diff --git "a/checkpoint-2900160/trainer_state.json" "b/checkpoint-2900160/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2900160/trainer_state.json"
@@ -0,0 +1,34313 @@
+{
+  "best_metric": 3.9695701599121094,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/binding-c-command/lstm/4/checkpoints/checkpoint-2900160",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 2900160,
+  "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": 10.8192,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5546,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0633,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9997,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9488,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8868,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7197,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994131474623009e-05,
+      "loss": 6.6199,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993292879871958e-05,
+      "loss": 6.5189,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992454285120906e-05,
+      "loss": 6.4585,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.3972,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.3409,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.2844,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.2104,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.163,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0983,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0593,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 6.0109,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.963,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9268,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8896,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8531,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8222,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7837,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979875363855125e-05,
+      "loss": 5.767,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979036769104073e-05,
+      "loss": 5.7369,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.978198174353021e-05,
+      "loss": 5.7061,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6852,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97652262273129e-05,
+      "loss": 5.6579,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.975684027980238e-05,
+      "loss": 5.6402,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974845433229186e-05,
+      "loss": 5.6214,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.6,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5772,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.5553,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714926921053515e-05,
+      "loss": 5.5515,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 5.5412,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.5142,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4741,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4773,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 5.4474,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 5.4469,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 5.4298,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964790485618428e-05,
+      "loss": 5.4184,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963951890867376e-05,
+      "loss": 5.3919,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 5.397,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 5.3829,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9614377444945934e-05,
+      "loss": 5.3809,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9605991497435414e-05,
+      "loss": 5.3646,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9597605549924894e-05,
+      "loss": 5.3432,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 5.3366,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 5.3375,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.3155,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.3068,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.293,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.2897,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.2728,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2866,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2482,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2549,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2468,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2262,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.2275,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.215,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.1982,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.1911,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.2058,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.1829,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.174,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942991935732195e-05,
+      "loss": 5.1564,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1628,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1695,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.1521,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.147,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938800599857309e-05,
+      "loss": 5.1519,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1414,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1303,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1083,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.1152,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.0977,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.0969,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.0885,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0866,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0953,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0829,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.0585,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0579,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0611,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0572,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92622823011302e-05,
+      "loss": 5.0467,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0411,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0338,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0205,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922875488989185e-05,
+      "loss": 5.0268,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0192,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 5.0104,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.0165,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919522747865351e-05,
+      "loss": 4.9969,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918684153114299e-05,
+      "loss": 4.9935,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917845558363247e-05,
+      "loss": 4.9826,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917006963612195e-05,
+      "loss": 4.979,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.9876,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.9707,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.9678,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136542224883596e-05,
+      "loss": 4.962,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.9604,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.9446,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.932,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.9386,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.9442,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.9354,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907787334991742e-05,
+      "loss": 4.9324,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90694874024069e-05,
+      "loss": 4.917,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.924,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905271550738586e-05,
+      "loss": 4.9098,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904432955987534e-05,
+      "loss": 4.9135,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.9087,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.9069,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901920447495124e-05,
+      "loss": 4.8982,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901081852744072e-05,
+      "loss": 4.8929,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.8979,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.8973,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.883,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.8866,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968888789888125e-05,
+      "loss": 4.8754,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960502842377605e-05,
+      "loss": 4.8785,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.8588,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.8529,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.8583,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.8591,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8488,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.8603,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.8523,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.836,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8408,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.8459,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.8364,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885992060866256e-05,
+      "loss": 4.833,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885153466115204e-05,
+      "loss": 4.823,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884314871364151e-05,
+      "loss": 4.8269,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883476276613099e-05,
+      "loss": 4.83,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882637681862047e-05,
+      "loss": 4.812,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881799087110995e-05,
+      "loss": 4.806,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.8123,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.8038,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879284940738212e-05,
+      "loss": 4.8004,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.804,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.7949,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.793,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.7917,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7856,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.751091957092285,
+      "eval_runtime": 300.2998,
+      "eval_samples_per_second": 1270.7,
+      "eval_steps_per_second": 39.71,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8742566479926466e-05,
+      "loss": 4.7785,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.7829,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8725794584905426e-05,
+      "loss": 4.7896,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717425016198635e-05,
+      "loss": 4.7782,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709039068688115e-05,
+      "loss": 4.7761,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700653121177595e-05,
+      "loss": 4.7704,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692267173667075e-05,
+      "loss": 4.7738,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683881226156555e-05,
+      "loss": 4.752,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867551165744977e-05,
+      "loss": 4.7759,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866712570993925e-05,
+      "loss": 4.7613,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865873976242873e-05,
+      "loss": 4.7683,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865035381491821e-05,
+      "loss": 4.7653,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864198424621142e-05,
+      "loss": 4.746,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86335982987009e-05,
+      "loss": 4.7492,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862521235119038e-05,
+      "loss": 4.7424,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.737,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.7447,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860007088746255e-05,
+      "loss": 4.7314,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859168493995203e-05,
+      "loss": 4.7305,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.7481,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.7319,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.7253,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.7188,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549755202399435e-05,
+      "loss": 4.7353,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.7133,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532999686182124e-05,
+      "loss": 4.7088,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852463011747533e-05,
+      "loss": 4.7095,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851624416996481e-05,
+      "loss": 4.712,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.7042,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.7032,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.7057,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.7044,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847431443241221e-05,
+      "loss": 4.6915,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846592848490169e-05,
+      "loss": 4.7043,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.7052,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844917296868439e-05,
