156,
+      "step": 1558016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4482380665673893e-05,
+      "loss": 3.8913,
+      "step": 1558528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4473994718163373e-05,
+      "loss": 3.8929,
+      "step": 1559040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4465608770652853e-05,
+      "loss": 3.9127,
+      "step": 1559552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4457222823142333e-05,
+      "loss": 3.9145,
+      "step": 1560064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4448836875631816e-05,
+      "loss": 3.8915,
+      "step": 1560576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4440450928121296e-05,
+      "loss": 3.9014,
+      "step": 1561088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4432064980610772e-05,
+      "loss": 3.8853,
+      "step": 1561600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4423679033100252e-05,
+      "loss": 3.9076,
+      "step": 1562112
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4415293085589732e-05,
+      "loss": 3.9088,
+      "step": 1562624
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4406923516882945e-05,
+      "loss": 3.9032,
+      "step": 1563136
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4398553948176157e-05,
+      "loss": 3.9049,
+      "step": 1563648
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4390168000665637e-05,
+      "loss": 3.9092,
+      "step": 1564160
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4381782053155117e-05,
+      "loss": 3.9228,
+      "step": 1564672
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4373396105644594e-05,
+      "loss": 3.8928,
+      "step": 1565184
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4365010158134074e-05,
+      "loss": 3.9107,
+      "step": 1565696
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4356624210623554e-05,
+      "loss": 3.9122,
+      "step": 1566208
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4348238263113034e-05,
+      "loss": 3.8894,
+      "step": 1566720
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4339852315602517e-05,
+      "loss": 3.912,
+      "step": 1567232
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4331482746895726e-05,
+      "loss": 3.9019,
+      "step": 1567744
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4323096799385206e-05,
+      "loss": 3.9164,
+      "step": 1568256
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4314710851874686e-05,
+      "loss": 3.9117,
+      "step": 1568768
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4306324904364166e-05,
+      "loss": 3.8935,
+      "step": 1569280
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.429795533565738e-05,
+      "loss": 3.9001,
+      "step": 1569792
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.428956938814686e-05,
+      "loss": 3.8979,
+      "step": 1570304
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4281199819440068e-05,
+      "loss": 3.9157,
+      "step": 1570816
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4272813871929547e-05,
+      "loss": 3.9046,
+      "step": 1571328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4264427924419027e-05,
+      "loss": 3.9006,
+      "step": 1571840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4256041976908507e-05,
+      "loss": 3.9006,
+      "step": 1572352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4247656029397987e-05,
+      "loss": 3.9053,
+      "step": 1572864
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.423927008188747e-05,
+      "loss": 3.9004,
+      "step": 1573376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.423088413437695e-05,
+      "loss": 3.8858,
+      "step": 1573888
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.422251456567016e-05,
+      "loss": 3.9018,
+      "step": 1574400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.421412861815964e-05,
+      "loss": 3.9098,
+      "step": 1574912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.420575904945285e-05,
+      "loss": 3.91,
+      "step": 1575424
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4197373101942332e-05,
+      "loss": 3.8888,
+      "step": 1575936
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4188987154431812e-05,
+      "loss": 3.8992,
+      "step": 1576448
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4180601206921292e-05,
+      "loss": 3.9003,
+      "step": 1576960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4172215259410772e-05,
+      "loss": 3.91,
+      "step": 1577472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.416382931190025e-05,
+      "loss": 3.9052,
+      "step": 1577984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.415544336438973e-05,
+      "loss": 3.9026,
+      "step": 1578496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.414705741687921e-05,
+      "loss": 3.8956,
+      "step": 1579008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.413867146936869e-05,
+      "loss": 3.9121,
+      "step": 1579520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4130285521858172e-05,
+      "loss": 3.8857,
+      "step": 1580032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.412191595315138e-05,
+      "loss": 3.9047,
+      "step": 1580544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.411353000564086e-05,
+      "loss": 3.8949,
+      "step": 1581056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4105160436934073e-05,
+      "loss": 3.9074,
+      "step": 1581568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.409677448942355e-05,
+      "loss": 3.8889,
+      "step": 1582080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4088388541913033e-05,
+      "loss": 3.9084,
+      "step": 1582592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4080002594402513e-05,
+      "loss": 3.8987,
+      "step": 1583104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4071616646891993e-05,
+      "loss": 3.8963,
+      "step": 1583616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4063230699381473e-05,
+      "loss": 3.8995,
+      "step": 1584128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4054844751870953e-05,
+      "loss": 3.9004,
+      "step": 1584640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.404645880436043e-05,
+      "loss": 3.9075,
+      "step": 1585152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.403807285684991e-05,
+      "loss": 3.9125,
+      "step": 1585664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.402968690933939e-05,
+      "loss": 3.9072,
+      "step": 1586176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4021300961828873e-05,
+      "loss": 3.9099,
+      "step": 1586688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4012931393122082e-05,
+      "loss": 3.9047,
+      "step": 1587200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4004545445611562e-05,
+      "loss": 3.8973,
+      "step": 1587712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3996159498101042e-05,
+      "loss": 3.8947,
+      "step": 1588224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3987789929394254e-05,
+      "loss": 3.9072,
+      "step": 1588736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3979403981883734e-05,
+      "loss": 3.899,
+      "step": 1589248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3971018034373214e-05,
+      "loss": 3.9063,
+      "step": 1589760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3962632086862694e-05,
+      "loss": 3.9006,
+      "step": 1590272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3954246139352174e-05,
+      "loss": 3.9066,
+      "step": 1590784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3945860191841654e-05,
+      "loss": 3.8999,
+      "step": 1591296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3937474244331134e-05,
+      "loss": 3.9001,
+      "step": 1591808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.392908829682061e-05,
+      "loss": 3.8973,
+      "step": 1592320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3920702349310094e-05,
+      "loss": 3.9098,
+      "step": 1592832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3912316401799574e-05,
+      "loss": 3.9021,
+      "step": 1593344
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3903930454289054e-05,
+      "loss": 3.8949,
+      "step": 1593856
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3895544506778534e-05,
+      "loss": 3.8861,
+      "step": 1594368
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3887191316875475e-05,
+      "loss": 3.9068,
+      "step": 1594880
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3878805369364955e-05,
+      "loss": 3.9109,
+      "step": 1595392
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3870419421854435e-05,
+      "loss": 3.9073,
+      "step": 1595904
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3862033474343915e-05,
+      "loss": 3.905,
+      "step": 1596416
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3853647526833395e-05,
+      "loss": 3.9098,
+      "step": 1596928
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3845277958126608e-05,
+      "loss": 3.9056,
+      "step": 1597440
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3836892010616084e-05,
+      "loss": 3.9091,
+      "step": 1597952
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3828506063105564e-05,
+      "loss": 3.897,
+      "step": 1598464
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3820120115595048e-05,
+      "loss": 3.904,
+      "step": 1598976
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3811734168084528e-05,
+      "loss": 3.9027,
+      "step": 1599488
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3803348220574008e-05,
+      "loss": 3.8969,
+      "step": 1600000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3794962273063488e-05,
+      "loss": 3.9093,
+      "step": 1600512
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.3786576325552967e-05,
+      "loss": 3.906,
+      "step": 1601024
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.377822313564991e-05,
+      "loss": 3.9009,
+      "step": 1601536
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.376983718813939e-05,
+      "loss": 3.9017,
+      "step": 1602048