+      "loss": 4.6979,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844078702117387e-05,
+      "loss": 4.6696,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843240107366335e-05,
+      "loss": 4.6797,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842401512615282e-05,
+      "loss": 4.6763,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84156291786423e-05,
+      "loss": 4.6817,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840725960993552e-05,
+      "loss": 4.6735,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398873662425e-05,
+      "loss": 4.6766,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839048771491447e-05,
+      "loss": 4.6646,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838210176740395e-05,
+      "loss": 4.6778,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373732198697167e-05,
+      "loss": 4.6731,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365346251186646e-05,
+      "loss": 4.674,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8356960303676126e-05,
+      "loss": 4.6729,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8348574356165606e-05,
+      "loss": 4.654,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8340188408655086e-05,
+      "loss": 4.6624,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331818839948295e-05,
+      "loss": 4.6669,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323432892437775e-05,
+      "loss": 4.6589,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315046944927255e-05,
+      "loss": 4.6557,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306660997416735e-05,
+      "loss": 4.6458,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298275049906215e-05,
+      "loss": 4.6525,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289905481199424e-05,
+      "loss": 4.6352,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281519533688904e-05,
+      "loss": 4.6662,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273133586178384e-05,
+      "loss": 4.6376,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826474763866787e-05,
+      "loss": 4.6482,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825636169115735e-05,
+      "loss": 4.6457,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824797574364683e-05,
+      "loss": 4.6238,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823958979613631e-05,
+      "loss": 4.6382,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823120384862579e-05,
+      "loss": 4.632,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222834279919e-05,
+      "loss": 4.6184,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821446471121221e-05,
+      "loss": 4.6188,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820607876370169e-05,
+      "loss": 4.6366,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819769281619117e-05,
+      "loss": 4.6228,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818930686868065e-05,
+      "loss": 4.6158,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818092092117013e-05,
+      "loss": 4.6097,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817253497365961e-05,
+      "loss": 4.6221,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816414902614909e-05,
+      "loss": 4.6296,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155779457442305e-05,
+      "loss": 4.621,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147409888735514e-05,
+      "loss": 4.6167,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8139023941224994e-05,
+      "loss": 4.6294,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8130637993714474e-05,
+      "loss": 4.6214,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8122252046203954e-05,
+      "loss": 4.6191,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811388247749716e-05,
+      "loss": 4.6,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810549652998664e-05,
+      "loss": 4.6118,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809712696127985e-05,
+      "loss": 4.6042,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808874101376933e-05,
+      "loss": 4.601,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808035506625881e-05,
+      "loss": 4.6007,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807196911874829e-05,
+      "loss": 4.6006,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806358317123777e-05,
+      "loss": 4.6155,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805519722372726e-05,
+      "loss": 4.6097,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804681127621674e-05,
+      "loss": 4.5858,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803842532870622e-05,
+      "loss": 4.5906,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80300393811957e-05,
+      "loss": 4.5931,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802165343368518e-05,
+      "loss": 4.5946,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801326748617466e-05,
+      "loss": 4.5941,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800489791746787e-05,
+      "loss": 4.5922,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996528348761076e-05,
+      "loss": 4.5854,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988142401250556e-05,
+      "loss": 4.5761,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979756453740036e-05,
+      "loss": 4.5875,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971370506229516e-05,
+      "loss": 4.5789,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7962984558718996e-05,
+      "loss": 4.5806,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954598611208476e-05,
+      "loss": 4.5862,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946212663697956e-05,
+      "loss": 4.5772,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937826716187436e-05,
+      "loss": 4.5711,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929440768676916e-05,
+      "loss": 4.5701,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7921054821166396e-05,
+      "loss": 4.5722,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912668873655876e-05,
+      "loss": 4.578,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7904299304949085e-05,
+      "loss": 4.5629,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895913357438565e-05,
+      "loss": 4.5638,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887527409928045e-05,
+      "loss": 4.5635,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879141462417525e-05,
+      "loss": 4.5631,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870771893710734e-05,
+      "loss": 4.5515,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862385946200214e-05,
+      "loss": 4.5395,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853999998689694e-05,
+      "loss": 4.5547,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784561405117918e-05,
+      "loss": 4.5577,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783722810366866e-05,
+      "loss": 4.5537,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782884215615814e-05,
+      "loss": 4.5529,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782047258745135e-05,
+      "loss": 4.5466,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781208663994083e-05,
+      "loss": 4.5524,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780370069243031e-05,
+      "loss": 4.5413,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779531474491979e-05,
+      "loss": 4.5449,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778692879740927e-05,
+      "loss": 4.5501,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777855922870248e-05,
+      "loss": 4.5488,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777017328119196e-05,
+      "loss": 4.5409,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776178733368144e-05,
+      "loss": 4.5392,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775340138617092e-05,
+      "loss": 4.5476,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77450154386604e-05,
+      "loss": 4.5479,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773662949114988e-05,
+      "loss": 4.532,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772824354363936e-05,
+      "loss": 4.5435,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719873974932574e-05,
+      "loss": 4.5326,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7711488027422054e-05,
+      "loss": 4.5438,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703102079911534e-05,
+      "loss": 4.5216,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694716132401014e-05,
+      "loss": 4.5202,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768634656369422e-05,
+      "loss": 4.5257,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767797699498743e-05,
+      "loss": 4.5323,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766959104747691e-05,
+      "loss": 4.5206,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766122147877012e-05,
+      "loss": 4.5402,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76528355312596e-05,
+      "loss": 4.5286,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764444958374909e-05,
+      "loss": 4.5211,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763606363623857e-05,
+      "loss": 4.5204,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762767768872805e-05,
+      "loss": 4.5328,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761929174121753e-05,
+      "loss": 4.5239,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761090579370701e-05,
+      "loss": 4.5218,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760251984619649e-05,
+      "loss": 4.5165,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759413389868596e-05,
+      "loss": 4.5188,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758574795117544e-05,
+      "loss": 4.5244,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757736200366492e-05,
+      "loss": 4.5112,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75689760561544e-05,
+      "loss": 4.5099,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756060648744761e-05,
+      "loss": 4.5144,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552236918740826e-05,
+      "loss": 4.5056,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543850971230306e-05,
+      "loss": 4.5059,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535465023719786e-05,
+      "loss": 4.5146,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527079076209266e-05,
+      "loss": 4.5051,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751870950750248e-05,
+      "loss": 4.5043,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751032355999196e-05,
+      "loss": 4.5037,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7501937612481435e-05,
+      "loss": 4.4985,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.47037410736084,
+      "eval_runtime": 311.1607,
+      "eval_samples_per_second": 1226.347,
+      "eval_steps_per_second": 38.324,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493551664970914e-05,
+      "loss": 4.4925,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485165717460394e-05,
+      "loss": 4.499,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476779769949874e-05,
+      "loss": 4.5113,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468393822439354e-05,
+      "loss": 4.5004,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460007874928834e-05,
+      "loss": 4.5007,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451621927418314e-05,
+      "loss": 4.4944,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443235979907794e-05,
+      "loss": 4.5018,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434850032397274e-05,
+      "loss": 4.4808,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426464084886754e-05,
+      "loss": 4.5026,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418078137376234e-05,
+      "loss": 4.5,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740969218986572e-05,
+      "loss": 4.4974,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74013062423552e-05,
+      "loss": 4.5047,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739293667364841e-05,
+      "loss": 4.4807,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738455072613789e-05,
+      "loss": 4.4853,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737616477862737e-05,
+      "loss": 4.4919,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736777883111685e-05,
+      "loss": 4.4716,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735939288360632e-05,
+      "loss": 4.485,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73510069360958e-05,
+      "loss": 4.4749,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734262098858528e-05,
+      "loss": 4.4756,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73342514198785e-05,
+      "loss": 4.4993,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732586547236797e-05,
+      "loss": 4.4782,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731747952485746e-05,
+      "loss": 4.4803,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730909357734694e-05,
+      "loss": 4.4678,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730070762983642e-05,
+      "loss": 4.4878,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.72923216823259e-05,
+      "loss": 4.4638,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728395211361911e-05,
+      "loss": 4.4664,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727556616610859e-05,
+      "loss": 4.4653,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726718021859807e-05,
+      "loss": 4.4743,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725879427108755e-05,
+      "loss": 4.4639,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725040832357703e-05,
+      "loss": 4.4626,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724202237606651e-05,
+      "loss": 4.4702,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723363642855599e-05,
+      "loss": 4.4717,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722525048104547e-05,
+      "loss": 4.4538,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721686453353495e-05,
+      "loss": 4.4713,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208494964828156e-05,
+      "loss": 4.4707,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720010901731764e-05,
+      "loss": 4.4636,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719172306980712e-05,
+      "loss": 4.4426,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71833371222966e-05,
+      "loss": 4.455,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717496755358981e-05,
+      "loss": 4.4434,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716658160607929e-05,
+      "loss": 4.4592,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715819565856877e-05,
+      "loss": 4.4507,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714980971105825e-05,
+      "loss": 4.4555,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714142376354773e-05,
+      "loss": 4.4423,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713305419484094e-05,
+      "loss": 4.4536,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712466824733042e-05,
+      "loss": 4.4541,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71162822998199e-05,
+      "loss": 4.4561,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710789635230938e-05,
+      "loss": 4.4567,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709952678360259e-05,
+      "loss": 4.4334,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709114083609208e-05,
+      "loss": 4.444,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708275488858156e-05,
+      "loss": 4.4561,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074368941071037e-05,
+      "loss": 4.4394,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7065982993560517e-05,
+      "loss": 4.4427,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705759704604999e-05,
+      "loss": 4.4374,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7049227477343206e-05,
+      "loss": 4.4424,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040841529832686e-05,
+      "loss": 4.43,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703245558232216e-05,
+      "loss": 4.449,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702406963481164e-05,
+      "loss": 4.435,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015700066104855e-05,
+      "loss": 4.4386,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007314118594335e-05,
+      "loss": 4.445,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998928171083814e-05,
+      "loss": 4.4187,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990542223573294e-05,
+      "loss": 4.4343,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982156276062774e-05,
+      "loss": 4.4292,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973770328552254e-05,
+      "loss": 4.4202,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965400759845463e-05,
+      "loss": 4.4177,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6957014812334943e-05,
+      "loss": 4.4322,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694862886482442e-05,
+      "loss": 4.4277,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69402429173139e-05,
+      "loss": 4.4167,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693185696980338e-05,
+      "loss": 4.4118,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692347102229286e-05,
+      "loss": 4.4289,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691508507478234e-05,
+      "loss": 4.4372,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690671550607555e-05,
+      "loss": 4.4239,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689832955856503e-05,
+      "loss": 4.4243,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688994361105451e-05,
+      "loss": 4.4362,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881557663544e-05,
+      "loss": 4.4297,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687317171603348e-05,
+      "loss": 4.4299,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686478576852296e-05,
+      "loss": 4.4061,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685639982101244e-05,
+      "loss": 4.4233,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684801387350192e-05,
+      "loss": 4.4239,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683966068359886e-05,
+      "loss": 4.4097,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831291114892066e-05,
+      "loss": 4.4174,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822905167381546e-05,
+      "loss": 4.4153,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6814519219871026e-05,
+      "loss": 4.4309,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806133272360506e-05,
+      "loss": 4.4222,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6797747324849986e-05,
+      "loss": 4.4028,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789361377339466e-05,
+      "loss": 4.4082,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678097542982895e-05,
+      "loss": 4.4114,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677258948231843e-05,
+      "loss": 4.4152,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676420353480791e-05,
+      "loss": 4.4159,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675583396610112e-05,
+      "loss": 4.41,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.67474480185906e-05,
+      "loss": 4.407,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673906207108008e-05,
+      "loss": 4.3993,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673067612356956e-05,
+      "loss": 4.41,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672229017605904e-05,
+      "loss": 4.3993,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671390422854852e-05,
+      "loss": 4.4083,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705518281037995e-05,
+      "loss": 4.412,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697132333527474e-05,
+      "loss": 4.4043,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668877914362442e-05,
+      "loss": 4.3975,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680393196113906e-05,
+      "loss": 4.3979,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6672007248603386e-05,
+      "loss": 4.4008,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663621301092866e-05,
+      "loss": 4.4042,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655251732386075e-05,
+      "loss": 4.3939,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6646865784875555e-05,
+      "loss": 4.3935,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6638496216168764e-05,
+      "loss": 4.3897,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6630110268658244e-05,
+      "loss": 4.3962,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621724321147724e-05,
+      "loss": 4.386,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613338373637204e-05,
+      "loss": 4.3751,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604952426126684e-05,
+      "loss": 4.3864,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596566478616164e-05,
+      "loss": 4.3938,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588180531105644e-05,
+      "loss": 4.3861,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579794583595124e-05,
+      "loss": 4.393,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571408636084604e-05,
+      "loss": 4.3817,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563022688574084e-05,
+      "loss": 4.3908,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554636741063564e-05,
+      "loss": 4.3802,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546250793553044e-05,
+      "loss": 4.38,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653788122484625e-05,
+      "loss": 4.3948,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652951165613947e-05,
+      "loss": 4.3902,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652112570862894e-05,
+      "loss": 4.3827,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651273976111842e-05,
+      "loss": 4.383,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65043538136079e-05,
+      "loss": 4.3837,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649596786609738e-05,
+      "loss": 4.3886,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648758191858686e-05,
+      "loss": 4.3756,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647919597107634e-05,
+      "loss": 4.388,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647081002356582e-05,
+      "loss": 4.3758,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646245683366277e-05,
+      "loss": 4.3876,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645407088615225e-05,
+      "loss": 4.364,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644568493864173e-05,
+      "loss": 4.3723,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437298991131207e-05,
+      "loss": 4.3692,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6428913043620687e-05,
+      "loss": 4.3767,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420543474913896e-05,
+      "loss": 4.3696,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412157527403376e-05,
+      "loss": 4.3893,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640378795869659e-05,
+      "loss": 4.3768,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395402011186065e-05,
+      "loss": 4.3719,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387016063675545e-05,
+      "loss": 4.3669,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637863011616503e-05,
+      "loss": 4.3816,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637024416865451e-05,
+      "loss": 4.3754,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636185822114399e-05,
+      "loss": 4.3784,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635347227363347e-05,
+      "loss": 4.3686,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634508632612295e-05,
+      "loss": 4.3699,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633671675741616e-05,
+      "loss": 4.3774,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632833080990564e-05,
+      "loss": 4.3649,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631994486239512e-05,
+      "loss": 4.3631,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63115589148846e-05,
+      "loss": 4.3728,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630318934617781e-05,
+      "loss": 4.3623,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629480339866729e-05,
+      "loss": 4.359,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628641745115677e-05,
+      "loss": 4.3652,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627803150364625e-05,
+      "loss": 4.368,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626964555613573e-05,
+      "loss": 4.36,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6261275987428945e-05,
+      "loss": 4.3584,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6252890039918425e-05,
+      "loss": 4.3544,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.335749626159668,
+      "eval_runtime": 293.658,
+      "eval_samples_per_second": 1299.44,
+      "eval_steps_per_second": 40.608,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244504092407905e-05,
+      "loss": 4.3614,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6236118144897385e-05,
+      "loss": 4.3551,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227748576190594e-05,
+      "loss": 4.3721,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219362628680074e-05,
+      "loss": 4.3579,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210976681169554e-05,
+      "loss": 4.3628,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202590733659034e-05,
+      "loss": 4.3549,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194204786148514e-05,
+      "loss": 4.3618,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185818838637994e-05,
+      "loss": 4.3455,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177432891127474e-05,
+      "loss": 4.3666,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6169046943616954e-05,
+      "loss": 4.3613,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160660996106434e-05,
+      "loss": 4.3602,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152275048595913e-05,
+      "loss": 4.368,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614392185869286e-05,
+      "loss": 4.3445,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613553591118234e-05,
+      "loss": 4.3508,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612714996367182e-05,
+      "loss": 4.3571,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61187640161613e-05,
+      "loss": 4.3428,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611037806865078e-05,
+      "loss": 4.3458,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610199212114025e-05,
+      "loss": 4.3415,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609360617362973e-05,
+      "loss": 4.3438,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608522022611921e-05,