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.37614676194326e-05,
+      "loss": 3.9024,
+      "step": 1602560
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.009007930755615,
+      "eval_runtime": 302.2224,
+      "eval_samples_per_second": 1262.616,
+      "eval_steps_per_second": 39.458,
+      "step": 1602720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.375308167192208e-05,
+      "loss": 3.9071,
+      "step": 1603072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3744695724411558e-05,
+      "loss": 3.9021,
+      "step": 1603584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3736309776901038e-05,
+      "loss": 3.9072,
+      "step": 1604096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3727923829390518e-05,
+      "loss": 3.9001,
+      "step": 1604608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3719537881879998e-05,
+      "loss": 3.9144,
+      "step": 1605120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.371115193436948e-05,
+      "loss": 3.9028,
+      "step": 1605632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.370276598685896e-05,
+      "loss": 3.8929,
+      "step": 1606144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.369438003934844e-05,
+      "loss": 3.8905,
+      "step": 1606656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3685994091837918e-05,
+      "loss": 3.9094,
+      "step": 1607168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3677608144327398e-05,
+      "loss": 3.9077,
+      "step": 1607680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3669222196816878e-05,
+      "loss": 3.901,
+      "step": 1608192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3660836249306358e-05,
+      "loss": 3.9064,
+      "step": 1608704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.365246668059957e-05,
+      "loss": 3.9031,
+      "step": 1609216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.364408073308905e-05,
+      "loss": 3.9015,
+      "step": 1609728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.363569478557853e-05,
+      "loss": 3.8991,
+      "step": 1610240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.362730883806801e-05,
+      "loss": 3.8947,
+      "step": 1610752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.361892289055749e-05,
+      "loss": 3.9005,
+      "step": 1611264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.361053694304697e-05,
+      "loss": 3.8944,
+      "step": 1611776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.360215099553645e-05,
+      "loss": 3.9066,
+      "step": 1612288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.359376504802593e-05,
+      "loss": 3.9102,
+      "step": 1612800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.358537910051541e-05,
+      "loss": 3.9123,
+      "step": 1613312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3577009531808622e-05,
+      "loss": 3.9013,
+      "step": 1613824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.35686235842981e-05,
+      "loss": 3.8982,
+      "step": 1614336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.356023763678758e-05,
+      "loss": 3.9049,
+      "step": 1614848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.355185168927706e-05,
+      "loss": 3.8971,
+      "step": 1615360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.354348212057027e-05,
+      "loss": 3.8995,
+      "step": 1615872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.353509617305975e-05,
+      "loss": 3.8853,
+      "step": 1616384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.352671022554923e-05,
+      "loss": 3.9075,
+      "step": 1616896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.351832427803871e-05,
+      "loss": 3.8954,
+      "step": 1617408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.350993833052819e-05,
+      "loss": 3.8893,
+      "step": 1617920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.350155238301767e-05,
+      "loss": 3.9019,
+      "step": 1618432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.349316643550715e-05,
+      "loss": 3.9067,
+      "step": 1618944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.348478048799663e-05,
+      "loss": 3.908,
+      "step": 1619456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.347639454048611e-05,
+      "loss": 3.9036,
+      "step": 1619968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3468024971779323e-05,
+      "loss": 3.8991,
+      "step": 1620480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3459655403072532e-05,
+      "loss": 3.9069,
+      "step": 1620992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3451269455562012e-05,
+      "loss": 3.9041,
+      "step": 1621504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3442883508051496e-05,
+      "loss": 3.8893,
+      "step": 1622016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3434497560540976e-05,
+      "loss": 3.8897,
+      "step": 1622528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3426111613030452e-05,
+      "loss": 3.8888,
+      "step": 1623040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3417725665519932e-05,
+      "loss": 3.8957,
+      "step": 1623552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3409339718009412e-05,
+      "loss": 3.9059,
+      "step": 1624064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3400953770498892e-05,
+      "loss": 3.895,
+      "step": 1624576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3392600580595837e-05,
+      "loss": 3.8909,
+      "step": 1625088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3384214633085317e-05,
+      "loss": 3.9044,
+      "step": 1625600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3375828685574797e-05,
+      "loss": 3.8939,
+      "step": 1626112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3367442738064277e-05,
+      "loss": 3.9088,
+      "step": 1626624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3359056790553754e-05,
+      "loss": 3.8909,
+      "step": 1627136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3350670843043234e-05,
+      "loss": 3.8734,
+      "step": 1627648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3342284895532713e-05,
+      "loss": 3.9127,
+      "step": 1628160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3333898948022197e-05,
+      "loss": 3.9055,
+      "step": 1628672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3325513000511677e-05,
+      "loss": 3.9039,
+      "step": 1629184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3317127053001157e-05,
+      "loss": 3.8977,
+      "step": 1629696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3308757484294366e-05,
+      "loss": 3.8937,
+      "step": 1630208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3300371536783846e-05,
+      "loss": 3.8928,
+      "step": 1630720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3291985589273326e-05,
+      "loss": 3.8954,
+      "step": 1631232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3283599641762806e-05,
+      "loss": 3.8898,
+      "step": 1631744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3275213694252286e-05,
+      "loss": 3.8952,
+      "step": 1632256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3266827746741766e-05,
+      "loss": 3.9024,
+      "step": 1632768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3258441799231246e-05,
+      "loss": 3.8969,
+      "step": 1633280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3250072230524455e-05,
+      "loss": 3.8877,
+      "step": 1633792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3241686283013935e-05,
+      "loss": 3.9067,
+      "step": 1634304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3233300335503415e-05,
+      "loss": 3.8871,
+      "step": 1634816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3224914387992898e-05,
+      "loss": 3.8876,
+      "step": 1635328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3216528440482378e-05,
+      "loss": 3.9062,
+      "step": 1635840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3208142492971858e-05,
+      "loss": 3.9043,
+      "step": 1636352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3199756545461338e-05,
+      "loss": 3.8848,
+      "step": 1636864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3191370597950814e-05,
+      "loss": 3.8949,
+      "step": 1637376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3183001029244027e-05,
+      "loss": 3.8805,
+      "step": 1637888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3174615081733507e-05,
+      "loss": 3.8972,
+      "step": 1638400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.316622913422299e-05,
+      "loss": 3.8999,
+      "step": 1638912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.31578595655162e-05,
+      "loss": 3.8996,
+      "step": 1639424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.314947361800568e-05,
+      "loss": 3.8991,
+      "step": 1639936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.314108767049516e-05,
+      "loss": 3.9035,
+      "step": 1640448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3132701722984636e-05,
+      "loss": 3.9118,
+      "step": 1640960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.312431577547412e-05,
+      "loss": 3.8877,
+      "step": 1641472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.311594620676733e-05,
+      "loss": 3.9055,
+      "step": 1641984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.310757663806054e-05,
+      "loss": 3.9051,
+      "step": 1642496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.309919069055002e-05,
+      "loss": 3.8831,
+      "step": 1643008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.30908047430395e-05,
+      "loss": 3.9061,
+      "step": 1643520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.308241879552898e-05,
+      "loss": 3.8953,
+      "step": 1644032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.307403284801846e-05,
+      "loss": 3.909,
+      "step": 1644544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.306564690050794e-05,
+      "loss": 3.9057,
+      "step": 1645056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.305726095299742e-05,
+      "loss": 3.8902,
+      "step": 1645568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.30488750054869e-05,
+      "loss": 3.8896,
+      "step": 1646080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.304048905797638e-05,
+      "loss": 3.8931,
+      "step": 1646592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.303210311046586e-05,
+      "loss": 3.9098,
+      "step": 1647104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.302371716295534e-05,
+      "loss": 3.8956,
+      "step": 1647616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3015347594248553e-05,
+      "loss": 3.8944,
+      "step": 1648128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3006961646738033e-05,
+      "loss": 3.894,
+      "step": 1648640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2998575699227513e-05,
+      "loss": 3.9005,
+      "step": 1649152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2990189751716993e-05,
+      "loss": 3.8933,
+      "step": 1649664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.298180380420647e-05,
+      "loss": 3.8777,
+      "step": 1650176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.297341785669595e-05,
+      "loss": 3.894,
+      "step": 1650688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.296503190918543e-05,
+      "loss": 3.9026,
+      "step": 1651200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2956645961674912e-05,
+      "loss": 3.9026,
+      "step": 1651712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.294827639296812e-05,
+      "loss": 3.8857,
+      "step": 1652224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.29398904454576e-05,
+      "loss": 3.8905,
+      "step": 1652736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.293150449794708e-05,
+      "loss": 3.8963,
+      "step": 1653248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.292313492924029e-05,
+      "loss": 3.9005,
+      "step": 1653760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2914748981729774e-05,
+      "loss": 3.9034,
+      "step": 1654272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2906379413022986e-05,
+      "loss": 3.8902,
+      "step": 1654784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2897993465512466e-05,
+      "loss": 3.8963,
+      "step": 1655296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2889607518001943e-05,
+      "loss": 3.9029,
+      "step": 1655808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2881221570491423e-05,
+      "loss": 3.8822,
+      "step": 1656320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2872835622980903e-05,
+      "loss": 3.8955,
+      "step": 1656832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2864449675470383e-05,
+      "loss": 3.8852,
+      "step": 1657344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2856063727959866e-05,
+      "loss": 3.9004,
+      "step": 1657856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2847677780449346e-05,
+      "loss": 3.8853,
+      "step": 1658368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2839291832938826e-05,
+      "loss": 3.901,
+      "step": 1658880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2830922264232035e-05,
+      "loss": 3.8903,
+      "step": 1659392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2822536316721515e-05,
+      "loss": 3.8907,
+      "step": 1659904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2814150369210995e-05,
+      "loss": 3.8922,
+      "step": 1660416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2805764421700475e-05,
+      "loss": 3.8967,
+      "step": 1660928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2797378474189955e-05,
+      "loss": 3.8948,
+      "step": 1661440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2788992526679435e-05,
+      "loss": 3.9105,
+      "step": 1661952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2780639336776376e-05,
+      "loss": 3.9018,
+      "step": 1662464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2772253389265856e-05,
+      "loss": 3.9031,
+      "step": 1662976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2763867441755336e-05,
+      "loss": 3.8936,
+      "step": 1663488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.275548149424482e-05,
+      "loss": 3.8962,
+      "step": 1664000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2747095546734296e-05,
+      "loss": 3.8878,
+      "step": 1664512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2738709599223776e-05,
+      "loss": 3.8953,
+      "step": 1665024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2730323651713256e-05,
+      "loss": 3.895,
+      "step": 1665536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2721937704202736e-05,
+      "loss": 3.8983,
+      "step": 1666048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2713551756692216e-05,
+      "loss": 3.896,
+      "step": 1666560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.270518218798543e-05,
+      "loss": 3.8962,
+      "step": 1667072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.269679624047491e-05,
+      "loss": 3.8944,
+      "step": 1667584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.268842667176812e-05,
+      "loss": 3.8986,
+      "step": 1668096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2680040724257598e-05,
+      "loss": 3.89,
+      "step": 1668608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2671654776747078e-05,
+      "loss": 3.8999,
+      "step": 1669120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2663268829236557e-05,
+      "loss": 3.8951,
+      "step": 1669632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2654882881726037e-05,
+      "loss": 3.891,
+      "step": 1670144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.264649693421552e-05,
+      "loss": 3.8775,
+      "step": 1670656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2638110986705e-05,
+      "loss": 3.9041,
+      "step": 1671168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2629725039194477e-05,
+      "loss": 3.9019,
+      "step": 1671680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2621339091683957e-05,
+      "loss": 3.9028,
+      "step": 1672192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.261296952297717e-05,
+      "loss": 3.8976,
+      "step": 1672704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.260458357546665e-05,
+      "loss": 3.9017,
+      "step": 1673216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.259619762795613e-05,
+      "loss": 3.8963,
+      "step": 1673728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2587828059249342e-05,
+      "loss": 3.9068,
+      "step": 1674240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2579442111738822e-05,
+      "loss": 3.8894,
+      "step": 1674752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2571056164228302e-05,
+      "loss": 3.8961,
+      "step": 1675264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.256267021671778e-05,
+      "loss": 3.8961,
+      "step": 1675776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.255428426920726e-05,
+      "loss": 3.8927,
+      "step": 1676288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.254589832169674e-05,
+      "loss": 3.9041,
+      "step": 1676800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2537512374186222e-05,
+      "loss": 3.8952,
+      "step": 1677312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2529126426675702e-05,
+      "loss": 3.9026,
+      "step": 1677824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2520740479165182e-05,
+      "loss": 3.8898,
+      "step": 1678336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.251235453165466e-05,
+      "loss": 3.8964,
+      "step": 1678848
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.006854057312012,
+      "eval_runtime": 299.2348,
+      "eval_samples_per_second": 1275.223,
+      "eval_steps_per_second": 39.852,
+      "step": 1679040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.250398496294787e-05,
+      "loss": 3.9046,
+      "step": 1679360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.249559901543735e-05,
+      "loss": 3.8951,
+      "step": 1679872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.248721306792683e-05,
+      "loss": 3.8994,
+      "step": 1680384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.247882712041631e-05,
+      "loss": 3.8949,
+      "step": 1680896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.247044117290579e-05,
+      "loss": 3.9084,
+      "step": 1681408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.246205522539527e-05,
+      "loss": 3.8959,
+      "step": 1681920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.245366927788475e-05,
+      "loss": 3.8892,
+      "step": 1682432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.244528333037423e-05,
+      "loss": 3.8837,
+      "step": 1682944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.243689738286371e-05,
+      "loss": 3.9019,
+      "step": 1683456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.242851143535319e-05,
+      "loss": 3.9004,
+      "step": 1683968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.242012548784267e-05,
+      "loss": 3.8938,
+      "step": 1684480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.241173954033215e-05,
+      "loss": 3.9013,
+      "step": 1684992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2403369971625363e-05,
+      "loss": 3.8958,
+      "step": 1685504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.239498402411484e-05,
+      "loss": 3.8955,
+      "step": 1686016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.238659807660432e-05,
+      "loss": 3.8892,
+      "step": 1686528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.23782121290938e-05,
+      "loss": 3.8891,
+      "step": 1687040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2369826181583283e-05,
+      "loss": 3.893,
+      "step": 1687552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2361440234072763e-05,
+      "loss": 3.8903,
+      "step": 1688064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2353070665365972e-05,
+      "loss": 3.8993,
+      "step": 1688576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.234468471785545e-05,
+      "loss": 3.906,