+      "loss": 4.368,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607683427860869e-05,
+      "loss": 4.3469,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606844833109817e-05,
+      "loss": 4.3454,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606007876239139e-05,
+      "loss": 4.3394,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605169281488087e-05,
+      "loss": 4.352,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604330686737035e-05,
+      "loss": 4.3372,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603492091985983e-05,
+      "loss": 4.3348,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602653497234931e-05,
+      "loss": 4.3375,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601814902483879e-05,
+      "loss": 4.3454,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009779456131996e-05,
+      "loss": 4.3318,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001393508621476e-05,
+      "loss": 4.3374,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5993007561110956e-05,
+      "loss": 4.339,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5984621613600436e-05,
+      "loss": 4.3438,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976235666089916e-05,
+      "loss": 4.3339,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967866097383125e-05,
+      "loss": 4.3386,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959480149872605e-05,
+      "loss": 4.3429,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595109420236209e-05,
+      "loss": 4.3396,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594270825485157e-05,
+      "loss": 4.3178,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593432230734105e-05,
+      "loss": 4.3343,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592593635983053e-05,
+      "loss": 4.3187,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591756679112374e-05,
+      "loss": 4.3361,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590918084361322e-05,
+      "loss": 4.3244,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59007948961027e-05,
+      "loss": 4.3334,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589240894859218e-05,
+      "loss": 4.3245,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588402300108166e-05,
+      "loss": 4.3279,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587563705357114e-05,
+      "loss": 4.3341,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867251106060614e-05,
+      "loss": 4.3304,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858865158550094e-05,
+      "loss": 4.3328,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585049558984331e-05,
+      "loss": 4.3194,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5842126021136525e-05,
+      "loss": 4.32,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5833756452429734e-05,
+      "loss": 4.3357,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5825370504919214e-05,
+      "loss": 4.3223,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5816984557408694e-05,
+      "loss": 4.3196,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808598609898174e-05,
+      "loss": 4.3206,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800212662387654e-05,
+      "loss": 4.3248,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791826714877134e-05,
+      "loss": 4.3101,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578344076736661e-05,
+      "loss": 4.3291,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577507119865982e-05,
+      "loss": 4.3178,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57666852511493e-05,
+      "loss": 4.3215,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575829930363878e-05,
+      "loss": 4.3259,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574991335612826e-05,
+      "loss": 4.3086,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574152740861774e-05,
+      "loss": 4.3117,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573314146110722e-05,
+      "loss": 4.3178,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57247555135967e-05,
+      "loss": 4.3048,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571636956608618e-05,
+      "loss": 4.3051,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570798361857566e-05,
+      "loss": 4.3144,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569961404986887e-05,
+      "loss": 4.3162,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569122810235835e-05,
+      "loss": 4.3021,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568284215484783e-05,
+      "loss": 4.2994,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567445620733731e-05,
+      "loss": 4.3144,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566607025982679e-05,
+      "loss": 4.3233,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565770069112e-05,
+      "loss": 4.3117,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564931474360948e-05,
+      "loss": 4.3163,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564092879609896e-05,
+      "loss": 4.3202,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563254284858845e-05,
+      "loss": 4.32,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624173279881657e-05,
+      "loss": 4.321,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615787332371137e-05,
+      "loss": 4.2944,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5607401384860617e-05,
+      "loss": 4.3118,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5599015437350096e-05,
+      "loss": 4.3175,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5590629489839576e-05,
+      "loss": 4.2982,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582259921132786e-05,
+      "loss": 4.3137,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5573873973622265e-05,
+      "loss": 4.3046,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5565488026111745e-05,
+      "loss": 4.3194,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557102078601225e-05,
+      "loss": 4.3163,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548716131090705e-05,
+      "loss": 4.2924,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5540330183580185e-05,
+      "loss": 4.3038,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5531944236069665e-05,
+      "loss": 4.3011,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523558288559145e-05,
+      "loss": 4.3061,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551518871985236e-05,
+      "loss": 4.3091,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550680277234184e-05,
+      "loss": 4.3018,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549841682483132e-05,
+      "loss": 4.3003,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54900308773208e-05,
+      "loss": 4.2943,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548166130861401e-05,
+      "loss": 4.302,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547327536110349e-05,
+      "loss": 4.2929,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54649057923967e-05,
+      "loss": 4.304,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545651984488618e-05,
+      "loss": 4.3085,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544813389737566e-05,
+      "loss": 4.2985,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543974794986514e-05,
+      "loss": 4.2937,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543136200235462e-05,
+      "loss": 4.2971,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54229760548441e-05,
+      "loss": 4.2947,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541459010733358e-05,
+      "loss": 4.2998,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540620415982306e-05,
+      "loss": 4.2957,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397834591116275e-05,
+      "loss": 4.2911,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538944864360575e-05,
+      "loss": 4.2838,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538106269609523e-05,
+      "loss": 4.2972,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537267674858471e-05,
+      "loss": 4.2774,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536429080107419e-05,
+      "loss": 4.2762,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53559212323674e-05,
+      "loss": 4.2828,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534755166366061e-05,
+      "loss": 4.29,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533916571615009e-05,
+      "loss": 4.2858,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533077976863957e-05,
+      "loss": 4.2986,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532239382112905e-05,
+      "loss": 4.2828,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531400787361853e-05,
+      "loss": 4.2829,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530562192610801e-05,
+      "loss": 4.2853,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529723597859749e-05,
+      "loss": 4.2757,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52888664098907e-05,
+      "loss": 4.2974,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528048046238018e-05,
+      "loss": 4.2903,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527209451486966e-05,
+      "loss": 4.2842,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526370856735914e-05,
+      "loss": 4.2865,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525532261984862e-05,
+      "loss": 4.2775,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52469366723381e-05,
+      "loss": 4.2909,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523856710363131e-05,
+      "loss": 4.2786,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523018115612079e-05,
+      "loss": 4.2947,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522179520861027e-05,
+      "loss": 4.2768,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521340926109976e-05,
+      "loss": 4.2852,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520502331358924e-05,
+      "loss": 4.2727,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519663736607872e-05,
+      "loss": 4.2756,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51882514185682e-05,
+      "loss": 4.2706,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517986547105768e-05,
+      "loss": 4.2829,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517147952354716e-05,
+      "loss": 4.2715,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5163109954840366e-05,
+      "loss": 4.291,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154724007329846e-05,
+      "loss": 4.2852,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146338059819326e-05,
+      "loss": 4.2791,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5137952112308806e-05,
+      "loss": 4.2679,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512956616479828e-05,
+      "loss": 4.286,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512118021728776e-05,
+      "loss": 4.2797,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112810648580975e-05,
+      "loss": 4.2872,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104424701070455e-05,
+      "loss": 4.2748,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096038753559935e-05,
+      "loss": 4.2738,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087652806049415e-05,
+      "loss": 4.2824,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507928323734263e-05,
+      "loss": 4.2788,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070897289832104e-05,
+      "loss": 4.2672,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062511342321584e-05,
+      "loss": 4.2788,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5054125394811064e-05,
+      "loss": 4.269,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5045739447300544e-05,
+      "loss": 4.2685,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503736987859375e-05,
+      "loss": 4.2737,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502898393108323e-05,
+      "loss": 4.2784,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502059798357271e-05,
+      "loss": 4.2646,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501221203606219e-05,
+      "loss": 4.2643,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500382608855168e-05,
+      "loss": 4.2708,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.252523899078369,
+      "eval_runtime": 307.6907,
+      "eval_samples_per_second": 1240.177,
+      "eval_steps_per_second": 38.756,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499544014104116e-05,
+      "loss": 4.2654,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498705419353064e-05,
+      "loss": 4.2663,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497866824602012e-05,
+      "loss": 4.2789,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49702822985096e-05,
+      "loss": 4.2665,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496189635099908e-05,
+      "loss": 4.278,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495351040348856e-05,
+      "loss": 4.2646,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494512445597804e-05,
+      "loss": 4.2689,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493673850846752e-05,
+      "loss": 4.2573,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492835256095699e-05,
+      "loss": 4.2801,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491996661344647e-05,
+      "loss": 4.2707,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491158066593595e-05,
+      "loss": 4.2691,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490319471842543e-05,