+      "step": 1689088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.233629877034493e-05,
+      "loss": 3.9088,
+      "step": 1689600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.232791282283441e-05,
+      "loss": 3.8962,
+      "step": 1690112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.231952687532389e-05,
+      "loss": 3.892,
+      "step": 1690624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.231114092781337e-05,
+      "loss": 3.8955,
+      "step": 1691136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.230275498030285e-05,
+      "loss": 3.8942,
+      "step": 1691648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.229436903279233e-05,
+      "loss": 3.8952,
+      "step": 1692160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2285999464085544e-05,
+      "loss": 3.8752,
+      "step": 1692672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2277629895378753e-05,
+      "loss": 3.9009,
+      "step": 1693184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2269243947868236e-05,
+      "loss": 3.8916,
+      "step": 1693696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2260858000357716e-05,
+      "loss": 3.8827,
+      "step": 1694208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2252472052847196e-05,
+      "loss": 3.8941,
+      "step": 1694720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2244086105336673e-05,
+      "loss": 3.9002,
+      "step": 1695232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2235700157826153e-05,
+      "loss": 3.8973,
+      "step": 1695744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2227314210315633e-05,
+      "loss": 3.8989,
+      "step": 1696256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2218928262805113e-05,
+      "loss": 3.8911,
+      "step": 1696768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2210558694098325e-05,
+      "loss": 3.9022,
+      "step": 1697280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2202172746587805e-05,
+      "loss": 3.8964,
+      "step": 1697792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2193786799077285e-05,
+      "loss": 3.883,
+      "step": 1698304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2185400851566765e-05,
+      "loss": 3.8851,
+      "step": 1698816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2177014904056245e-05,
+      "loss": 3.8824,
+      "step": 1699328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2168628956545725e-05,
+      "loss": 3.8895,
+      "step": 1699840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.21602430090352e-05,
+      "loss": 3.9001,
+      "step": 1700352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2151873440328417e-05,
+      "loss": 3.8881,
+      "step": 1700864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2143487492817897e-05,
+      "loss": 3.886,
+      "step": 1701376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2135101545307377e-05,
+      "loss": 3.894,
+      "step": 1701888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2126715597796854e-05,
+      "loss": 3.8873,
+      "step": 1702400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2118346029090066e-05,
+      "loss": 3.9034,
+      "step": 1702912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2109960081579546e-05,
+      "loss": 3.887,
+      "step": 1703424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2101574134069026e-05,
+      "loss": 3.8649,
+      "step": 1703936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2093188186558506e-05,
+      "loss": 3.9044,
+      "step": 1704448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2084802239047986e-05,
+      "loss": 3.9003,
+      "step": 1704960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.20764326703412e-05,
+      "loss": 3.8979,
+      "step": 1705472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2068046722830675e-05,
+      "loss": 3.8916,
+      "step": 1705984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.205967715412389e-05,
+      "loss": 3.8885,
+      "step": 1706496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.205129120661337e-05,
+      "loss": 3.8851,
+      "step": 1707008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.204290525910285e-05,
+      "loss": 3.892,
+      "step": 1707520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2034519311592328e-05,
+      "loss": 3.8864,
+      "step": 1708032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2026133364081808e-05,
+      "loss": 3.8858,
+      "step": 1708544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2017747416571287e-05,
+      "loss": 3.8959,
+      "step": 1709056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2009361469060767e-05,
+      "loss": 3.8938,
+      "step": 1709568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2000975521550247e-05,
+      "loss": 3.8798,
+      "step": 1710080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.199258957403973e-05,
+      "loss": 3.899,
+      "step": 1710592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.198422000533294e-05,
+      "loss": 3.8792,
+      "step": 1711104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.197583405782242e-05,
+      "loss": 3.8819,
+      "step": 1711616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.19674481103119e-05,
+      "loss": 3.8978,
+      "step": 1712128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.195906216280138e-05,
+      "loss": 3.8976,
+      "step": 1712640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1950676215290856e-05,
+      "loss": 3.8779,
+      "step": 1713152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.194229026778034e-05,
+      "loss": 3.8856,
+      "step": 1713664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1933920699073552e-05,
+      "loss": 3.8729,
+      "step": 1714176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1925534751563032e-05,
+      "loss": 3.8936,
+      "step": 1714688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.191714880405251e-05,
+      "loss": 3.8946,
+      "step": 1715200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.190876285654199e-05,
+      "loss": 3.8941,
+      "step": 1715712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.19003932878352e-05,
+      "loss": 3.8901,
+      "step": 1716224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.189200734032468e-05,
+      "loss": 3.8955,
+      "step": 1716736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.188362139281416e-05,
+      "loss": 3.9039,
+      "step": 1717248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.187523544530364e-05,
+      "loss": 3.8836,
+      "step": 1717760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.186684949779312e-05,
+      "loss": 3.8977,
+      "step": 1718272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.18584635502826e-05,
+      "loss": 3.8985,
+      "step": 1718784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.185007760277208e-05,
+      "loss": 3.8803,
+      "step": 1719296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.184169165526156e-05,
+      "loss": 3.895,
+      "step": 1719808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1833322086554773e-05,
+      "loss": 3.8952,
+      "step": 1720320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1824936139044253e-05,
+      "loss": 3.8975,
+      "step": 1720832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1816550191533733e-05,
+      "loss": 3.9016,
+      "step": 1721344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1808180622826942e-05,
+      "loss": 3.8831,
+      "step": 1721856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1799794675316422e-05,
+      "loss": 3.8842,
+      "step": 1722368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1791408727805902e-05,
+      "loss": 3.8888,
+      "step": 1722880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1783022780295385e-05,
+      "loss": 3.896,
+      "step": 1723392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1774636832784862e-05,
+      "loss": 3.8929,
+      "step": 1723904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1766267264078075e-05,
+      "loss": 3.8887,
+      "step": 1724416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1757881316567554e-05,
+      "loss": 3.8903,
+      "step": 1724928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1749495369057034e-05,
+      "loss": 3.8933,
+      "step": 1725440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1741109421546514e-05,
+      "loss": 3.8868,
+      "step": 1725952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1732723474035994e-05,
+      "loss": 3.8696,
+      "step": 1726464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1724337526525474e-05,
+      "loss": 3.8894,
+      "step": 1726976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1715967957818687e-05,
+      "loss": 3.8945,
+      "step": 1727488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1707582010308163e-05,
+      "loss": 3.8952,
+      "step": 1728000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1699196062797643e-05,
+      "loss": 3.8817,
+      "step": 1728512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1690810115287123e-05,
+      "loss": 3.8833,
+      "step": 1729024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1682424167776603e-05,
+      "loss": 3.8884,
+      "step": 1729536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1674038220266087e-05,
+      "loss": 3.8914,
+      "step": 1730048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1665652272755567e-05,
+      "loss": 3.8976,
+      "step": 1730560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1657266325245043e-05,
+      "loss": 3.8831,
+      "step": 1731072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1648896756538256e-05,
+      "loss": 3.8901,
+      "step": 1731584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1640527187831465e-05,
+      "loss": 3.8959,
+      "step": 1732096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1632141240320948e-05,
+      "loss": 3.8759,
+      "step": 1732608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1623755292810428e-05,
+      "loss": 3.8907,
+      "step": 1733120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1615369345299908e-05,
+      "loss": 3.8751,
+      "step": 1733632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1606983397789388e-05,
+      "loss": 3.8962,
+      "step": 1734144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1598597450278864e-05,
+      "loss": 3.879,
+      "step": 1734656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1590227881572077e-05,
+      "loss": 3.8937,
+      "step": 1735168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1581841934061557e-05,
+      "loss": 3.8843,
+      "step": 1735680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.157345598655104e-05,
+      "loss": 3.885,
+      "step": 1736192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1565070039040517e-05,
+      "loss": 3.8852,
+      "step": 1736704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1556684091529997e-05,
+      "loss": 3.8918,
+      "step": 1737216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1548298144019477e-05,
+      "loss": 3.8892,
+      "step": 1737728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1539912196508957e-05,
+      "loss": 3.9043,
+      "step": 1738240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.153154262780217e-05,
+      "loss": 3.8881,
+      "step": 1738752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.152315668029165e-05,
+      "loss": 3.901,
+      "step": 1739264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.151477073278113e-05,
+      "loss": 3.8887,
+      "step": 1739776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.150638478527061e-05,
+      "loss": 3.8875,
+      "step": 1740288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.149799883776009e-05,
+      "loss": 3.8885,
+      "step": 1740800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1489629269053298e-05,
+      "loss": 3.8841,
+      "step": 1741312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1481243321542778e-05,
+      "loss": 3.8848,
+      "step": 1741824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.147285737403226e-05,
+      "loss": 3.8941,
+      "step": 1742336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.146447142652174e-05,
+      "loss": 3.8872,
+      "step": 1742848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.145610185781495e-05,
+      "loss": 3.8918,
+      "step": 1743360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.144771591030443e-05,
+      "loss": 3.8859,
+      "step": 1743872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.143932996279391e-05,
+      "loss": 3.8912,
+      "step": 1744384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.143094401528339e-05,
+      "loss": 3.8885,
+      "step": 1744896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.142255806777287e-05,
+      "loss": 3.8888,
+      "step": 1745408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1414188499066083e-05,
+      "loss": 3.8865,
+      "step": 1745920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1405802551555563e-05,
+      "loss": 3.8865,
+      "step": 1746432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1397416604045043e-05,
+      "loss": 3.8727,
+      "step": 1746944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.138903065653452e-05,
+      "loss": 3.8958,
+      "step": 1747456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1380644709024e-05,
+      "loss": 3.896,
+      "step": 1747968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.137225876151348e-05,
+      "loss": 3.8982,
+      "step": 1748480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1363889192806695e-05,
+      "loss": 3.8881,
+      "step": 1748992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.135550324529617e-05,
+      "loss": 3.896,
+      "step": 1749504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.134711729778565e-05,
+      "loss": 3.8932,
+      "step": 1750016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.133873135027513e-05,
+      "loss": 3.9009,
+      "step": 1750528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.133034540276461e-05,
+      "loss": 3.8824,
+      "step": 1751040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.132195945525409e-05,
+      "loss": 3.887,
+      "step": 1751552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.131357350774357e-05,
+      "loss": 3.8902,
+      "step": 1752064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1305187560233055e-05,
+      "loss": 3.8901,
+      "step": 1752576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1296834370329993e-05,
+      "loss": 3.893,
+      "step": 1753088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1288448422819473e-05,
+      "loss": 3.8918,
+      "step": 1753600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1280062475308953e-05,
+      "loss": 3.894,
+      "step": 1754112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1271676527798433e-05,
+      "loss": 3.8841,
+      "step": 1754624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1263290580287916e-05,
+      "loss": 3.8947,
+      "step": 1755136
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0043206214904785,
+      "eval_runtime": 294.895,
+      "eval_samples_per_second": 1293.989,
+      "eval_steps_per_second": 40.438,
+      "step": 1755360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1254904632777396e-05,
+      "loss": 3.8966,
+      "step": 1755648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1246518685266876e-05,
+      "loss": 3.8938,
+      "step": 1756160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1238132737756353e-05,
+      "loss": 3.8887,
+      "step": 1756672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1229746790245833e-05,
+      "loss": 3.8889,
+      "step": 1757184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1221360842735313e-05,
+      "loss": 3.9006,
+      "step": 1757696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1212974895224792e-05,
+      "loss": 3.8936,
+      "step": 1758208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1204588947714272e-05,
+      "loss": 3.8842,
+      "step": 1758720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1196203000203756e-05,
+      "loss": 3.8807,
+      "step": 1759232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1187817052693236e-05,
+      "loss": 3.8912,
+      "step": 1759744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1179431105182712e-05,
+      "loss": 3.8997,
+      "step": 1760256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1171045157672192e-05,
+      "loss": 3.8841,
+      "step": 1760768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1162659210161672e-05,
+      "loss": 3.8952,
+      "step": 1761280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.115428964145488e-05,
+      "loss": 3.896,
+      "step": 1761792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1145903693944365e-05,
+      "loss": 3.8841,
+      "step": 1762304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1137517746433845e-05,
+      "loss": 3.8849,
+      "step": 1762816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1129131798923325e-05,
+      "loss": 3.8819,
+      "step": 1763328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1120745851412805e-05,
+      "loss": 3.8864,
+      "step": 1763840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1112359903902284e-05,
+      "loss": 3.8833,
+      "step": 1764352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1103973956391764e-05,
+      "loss": 3.8903,
+      "step": 1764864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.109558800888124e-05,
+      "loss": 3.9026,
+      "step": 1765376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1087202061370724e-05,
+      "loss": 3.9002,
+      "step": 1765888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1078832492663937e-05,
+      "loss": 3.8912,
+      "step": 1766400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1070446545153417e-05,
+      "loss": 3.8853,
+      "step": 1766912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1062060597642893e-05,
+      "loss": 3.8859,
+      "step": 1767424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1053674650132373e-05,
+      "loss": 3.888,
+      "step": 1767936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1045305081425586e-05,
+      "loss": 3.8862,
+      "step": 1768448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1036919133915066e-05,
+      "loss": 3.8717,
+      "step": 1768960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1028533186404546e-05,
+      "loss": 3.8953,
+      "step": 1769472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1020147238894026e-05,
+      "loss": 3.8852,
+      "step": 1769984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1011761291383506e-05,
+      "loss": 3.8789,
+      "step": 1770496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1003375343872986e-05,
+      "loss": 3.8803,
+      "step": 1771008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0994989396362466e-05,
+      "loss": 3.8987,
+      "step": 1771520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0986619827655678e-05,
+      "loss": 3.8929,
+      "step": 1772032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0978233880145158e-05,
+      "loss": 3.8917,
+      "step": 1772544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0969847932634638e-05,
+      "loss": 3.8836,
+      "step": 1773056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0961478363927847e-05,
+      "loss": 3.8928,
+      "step": 1773568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0953092416417327e-05,
+      "loss": 3.8929,
+      "step": 1774080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0944706468906807e-05,
+      "loss": 3.8761,
+      "step": 1774592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0936320521396287e-05,
+      "loss": 3.8772,
+      "step": 1775104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0927934573885767e-05,
+      "loss": 3.8764,
+      "step": 1775616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0919548626375247e-05,
+      "loss": 3.8843,
+      "step": 1776128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0911162678864727e-05,
+      "loss": 3.89,
+      "step": 1776640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0902776731354207e-05,