+      "loss": 4.2816,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489482514971865e-05,
+      "loss": 4.2533,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488643920220813e-05,
+      "loss": 4.2659,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487805325469761e-05,
+      "loss": 4.2675,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869683685990824e-05,
+      "loss": 4.2561,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48612977384803e-05,
+      "loss": 4.258,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485291179096978e-05,
+      "loss": 4.2544,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484452584345926e-05,
+      "loss": 4.2597,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483613989594874e-05,
+      "loss": 4.2763,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4827753948438217e-05,
+      "loss": 4.2597,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4819368000927697e-05,
+      "loss": 4.2597,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4810982053417177e-05,
+      "loss": 4.2537,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4802596105906656e-05,
+      "loss": 4.2653,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4794210158396136e-05,
+      "loss": 4.2524,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4785824210885616e-05,
+      "loss": 4.2515,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4777438263375096e-05,
+      "loss": 4.2501,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4769052315864576e-05,
+      "loss": 4.2587,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4760666368354056e-05,
+      "loss": 4.2473,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475229679964727e-05,
+      "loss": 4.2546,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474391085213675e-05,
+      "loss": 4.2539,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473552490462623e-05,
+      "loss": 4.2607,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472713895711571e-05,
+      "loss": 4.2511,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471876938840892e-05,
+      "loss": 4.256,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47103834408984e-05,
+      "loss": 4.2599,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470199749338788e-05,
+      "loss": 4.2558,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693611545877354e-05,
+      "loss": 4.2392,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685225598366834e-05,
+      "loss": 4.2486,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676839650856314e-05,
+      "loss": 4.2363,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466847008214953e-05,
+      "loss": 4.2505,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466008413463901e-05,
+      "loss": 4.2442,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465169818712849e-05,
+      "loss": 4.2512,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464331223961797e-05,
+      "loss": 4.243,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463492629210745e-05,
+      "loss": 4.2495,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462654034459693e-05,
+      "loss": 4.2505,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461815439708641e-05,
+      "loss": 4.2492,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460978482837962e-05,
+      "loss": 4.2525,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46013988808691e-05,
+      "loss": 4.2418,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459302931216231e-05,
+      "loss": 4.2321,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458464336465179e-05,
+      "loss": 4.2565,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457625741714127e-05,
+      "loss": 4.2431,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456787146963075e-05,
+      "loss": 4.2403,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559501900923964e-05,
+      "loss": 4.2407,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551115953413444e-05,
+      "loss": 4.2455,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542730005902923e-05,
+      "loss": 4.2305,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453436043719613e-05,
+      "loss": 4.2484,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452597448968561e-05,
+      "loss": 4.2374,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451758854217509e-05,
+      "loss": 4.2405,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450920259466457e-05,
+      "loss": 4.2481,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450081664715405e-05,
+      "loss": 4.2335,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449243069964353e-05,
+      "loss": 4.2302,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448404475213301e-05,
+      "loss": 4.2448,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447565880462249e-05,
+      "loss": 4.2218,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446727285711197e-05,
+      "loss": 4.2276,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445888690960145e-05,
+      "loss": 4.2377,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445050096209093e-05,
+      "loss": 4.2388,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444211501458041e-05,
+      "loss": 4.2254,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44337290670699e-05,
+      "loss": 4.2241,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442534311955937e-05,
+      "loss": 4.2345,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441695717204885e-05,
+      "loss": 4.243,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440858760334207e-05,
+      "loss": 4.2375,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440020165583154e-05,
+      "loss": 4.2372,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439183208712476e-05,
+      "loss": 4.245,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438344613961424e-05,
+      "loss": 4.2403,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437506019210372e-05,
+      "loss": 4.2498,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436667424459319e-05,
+      "loss": 4.2194,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435828829708267e-05,
+      "loss": 4.2367,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434990234957216e-05,
+      "loss": 4.2401,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434151640206164e-05,
+      "loss": 4.221,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333130454551117e-05,
+      "loss": 4.2413,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4324760885844326e-05,
+      "loss": 4.2288,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431639131713754e-05,
+      "loss": 4.2423,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308005369627015e-05,
+      "loss": 4.2446,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299619422116495e-05,
+      "loss": 4.2184,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291233474605975e-05,
+      "loss": 4.2316,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4282847527095455e-05,
+      "loss": 4.2226,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274461579584935e-05,
+      "loss": 4.2345,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266092010878144e-05,
+      "loss": 4.2343,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425772244217136e-05,
+      "loss": 4.2265,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424933649466084e-05,
+      "loss": 4.2284,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424095054715032e-05,
+      "loss": 4.2239,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42325645996398e-05,
+      "loss": 4.2266,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422417865212928e-05,
+      "loss": 4.2197,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421579270461876e-05,
+      "loss": 4.2341,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420742313591197e-05,
+      "loss": 4.2347,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4199053567205184e-05,
+      "loss": 4.2261,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190667619694664e-05,
+      "loss": 4.2214,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418228167218414e-05,
+      "loss": 4.2212,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417389572467362e-05,
+      "loss": 4.2254,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41655097771631e-05,
+      "loss": 4.2296,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415712382965258e-05,
+      "loss": 4.2247,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414873788214206e-05,
+      "loss": 4.2198,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4140351934631544e-05,
+      "loss": 4.2122,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4131965987121024e-05,
+      "loss": 4.2232,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123580039610504e-05,
+      "loss": 4.2059,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115194092099984e-05,
+      "loss": 4.211,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4106808144589464e-05,
+      "loss": 4.2071,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409843857588267e-05,
+      "loss": 4.2217,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409005262837215e-05,
+      "loss": 4.2102,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408166668086163e-05,
+      "loss": 4.2301,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407328073335111e-05,
+      "loss": 4.211,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406491116464432e-05,
+      "loss": 4.2157,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40565252171338e-05,
+      "loss": 4.2166,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404813926962328e-05,
+      "loss": 4.204,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403975332211276e-05,
+      "loss": 4.2273,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403136737460224e-05,
+      "loss": 4.2238,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402299780589546e-05,
+      "loss": 4.2163,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401461185838494e-05,
+      "loss": 4.2155,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400622591087442e-05,
+      "loss": 4.2092,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39978399633639e-05,
+      "loss": 4.2228,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398945401585338e-05,
+      "loss": 4.2051,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398106806834285e-05,
+      "loss": 4.2251,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397268212083233e-05,
+      "loss": 4.2118,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396429617332181e-05,
+      "loss": 4.2143,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395591022581129e-05,
+      "loss": 4.2072,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394752427830077e-05,
+      "loss": 4.2077,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393913833079025e-05,
+      "loss": 4.2003,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393075238327973e-05,
+      "loss": 4.2162,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922382814572946e-05,
+      "loss": 4.2054,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3913996867062426e-05,
+      "loss": 4.2221,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905610919551906e-05,
+      "loss": 4.2166,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897224972041386e-05,
+      "loss": 4.2137,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3888855403334595e-05,
+      "loss": 4.2004,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880485834627804e-05,
+      "loss": 4.2173,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872099887117284e-05,
+      "loss": 4.2154,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863713939606764e-05,
+      "loss": 4.2194,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855327992096244e-05,
+      "loss": 4.2077,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846942044585724e-05,
+      "loss": 4.2064,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838556097075204e-05,
+      "loss": 4.215,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830170149564684e-05,
+      "loss": 4.2155,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821784202054164e-05,
+      "loss": 4.1996,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813398254543644e-05,
+      "loss": 4.2141,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380502868583686e-05,
+      "loss": 4.2005,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379664273832634e-05,
+      "loss": 4.2032,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378825679081582e-05,
+      "loss": 4.2132,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37798708433053e-05,
+      "loss": 4.2054,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377148489579478e-05,
+      "loss": 4.2025,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376311532708799e-05,
+      "loss": 4.2005,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375472937957747e-05,
+      "loss": 4.2053,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.195892810821533,
+      "eval_runtime": 293.9942,
+      "eval_samples_per_second": 1297.954,
+      "eval_steps_per_second": 40.562,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374634343206695e-05,
+      "loss": 4.2045,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373795748455643e-05,
+      "loss": 4.1997,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372957153704591e-05,
+      "loss": 4.2153,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372118558953539e-05,