+      "loss": 3.8849,
+      "step": 1777152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0894390783843687e-05,
+      "loss": 3.8794,
+      "step": 1777664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0886021215136896e-05,
+      "loss": 3.8896,
+      "step": 1778176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.087763526762638e-05,
+      "loss": 3.8806,
+      "step": 1778688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.086924932011586e-05,
+      "loss": 3.8937,
+      "step": 1779200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.086086337260534e-05,
+      "loss": 3.8878,
+      "step": 1779712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.085247742509482e-05,
+      "loss": 3.8587,
+      "step": 1780224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.08440914775843e-05,
+      "loss": 3.895,
+      "step": 1780736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0835721908877508e-05,
+      "loss": 3.8941,
+      "step": 1781248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0827335961366988e-05,
+      "loss": 3.895,
+      "step": 1781760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.081895001385647e-05,
+      "loss": 3.8829,
+      "step": 1782272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.081056406634595e-05,
+      "loss": 3.8781,
+      "step": 1782784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0802178118835428e-05,
+      "loss": 3.8812,
+      "step": 1783296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.079380855012864e-05,
+      "loss": 3.8851,
+      "step": 1783808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.078542260261812e-05,
+      "loss": 3.8831,
+      "step": 1784320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.07770366551076e-05,
+      "loss": 3.8754,
+      "step": 1784832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.076865070759708e-05,
+      "loss": 3.8892,
+      "step": 1785344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.076026476008656e-05,
+      "loss": 3.8886,
+      "step": 1785856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.075187881257604e-05,
+      "loss": 3.8725,
+      "step": 1786368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.074349286506552e-05,
+      "loss": 3.8937,
+      "step": 1786880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.073512329635873e-05,
+      "loss": 3.8754,
+      "step": 1787392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.072673734884821e-05,
+      "loss": 3.8725,
+      "step": 1787904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0718367780141425e-05,
+      "loss": 3.8921,
+      "step": 1788416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.07099818326309e-05,
+      "loss": 3.8901,
+      "step": 1788928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.070159588512038e-05,
+      "loss": 3.874,
+      "step": 1789440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.069320993760986e-05,
+      "loss": 3.8831,
+      "step": 1789952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.068482399009934e-05,
+      "loss": 3.868,
+      "step": 1790464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.067643804258882e-05,
+      "loss": 3.8858,
+      "step": 1790976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.06680520950783e-05,
+      "loss": 3.8879,
+      "step": 1791488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.065966614756778e-05,
+      "loss": 3.8892,
+      "step": 1792000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0651296578860994e-05,
+      "loss": 3.8835,
+      "step": 1792512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0642910631350474e-05,
+      "loss": 3.8874,
+      "step": 1793024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0634524683839954e-05,
+      "loss": 3.9033,
+      "step": 1793536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.062613873632943e-05,
+      "loss": 3.8728,
+      "step": 1794048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.061775278881891e-05,
+      "loss": 3.8934,
+      "step": 1794560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.060936684130839e-05,
+      "loss": 3.8919,
+      "step": 1795072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0600980893797874e-05,
+      "loss": 3.8762,
+      "step": 1795584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0592611325091083e-05,
+      "loss": 3.8888,
+      "step": 1796096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0584225377580563e-05,
+      "loss": 3.8886,
+      "step": 1796608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0575839430070043e-05,
+      "loss": 3.8899,
+      "step": 1797120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0567453482559522e-05,
+      "loss": 3.897,
+      "step": 1797632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0559067535049002e-05,
+      "loss": 3.8737,
+      "step": 1798144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0550697966342215e-05,
+      "loss": 3.8788,
+      "step": 1798656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0542312018831695e-05,
+      "loss": 3.8804,
+      "step": 1799168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0533942450124904e-05,
+      "loss": 3.8921,
+      "step": 1799680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0525556502614384e-05,
+      "loss": 3.8827,
+      "step": 1800192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0517170555103864e-05,
+      "loss": 3.8861,
+      "step": 1800704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0508784607593344e-05,
+      "loss": 3.8864,
+      "step": 1801216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0500398660082827e-05,
+      "loss": 3.8792,
+      "step": 1801728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0492012712572307e-05,
+      "loss": 3.8831,
+      "step": 1802240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0483643143865516e-05,
+      "loss": 3.8652,
+      "step": 1802752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0475257196354996e-05,
+      "loss": 3.8813,
+      "step": 1803264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0466871248844476e-05,
+      "loss": 3.8862,
+      "step": 1803776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0458485301333956e-05,
+      "loss": 3.8897,
+      "step": 1804288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0450099353823436e-05,
+      "loss": 3.8763,
+      "step": 1804800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0441713406312916e-05,
+      "loss": 3.8741,
+      "step": 1805312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0433327458802396e-05,
+      "loss": 3.883,
+      "step": 1805824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0424941511291876e-05,
+      "loss": 3.8861,
+      "step": 1806336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0416555563781356e-05,
+      "loss": 3.8879,
+      "step": 1806848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0408169616270836e-05,
+      "loss": 3.8833,
+      "step": 1807360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0399783668760316e-05,
+      "loss": 3.8854,
+      "step": 1807872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0391397721249796e-05,
+      "loss": 3.8833,
+      "step": 1808384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0383028152543008e-05,
+      "loss": 3.8733,
+      "step": 1808896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0374642205032488e-05,
+      "loss": 3.8798,
+      "step": 1809408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0366272636325697e-05,
+      "loss": 3.872,
+      "step": 1809920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0357886688815177e-05,
+      "loss": 3.8925,
+      "step": 1810432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0349500741304657e-05,
+      "loss": 3.8724,
+      "step": 1810944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0341114793794137e-05,
+      "loss": 3.8847,
+      "step": 1811456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0332728846283617e-05,
+      "loss": 3.8791,
+      "step": 1811968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0324342898773097e-05,
+      "loss": 3.8819,
+      "step": 1812480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0315956951262577e-05,
+      "loss": 3.8786,
+      "step": 1812992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0307571003752057e-05,
+      "loss": 3.8839,
+      "step": 1813504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0299185056241537e-05,
+      "loss": 3.8839,
+      "step": 1814016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0290799108731017e-05,
+      "loss": 3.8973,
+      "step": 1814528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.028242954002423e-05,
+      "loss": 3.8817,
+      "step": 1815040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.027404359251371e-05,
+      "loss": 3.8956,
+      "step": 1815552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.026565764500319e-05,
+      "loss": 3.8799,
+      "step": 1816064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.025727169749267e-05,
+      "loss": 3.8813,
+      "step": 1816576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.024888574998215e-05,
+      "loss": 3.8819,
+      "step": 1817088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.024053256007909e-05,
+      "loss": 3.8835,
+      "step": 1817600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.023214661256857e-05,
+      "loss": 3.8756,
+      "step": 1818112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.022376066505805e-05,
+      "loss": 3.8866,
+      "step": 1818624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.021537471754753e-05,
+      "loss": 3.8778,
+      "step": 1819136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.020698877003701e-05,
+      "loss": 3.886,
+      "step": 1819648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.019860282252649e-05,
+      "loss": 3.8854,
+      "step": 1820160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.019021687501597e-05,
+      "loss": 3.8811,
+      "step": 1820672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.018183092750545e-05,
+      "loss": 3.8858,
+      "step": 1821184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.017344497999493e-05,
+      "loss": 3.8864,
+      "step": 1821696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0165075411288143e-05,
+      "loss": 3.8772,
+      "step": 1822208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0156689463777623e-05,
+      "loss": 3.8824,
+      "step": 1822720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.01483035162671e-05,
+      "loss": 3.8634,
+      "step": 1823232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.013991756875658e-05,
+      "loss": 3.888,
+      "step": 1823744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.013153162124606e-05,
+      "loss": 3.891,
+      "step": 1824256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0123162052539272e-05,
+      "loss": 3.8928,
+      "step": 1824768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0114776105028752e-05,
+      "loss": 3.8823,
+      "step": 1825280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0106390157518232e-05,
+      "loss": 3.8879,
+      "step": 1825792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0098004210007712e-05,
+      "loss": 3.8885,
+      "step": 1826304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.008961826249719e-05,
+      "loss": 3.8934,
+      "step": 1826816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.008123231498667e-05,
+      "loss": 3.8765,
+      "step": 1827328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.007284636747615e-05,
+      "loss": 3.8819,
+      "step": 1827840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0064476798769364e-05,
+      "loss": 3.8831,
+      "step": 1828352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0056090851258844e-05,
+      "loss": 3.8876,
+      "step": 1828864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0047704903748324e-05,
+      "loss": 3.8845,
+      "step": 1829376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0039318956237804e-05,
+      "loss": 3.8803,
+      "step": 1829888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.003093300872728e-05,
+      "loss": 3.8911,
+      "step": 1830400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0022563440020496e-05,
+      "loss": 3.8767,
+      "step": 1830912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0014177492509976e-05,
+      "loss": 3.8923,
+      "step": 1831424
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.002368927001953,
+      "eval_runtime": 291.7282,
+      "eval_samples_per_second": 1308.036,
+      "eval_steps_per_second": 40.877,
+      "step": 1831680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0005791544999453e-05,
+      "loss": 3.8865,
+      "step": 1831936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9997405597488933e-05,
+      "loss": 3.8846,
+      "step": 1832448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9989019649978413e-05,
+      "loss": 3.8842,
+      "step": 1832960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9980633702467893e-05,
+      "loss": 3.8839,
+      "step": 1833472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9972247754957373e-05,
+      "loss": 3.8958,
+      "step": 1833984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9963861807446853e-05,
+      "loss": 3.8868,
+      "step": 1834496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9955475859936336e-05,
+      "loss": 3.8758,
+      "step": 1835008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9947089912425813e-05,
+      "loss": 3.8728,
+      "step": 1835520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9938703964915293e-05,
+      "loss": 3.882,
+      "step": 1836032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9930318017404773e-05,
+      "loss": 3.8942,
+      "step": 1836544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9921932069894252e-05,
+      "loss": 3.8822,
+      "step": 1837056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9913546122383732e-05,
+      "loss": 3.8872,
+      "step": 1837568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9905176553676945e-05,
+      "loss": 3.8886,
+      "step": 1838080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9896790606166425e-05,
+      "loss": 3.8783,
+      "step": 1838592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9888404658655905e-05,
+      "loss": 3.8775,
+      "step": 1839104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9880018711145385e-05,
+      "loss": 3.8748,
+      "step": 1839616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9871632763634865e-05,
+      "loss": 3.8832,
+      "step": 1840128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.986324681612434e-05,
+      "loss": 3.8769,
+      "step": 1840640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.985486086861382e-05,
+      "loss": 3.8845,
+      "step": 1841152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9846491299907037e-05,
+      "loss": 3.8939,
+      "step": 1841664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9838105352396517e-05,
+      "loss": 3.8937,
+      "step": 1842176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9829719404885994e-05,
+      "loss": 3.8822,
+      "step": 1842688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9821333457375474e-05,
+      "loss": 3.8825,
+      "step": 1843200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9812947509864954e-05,
+      "loss": 3.8822,
+      "step": 1843712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9804561562354434e-05,
+      "loss": 3.8814,
+      "step": 1844224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9796175614843913e-05,
+      "loss": 3.8811,
+      "step": 1844736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9787789667333397e-05,
+      "loss": 3.868,
+      "step": 1845248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9779403719822873e-05,
+      "loss": 3.8905,
+      "step": 1845760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9771034151116086e-05,
+      "loss": 3.8825,
+      "step": 1846272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9762648203605566e-05,
+      "loss": 3.87,
+      "step": 1846784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9754262256095046e-05,
+      "loss": 3.8741,
+      "step": 1847296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9745876308584522e-05,
+      "loss": 3.8907,
+      "step": 1847808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9737490361074006e-05,
+      "loss": 3.8881,
+      "step": 1848320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9729104413563486e-05,
+      "loss": 3.8866,
+      "step": 1848832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9720734844856698e-05,
+      "loss": 3.8794,
+      "step": 1849344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9712365276149907e-05,
+      "loss": 3.88,
+      "step": 1849856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9703979328639387e-05,
+      "loss": 3.8918,
+      "step": 1850368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9695593381128867e-05,
+      "loss": 3.8722,
+      "step": 1850880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9687207433618347e-05,
+      "loss": 3.8708,
+      "step": 1851392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9678821486107827e-05,
+      "loss": 3.8735,
+      "step": 1851904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9670435538597307e-05,
+      "loss": 3.8765,
+      "step": 1852416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9662049591086787e-05,
+      "loss": 3.8845,
+      "step": 1852928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9653663643576267e-05,
+      "loss": 3.8745,
+      "step": 1853440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9645277696065747e-05,
+      "loss": 3.8826,
+      "step": 1853952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9636891748555227e-05,
+      "loss": 3.88,
+      "step": 1854464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9628505801044707e-05,
+      "loss": 3.8748,
+      "step": 1854976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.962013623233792e-05,
+      "loss": 3.8922,
+      "step": 1855488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.96117502848274e-05,
+      "loss": 3.8768,
+      "step": 1856000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.960336433731688e-05,
+      "loss": 3.8558,
+      "step": 1856512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9594978389806356e-05,
+      "loss": 3.8882,
+      "step": 1857024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9586592442295836e-05,
+      "loss": 3.8835,
+      "step": 1857536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9578206494785316e-05,
+      "loss": 3.8907,
+      "step": 1858048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.95698205472748e-05,
+      "loss": 3.8795,
+      "step": 1858560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9561450978568008e-05,
+      "loss": 3.8716,
+      "step": 1859072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9553065031057488e-05,
+      "loss": 3.8758,
+      "step": 1859584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9544679083546968e-05,
+      "loss": 3.8761,
+      "step": 1860096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9536293136036448e-05,
+      "loss": 3.8816,
+      "step": 1860608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9527907188525928e-05,
+      "loss": 3.8665,
+      "step": 1861120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.951953761981914e-05,
+      "loss": 3.8875,
+      "step": 1861632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.951115167230862e-05,
+      "loss": 3.8812,
+      "step": 1862144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.95027657247981e-05,
+      "loss": 3.8678,
+      "step": 1862656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.949437977728758e-05,
+      "loss": 3.8865,
+      "step": 1863168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.948599382977706e-05,
+      "loss": 3.8689,
+      "step": 1863680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9477607882266537e-05,
+      "loss": 3.8681,
+      "step": 1864192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9469221934756017e-05,
+      "loss": 3.8829,