+      "loss": 4.2058,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371279964202487e-05,
+      "loss": 4.2073,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370441369451435e-05,
+      "loss": 4.2051,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3696044125807564e-05,
+      "loss": 4.2005,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368765817829704e-05,
+      "loss": 4.1948,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367927223078652e-05,
+      "loss": 4.2097,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3670886283276e-05,
+      "loss": 4.2114,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366250033576548e-05,
+      "loss": 4.2056,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365411438825496e-05,
+      "loss": 4.2163,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3645744819548166e-05,
+      "loss": 4.192,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3637358872037646e-05,
+      "loss": 4.2013,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3628972924527126e-05,
+      "loss": 4.2081,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3620586977016606e-05,
+      "loss": 4.1896,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3612201029506086e-05,
+      "loss": 4.1959,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3603815081995566e-05,
+      "loss": 4.1932,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359544551328878e-05,
+      "loss": 4.1964,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358705956577826e-05,
+      "loss": 4.2105,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357867361826774e-05,
+      "loss": 4.2015,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357028767075722e-05,
+      "loss": 4.2018,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35619017232467e-05,
+      "loss": 4.1897,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355351577573618e-05,
+      "loss": 4.2059,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354512982822566e-05,
+      "loss": 4.1899,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353676025951887e-05,
+      "loss": 4.1902,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352837431200835e-05,
+      "loss": 4.1921,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351998836449783e-05,
+      "loss": 4.1967,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351161879579104e-05,
+      "loss": 4.1851,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350323284828052e-05,
+      "loss": 4.193,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3494846900770006e-05,
+      "loss": 4.1964,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486460953259486e-05,
+      "loss": 4.1973,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478075005748966e-05,
+      "loss": 4.1938,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3469689058238446e-05,
+      "loss": 4.1936,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461303110727926e-05,
+      "loss": 4.1991,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34529171632174e-05,
+      "loss": 4.1917,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3444547594510615e-05,
+      "loss": 4.1836,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3436161647000095e-05,
+      "loss": 4.1854,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3427792078293304e-05,
+      "loss": 4.1812,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419406130782784e-05,
+      "loss": 4.1889,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3411020183272264e-05,
+      "loss": 4.1857,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402634235761744e-05,
+      "loss": 4.1925,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3394248288251224e-05,
+      "loss": 4.1842,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385862340740704e-05,
+      "loss": 4.1878,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377476393230184e-05,
+      "loss": 4.196,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3369090445719664e-05,
+      "loss": 4.1874,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3360704498209144e-05,
+      "loss": 4.1946,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3352318550698624e-05,
+      "loss": 4.1812,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3343932603188104e-05,
+      "loss": 4.1743,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333556303448131e-05,
+      "loss": 4.2018,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332717708697079e-05,
+      "loss": 4.1836,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331879113946027e-05,
+      "loss": 4.1824,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331040519194975e-05,
+      "loss": 4.181,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330201924443923e-05,
+      "loss": 4.189,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329363329692871e-05,
+      "loss": 4.1709,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328524734941819e-05,
+      "loss": 4.1888,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327687778071141e-05,
+      "loss": 4.1833,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326849183320089e-05,
+      "loss": 4.1799,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326010588569037e-05,
+      "loss": 4.1951,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325171993817985e-05,
+      "loss": 4.1729,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324333399066933e-05,
+      "loss": 4.1747,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323494804315881e-05,
+      "loss": 4.189,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322656209564829e-05,
+      "loss": 4.1693,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32181925269415e-05,
+      "loss": 4.1714,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320980657943098e-05,
+      "loss": 4.1768,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320142063192046e-05,
+      "loss": 4.1828,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319303468440994e-05,
+      "loss": 4.1721,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318464873689941e-05,
+      "loss": 4.1701,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176279168192626e-05,
+      "loss": 4.1729,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3167893220682106e-05,
+      "loss": 4.1846,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3159507273171586e-05,
+      "loss": 4.1808,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151121325661066e-05,
+      "loss": 4.1848,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3142735378150546e-05,
+      "loss": 4.1851,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313436580944376e-05,
+      "loss": 4.1852,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3125979861933235e-05,
+      "loss": 4.1946,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311761029322645e-05,
+      "loss": 4.162,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310922434571593e-05,
+      "loss": 4.1865,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310083839820541e-05,
+      "loss": 4.1893,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3092452450694884e-05,
+      "loss": 4.1597,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3084066503184364e-05,
+      "loss": 4.1875,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3075680555673844e-05,
+      "loss": 4.1741,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306729460816333e-05,
+      "loss": 4.187,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305892503945654e-05,
+      "loss": 4.1892,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305053909194602e-05,
+      "loss": 4.1664,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30421531444355e-05,
+      "loss": 4.1719,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303376719692498e-05,
+      "loss": 4.1701,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302538124941446e-05,
+      "loss": 4.1797,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301699530190394e-05,
+      "loss": 4.1768,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300862573319715e-05,
+      "loss": 4.1755,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300023978568663e-05,
+      "loss": 4.17,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299185383817611e-05,
+      "loss": 4.1716,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.298346789066559e-05,
+      "loss": 4.1723,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297508194315507e-05,
+      "loss": 4.1651,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296669599564455e-05,
+      "loss": 4.1832,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295831004813403e-05,
+      "loss": 4.1789,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2949924100623515e-05,
+      "loss": 4.1786,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2941538153112995e-05,
+      "loss": 4.1634,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2933152205602475e-05,
+      "loss": 4.1707,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292476625809195e-05,
+      "loss": 4.1705,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291638031058143e-05,
+      "loss": 4.174,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2908010741874644e-05,
+      "loss": 4.1735,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289964117316785e-05,
+      "loss": 4.167,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289125522565733e-05,
+      "loss": 4.1599,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288286927814681e-05,
+      "loss": 4.1708,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287448333063629e-05,
+      "loss": 4.1584,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2866097383125766e-05,
+      "loss": 4.1553,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285772781441898e-05,
+      "loss": 4.1522,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284934186690847e-05,
+      "loss": 4.1689,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284095591939795e-05,
+      "loss": 4.1631,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283256997188742e-05,
+      "loss": 4.176,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28241840243769e-05,
+      "loss": 4.1603,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281579807686638e-05,
+      "loss": 4.1639,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280741212935586e-05,
+      "loss": 4.1664,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279904256064907e-05,
+      "loss": 4.1532,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279065661313855e-05,
+      "loss": 4.1755,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278227066562803e-05,
+      "loss": 4.1717,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277390109692124e-05,
+      "loss": 4.164,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276551514941072e-05,
+      "loss": 4.1642,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275712920190021e-05,
+      "loss": 4.1605,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274874325438969e-05,
+      "loss": 4.1717,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274035730687917e-05,
+      "loss": 4.1504,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273197135936865e-05,
+      "loss": 4.1778,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2723585411858127e-05,
+      "loss": 4.1593,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2715199464347607e-05,
+      "loss": 4.1613,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2706813516837087e-05,
+      "loss": 4.1596,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2698427569326566e-05,
+      "loss": 4.1565,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2690058000619776e-05,
+      "loss": 4.1571,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2681672053109256e-05,
+      "loss": 4.163,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2673286105598735e-05,
+      "loss": 4.1591,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2664900158088215e-05,
+      "loss": 4.1704,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2656514210577695e-05,
+      "loss": 4.1628,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2648144641870904e-05,
+      "loss": 4.1694,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2639758694360384e-05,
+      "loss": 4.1474,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263137274684987e-05,
+      "loss": 4.1717,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262298679933935e-05,
+      "loss": 4.1625,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261460085182883e-05,
+      "loss": 4.168,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260621490431831e-05,
+      "loss": 4.1606,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2597828956807784e-05,
+      "loss": 4.1549,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2589443009297264e-05,
+      "loss": 4.1679,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258107344059048e-05,
+      "loss": 4.1669,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257268749307996e-05,
+      "loss": 4.1495,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256430154556943e-05,
+      "loss": 4.1601,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255591559805891e-05,
+      "loss": 4.1553,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254752965054839e-05,
+      "loss": 4.1531,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253917646064534e-05,
+      "loss": 4.1642,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2530790513134825e-05,
+      "loss": 4.1557,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2522404565624305e-05,