+      "step": 1864704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.94608359872455e-05,
+      "loss": 3.8881,
+      "step": 1865216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.945246641853871e-05,
+      "loss": 3.8622,
+      "step": 1865728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.944408047102819e-05,
+      "loss": 3.881,
+      "step": 1866240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.943569452351767e-05,
+      "loss": 3.8613,
+      "step": 1866752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.942730857600715e-05,
+      "loss": 3.879,
+      "step": 1867264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.941892262849663e-05,
+      "loss": 3.8841,
+      "step": 1867776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.941055305978984e-05,
+      "loss": 3.8777,
+      "step": 1868288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9402183491083054e-05,
+      "loss": 3.8846,
+      "step": 1868800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9393797543572534e-05,
+      "loss": 3.879,
+      "step": 1869312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.938541159606201e-05,
+      "loss": 3.9007,
+      "step": 1869824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.937702564855149e-05,
+      "loss": 3.8671,
+      "step": 1870336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.936863970104097e-05,
+      "loss": 3.8874,
+      "step": 1870848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9360253753530454e-05,
+      "loss": 3.8825,
+      "step": 1871360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9351867806019934e-05,
+      "loss": 3.8758,
+      "step": 1871872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9343481858509414e-05,
+      "loss": 3.8798,
+      "step": 1872384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.933509591099889e-05,
+      "loss": 3.8797,
+      "step": 1872896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9326726342292103e-05,
+      "loss": 3.8841,
+      "step": 1873408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9318340394781583e-05,
+      "loss": 3.8891,
+      "step": 1873920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9309954447271063e-05,
+      "loss": 3.8663,
+      "step": 1874432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9301568499760543e-05,
+      "loss": 3.8747,
+      "step": 1874944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9293182552250023e-05,
+      "loss": 3.8748,
+      "step": 1875456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9284796604739502e-05,
+      "loss": 3.8877,
+      "step": 1875968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.927642703603271e-05,
+      "loss": 3.8745,
+      "step": 1876480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.926804108852219e-05,
+      "loss": 3.8813,
+      "step": 1876992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9259655141011675e-05,
+      "loss": 3.883,
+      "step": 1877504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9251269193501155e-05,
+      "loss": 3.8766,
+      "step": 1878016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9242883245990635e-05,
+      "loss": 3.8782,
+      "step": 1878528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9234497298480115e-05,
+      "loss": 3.8595,
+      "step": 1879040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9226127729773324e-05,
+      "loss": 3.8746,
+      "step": 1879552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9217741782262804e-05,
+      "loss": 3.881,
+      "step": 1880064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9209355834752284e-05,
+      "loss": 3.885,
+      "step": 1880576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9200969887241764e-05,
+      "loss": 3.8698,
+      "step": 1881088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9192583939731247e-05,
+      "loss": 3.8688,
+      "step": 1881600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9184197992220724e-05,
+      "loss": 3.8755,
+      "step": 1882112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9175812044710204e-05,
+      "loss": 3.8829,
+      "step": 1882624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9167426097199684e-05,
+      "loss": 3.8831,
+      "step": 1883136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9159056528492893e-05,
+      "loss": 3.8754,
+      "step": 1883648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9150670580982376e-05,
+      "loss": 3.8813,
+      "step": 1884160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9142284633471856e-05,
+      "loss": 3.8742,
+      "step": 1884672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9133898685961336e-05,
+      "loss": 3.8731,
+      "step": 1885184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9125512738450816e-05,
+      "loss": 3.8716,
+      "step": 1885696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9117143169744025e-05,
+      "loss": 3.8698,
+      "step": 1886208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9108757222233505e-05,
+      "loss": 3.8868,
+      "step": 1886720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9100371274722985e-05,
+      "loss": 3.8674,
+      "step": 1887232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9091985327212468e-05,
+      "loss": 3.8785,
+      "step": 1887744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9083599379701948e-05,
+      "loss": 3.8702,
+      "step": 1888256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9075229810995157e-05,
+      "loss": 3.8793,
+      "step": 1888768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9066843863484637e-05,
+      "loss": 3.8729,
+      "step": 1889280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9058457915974117e-05,
+      "loss": 3.8768,
+      "step": 1889792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9050071968463597e-05,
+      "loss": 3.8788,
+      "step": 1890304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9041686020953077e-05,
+      "loss": 3.891,
+      "step": 1890816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.903331645224629e-05,
+      "loss": 3.8788,
+      "step": 1891328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.90249468835395e-05,
+      "loss": 3.8874,
+      "step": 1891840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.901656093602898e-05,
+      "loss": 3.877,
+      "step": 1892352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.900817498851846e-05,
+      "loss": 3.8757,
+      "step": 1892864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.899978904100794e-05,
+      "loss": 3.8783,
+      "step": 1893376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8991403093497422e-05,
+      "loss": 3.8772,
+      "step": 1893888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8983017145986902e-05,
+      "loss": 3.871,
+      "step": 1894400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.897463119847638e-05,
+      "loss": 3.8815,
+      "step": 1894912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.896624525096586e-05,
+      "loss": 3.8708,
+      "step": 1895424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8957859303455338e-05,
+      "loss": 3.8806,
+      "step": 1895936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8949473355944818e-05,
+      "loss": 3.8765,
+      "step": 1896448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.894110378723803e-05,
+      "loss": 3.8798,
+      "step": 1896960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.893271783972751e-05,
+      "loss": 3.8754,
+      "step": 1897472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.892433189221699e-05,
+      "loss": 3.8842,
+      "step": 1897984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.891594594470647e-05,
+      "loss": 3.8695,
+      "step": 1898496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.890757637599968e-05,
+      "loss": 3.8783,
+      "step": 1899008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.889919042848916e-05,
+      "loss": 3.8538,
+      "step": 1899520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.889080448097864e-05,
+      "loss": 3.8864,
+      "step": 1900032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8882418533468123e-05,
+      "loss": 3.8849,
+      "step": 1900544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8874032585957603e-05,
+      "loss": 3.8884,
+      "step": 1901056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8865646638447083e-05,
+      "loss": 3.8736,
+      "step": 1901568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.885726069093656e-05,
+      "loss": 3.8839,
+      "step": 1902080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.884887474342604e-05,
+      "loss": 3.8806,
+      "step": 1902592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8840505174719252e-05,
+      "loss": 3.8925,
+      "step": 1903104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8832135606012464e-05,
+      "loss": 3.8704,
+      "step": 1903616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8823749658501944e-05,
+      "loss": 3.8753,
+      "step": 1904128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8815363710991424e-05,
+      "loss": 3.8751,
+      "step": 1904640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8806977763480904e-05,
+      "loss": 3.8864,
+      "step": 1905152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8798608194774113e-05,
+      "loss": 3.8764,
+      "step": 1905664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8790222247263593e-05,
+      "loss": 3.8747,
+      "step": 1906176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8781836299753077e-05,
+      "loss": 3.8825,
+      "step": 1906688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8773450352242557e-05,
+      "loss": 3.875,
+      "step": 1907200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8765064404732033e-05,
+      "loss": 3.8834,
+      "step": 1907712
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.000287055969238,
+      "eval_runtime": 299.284,
+      "eval_samples_per_second": 1275.013,
+      "eval_steps_per_second": 39.845,
+      "step": 1908000
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 7.565553426794433e+17,
+  "trial_name": null,
+  "trial_params": null
+}