+      "loss": 4.1596,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2514018618113785e-05,
+      "loss": 4.1464,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250563267060326e-05,
+      "loss": 4.1611,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.155533790588379,
+      "eval_runtime": 298.8064,
+      "eval_samples_per_second": 1277.051,
+      "eval_steps_per_second": 39.909,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249724672309274e-05,
+      "loss": 4.1582,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248886077558222e-05,
+      "loss": 4.1495,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24804748280717e-05,
+      "loss": 4.1656,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247208888056118e-05,
+      "loss": 4.1594,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246370293305066e-05,
+      "loss": 4.1587,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245531698554014e-05,
+      "loss": 4.1574,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244693103802962e-05,
+      "loss": 4.1484,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24385450905191e-05,
+      "loss": 4.1521,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243015914300858e-05,
+      "loss": 4.1595,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242177319549806e-05,
+      "loss": 4.1668,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241338724798754e-05,
+      "loss": 4.1574,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2405001300477024e-05,
+      "loss": 4.1679,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239663173177023e-05,
+      "loss": 4.1477,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238824578425971e-05,
+      "loss": 4.1523,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237985983674919e-05,
+      "loss": 4.1627,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237147388923867e-05,
+      "loss": 4.1415,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363087941728146e-05,
+      "loss": 4.1485,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2354701994217626e-05,
+      "loss": 4.1468,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346316046707106e-05,
+      "loss": 4.1508,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2337930099196586e-05,
+      "loss": 4.159,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2329560530489795e-05,
+      "loss": 4.1575,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321174582979275e-05,
+      "loss": 4.1554,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231278863546876e-05,
+      "loss": 4.1401,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230440268795824e-05,
+      "loss": 4.1576,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229601674044772e-05,
+      "loss": 4.1464,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22876307929372e-05,
+      "loss": 4.144,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227924484542668e-05,
+      "loss": 4.1404,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227085889791616e-05,
+      "loss": 4.1455,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226248932920937e-05,
+      "loss": 4.1468,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225410338169885e-05,
+      "loss": 4.142,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224573381299206e-05,
+      "loss": 4.1519,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223734786548154e-05,
+      "loss": 4.1526,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222896191797102e-05,
+      "loss": 4.1454,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22205759704605e-05,
+      "loss": 4.1524,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221219002294998e-05,
+      "loss": 4.15,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2203820454243196e-05,
+      "loss": 4.1473,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2195434506732676e-05,
+      "loss": 4.1428,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2187048559222156e-05,
+      "loss": 4.1347,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2178662611711635e-05,
+      "loss": 4.1337,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2170293043004845e-05,
+      "loss": 4.1426,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2161907095494325e-05,
+      "loss": 4.1419,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2153521147983804e-05,
+      "loss": 4.1452,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2145135200473284e-05,
+      "loss": 4.1396,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2136749252962764e-05,
+      "loss": 4.1413,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128379684255973e-05,
+      "loss": 4.1529,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212001011554918e-05,
+      "loss": 4.1413,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211162416803866e-05,
+      "loss": 4.1516,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210323822052815e-05,
+      "loss": 4.1357,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209485227301763e-05,
+      "loss": 4.126,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208646632550711e-05,
+      "loss": 4.1585,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207808037799659e-05,
+      "loss": 4.1427,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206969443048607e-05,
+      "loss": 4.135,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206130848297555e-05,
+      "loss": 4.1397,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205292253546503e-05,
+      "loss": 4.1399,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204453658795451e-05,
+      "loss": 4.1311,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203616701924772e-05,
+      "loss": 4.1412,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20277810717372e-05,
+      "loss": 4.1394,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201939512422668e-05,
+      "loss": 4.1385,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201100917671615e-05,
+      "loss": 4.149,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200262322920563e-05,
+      "loss": 4.1288,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199423728169512e-05,
+      "loss": 4.131,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19858513341846e-05,
+      "loss": 4.1475,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197746538667408e-05,
+      "loss": 4.1259,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196909581796729e-05,
+      "loss": 4.1255,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196070987045677e-05,
+      "loss": 4.134,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195232392294625e-05,
+      "loss": 4.1399,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194393797543573e-05,
+      "loss": 4.1253,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1935568406728936e-05,
+      "loss": 4.1294,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1927182459218416e-05,
+      "loss": 4.1246,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1918796511707896e-05,
+      "loss": 4.1466,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1910410564197376e-05,
+      "loss": 4.1384,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902040995490585e-05,
+      "loss": 4.1402,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189365504798007e-05,
+      "loss": 4.1412,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188526910046955e-05,
+      "loss": 4.1415,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187688315295903e-05,
+      "loss": 4.1519,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1868529963055976e-05,
+      "loss": 4.1226,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860144015545456e-05,
+      "loss": 4.1416,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185175806803493e-05,
+      "loss": 4.1434,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184337212052441e-05,
+      "loss": 4.122,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183498617301389e-05,
+      "loss": 4.1404,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182660022550337e-05,
+      "loss": 4.1332,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181821427799285e-05,
+      "loss": 4.1447,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180984470928606e-05,
+      "loss": 4.1462,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180145876177554e-05,
+      "loss": 4.1264,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1793072814265025e-05,
+      "loss": 4.125,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1784686866754505e-05,
+      "loss": 4.1287,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776300919243985e-05,
+      "loss": 4.1394,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1767914971733465e-05,
+      "loss": 4.1324,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17595617818304e-05,
+      "loss": 4.1343,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175117583431988e-05,
+      "loss": 4.1311,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174278988680936e-05,
+      "loss": 4.128,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173440393929884e-05,
+      "loss": 4.1338,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172601799178832e-05,
+      "loss": 4.1246,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17176320442778e-05,
+      "loss": 4.1406,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170924609676728e-05,
+      "loss": 4.1373,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170086014925676e-05,
+      "loss": 4.1368,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169247420174624e-05,
+      "loss": 4.1224,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168408825423572e-05,
+      "loss": 4.127,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167571868552894e-05,
+      "loss": 4.132,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166733273801842e-05,
+      "loss": 4.1322,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16589467905079e-05,
+      "loss": 4.134,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165056084299738e-05,
+      "loss": 4.122,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164217489548686e-05,
+      "loss": 4.1247,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163378894797634e-05,
+      "loss": 4.1262,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162540300046581e-05,
+      "loss": 4.1164,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161703343175903e-05,
+      "loss": 4.1236,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160864748424851e-05,
+      "loss": 4.1044,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160026153673799e-05,
+      "loss": 4.1295,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1591891968031197e-05,
+      "loss": 4.1246,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1583506020520677e-05,
+      "loss": 4.1307,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1575120073010156e-05,
+      "loss": 4.119,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156673412549964e-05,
+      "loss": 4.1255,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558348177989116e-05,
+      "loss": 4.1275,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1549962230478596e-05,
+      "loss": 4.1106,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1541576282968076e-05,
+      "loss": 4.137,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1533190335457556e-05,
+      "loss": 4.1288,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1524820766750765e-05,
+      "loss": 4.1257,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1516434819240245e-05,
+      "loss": 4.1276,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1508048871729725e-05,
+      "loss": 4.1137,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1499662924219205e-05,
+      "loss": 4.1341,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1491276976708685e-05,
+      "loss": 4.1152,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1482907408001894e-05,
+      "loss": 4.1341,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147453783929511e-05,
+      "loss": 4.1169,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146615189178459e-05,
+      "loss": 4.123,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145776594427407e-05,
+      "loss": 4.1169,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144937999676355e-05,
+      "loss": 4.1161,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144101042805676e-05,
+      "loss": 4.1185,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143262448054624e-05,
+      "loss": 4.119,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142423853303572e-05,
+      "loss": 4.1234,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14158525855252e-05,
+      "loss": 4.1279,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140746663801468e-05,
+      "loss": 4.1214,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139908069050416e-05,
+      "loss": 4.1303,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139069474299364e-05,
+      "loss": 4.1082,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138230879548312e-05,
+      "loss": 4.1308,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13739228479726e-05,
+      "loss": 4.1278,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136553690046208e-05,
+      "loss": 4.1245,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357167331755295e-05,
+      "loss": 4.12,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1348781384244775e-05,
+      "loss": 4.1202,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340395436734254e-05,
+      "loss": 4.1239,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332009489223734e-05,
+      "loss": 4.1289,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1323639920516944e-05,
+      "loss": 4.11,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315253973006424e-05,
+      "loss": 4.1216,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130688440429963e-05,
+      "loss": 4.1155,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129849845678911e-05,
+      "loss": 4.116,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129011250927859e-05,
+      "loss": 4.1218,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128172656176807e-05,
+      "loss": 4.1188,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127334061425755e-05,
+      "loss": 4.1175,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126495466674703e-05,
+      "loss": 4.1112,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125658509804025e-05,
+      "loss": 4.1268,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.12523889541626,
+      "eval_runtime": 299.4719,
+      "eval_samples_per_second": 1274.213,
+      "eval_steps_per_second": 39.82,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124819915052973e-05,
+      "loss": 4.1164,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123981320301921e-05,
+      "loss": 4.1101,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123142725550869e-05,
+      "loss": 4.1255,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122304130799817e-05,
+      "loss": 4.1202,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121465536048765e-05,
+      "loss": 4.1186,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120626941297712e-05,
+      "loss": 4.1212,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11978834654666e-05,
+      "loss": 4.1108,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118951389675982e-05,
+      "loss": 4.1117,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11811279492493e-05,
+      "loss": 4.1234,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1172758380542506e-05,
+      "loss": 4.1277,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1164372433031986e-05,
+      "loss": 4.1207,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115598648552147e-05,
+      "loss": 4.1283,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114761691681468e-05,
+      "loss": 4.1149,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113923096930416e-05,
+      "loss": 4.1131,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113084502179364e-05,
+      "loss": 4.1199,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112245907428312e-05,
+      "loss": 4.103,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1114073126772595e-05,
+      "loss": 4.1126,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1105687179262075e-05,
+      "loss": 4.1072,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109731761055529e-05,
+      "loss": 4.1133,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108893166304477e-05,
+      "loss": 4.1199,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1080545715534244e-05,
+      "loss": 4.1246,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1072159768023724e-05,
+      "loss": 4.1153,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106377382051321e-05,
+      "loss": 4.1025,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105538787300269e-05,
+      "loss": 4.1206,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104700192549217e-05,
+      "loss": 4.1052,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103861597798165e-05,
+      "loss": 4.109,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103023003047113e-05,
+      "loss": 4.1025,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102186046176434e-05,
+      "loss": 4.1104,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101347451425382e-05,
+      "loss": 4.1061,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10050885667433e-05,
+      "loss": 4.1085,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.099670261923278e-05,
+      "loss": 4.1104,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098831667172226e-05,
+      "loss": 4.1184,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097993072421174e-05,
+      "loss": 4.1107,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097154477670122e-05,
+      "loss": 4.1091,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09631588291907e-05,
+      "loss": 4.115,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095478926048391e-05,
+      "loss": 4.1114,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0946419691777124e-05,
+      "loss": 4.1092,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0938033744266604e-05,
+      "loss": 4.0938,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0929647796756084e-05,
+      "loss": 4.0992,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0921261849245564e-05,
+      "loss": 4.1035,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091289228053877e-05,
+      "loss": 4.1092,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090450633302825e-05,
+      "loss": 4.1087,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089612038551773e-05,
+      "loss": 4.1017,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088773443800721e-05,
+      "loss": 4.1035,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087934849049669e-05,
+      "loss": 4.1157,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08709789217899e-05,
+      "loss": 4.101,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086259297427938e-05,
+      "loss": 4.1231,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085420702676886e-05,
+      "loss": 4.0979,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084582107925834e-05,
+      "loss": 4.0857,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083743513174783e-05,
+      "loss": 4.1239,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082906556304104e-05,
+      "loss": 4.1033,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082067961553052e-05,
+      "loss": 4.1033,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081229366802e-05,
+      "loss": 4.1024,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080390772050948e-05,
+      "loss": 4.1025,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079552177299896e-05,
+      "loss": 4.0959,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078713582548843e-05,
+      "loss": 4.1013,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077874987797791e-05,
+      "loss": 4.1035,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0770380309271127e-05,
+      "loss": 4.1042,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0761994361760607e-05,
+      "loss": 4.1151,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0753624793053816e-05,
+      "loss": 4.0926,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0745238845543296e-05,
+      "loss": 4.093,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073685289803278e-05,
+      "loss": 4.1121,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0728466950522255e-05,
+      "loss": 4.087,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0720081003011735e-05,
+      "loss": 4.0879,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0711695055501215e-05,
+      "loss": 4.0978,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0703309107990695e-05,
+      "loss": 4.105,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0694923160480175e-05,
+      "loss": 4.0889,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0686553591773384e-05,
+      "loss": 4.098,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06781840230666e-05,
+      "loss": 4.0893,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066979807555608e-05,
+      "loss": 4.1081,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0661412128045553e-05,
+      "loss": 4.1063,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065302618053503e-05,
+      "loss": 4.1013,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064464023302452e-05,
+      "loss": 4.1055,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636254285514e-05,
+      "loss": 4.1063,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062788471680721e-05,
+      "loss": 4.1208,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061949876929669e-05,
+      "loss": 4.0861,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061111282178617e-05,
+      "loss": 4.1096,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060274325307938e-05,
+      "loss": 4.1037,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059435730556886e-05,
+      "loss": 4.0904,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058597135805834e-05,
+      "loss": 4.1027,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057758541054782e-05,
+      "loss": 4.1003,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05691994630373e-05,
+      "loss": 4.1077,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056081351552678e-05,
+      "loss": 4.1135,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055242756801626e-05,
+      "loss": 4.091,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054404162050574e-05,
+      "loss": 4.0873,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0535672051798954e-05,
+      "loss": 4.0964,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0527286104288434e-05,
+      "loss": 4.104,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0518900156777914e-05,
+      "loss": 4.1002,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051053058807112e-05,
+      "loss": 4.0963,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.05021446405606e-05,
+      "loss": 4.0908,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049375869305008e-05,
+      "loss": 4.0948,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048537274553956e-05,
+      "loss": 4.0981,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047698679802904e-05,
+      "loss": 4.0936,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046860085051852e-05,
+      "loss": 4.1044,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0460214903008e-05,
+      "loss": 4.1052,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045182895549748e-05,
+      "loss": 4.0989,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044344300798696e-05,
+      "loss": 4.0906,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043505706047644e-05,
+      "loss": 4.0968,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042668749176965e-05,
+      "loss": 4.0908,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041830154425914e-05,
+      "loss": 4.0977,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040991559674862e-05,
+      "loss": 4.1008,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040152964923809e-05,
+      "loss": 4.0925,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039316008053131e-05,
+      "loss": 4.0868,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038477413302079e-05,
+      "loss": 4.0876,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037638818551027e-05,
+      "loss": 4.0889,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036800223799974e-05,
+      "loss": 4.0878,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035961629048922e-05,
+      "loss": 4.0738,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03512303429787e-05,
+      "loss": 4.094,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034284439546818e-05,
+      "loss": 4.0948,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033445844795766e-05,
+      "loss": 4.0927,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0326088879250876e-05,
+      "loss": 4.0844,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031771931054409e-05,
+      "loss": 4.0999,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0309333363033565e-05,
+      "loss": 4.0875,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0300947415523045e-05,
+      "loss": 4.0796,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0292561468012525e-05,
+      "loss": 4.1015,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284175520502005e-05,
+      "loss": 4.0951,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0275805951795214e-05,
+      "loss": 4.092,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0267420004284694e-05,
+      "loss": 4.0971,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0259034056774174e-05,
+      "loss": 4.0804,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0250648109263654e-05,
+      "loss": 4.1007,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0242262161753134e-05,
+      "loss": 4.0853,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0233876214242614e-05,
+      "loss": 4.0981,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0225490266732094e-05,
+      "loss": 4.0854,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0217104319221574e-05,
+      "loss": 4.0921,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020873475051479e-05,
+      "loss": 4.0857,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020034880300427e-05,
+      "loss": 4.084,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019196285549375e-05,
+      "loss": 4.0838,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018359328678696e-05,
+      "loss": 4.088,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017520733927644e-05,
+      "loss": 4.0909,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016682139176592e-05,
+      "loss": 4.0994,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01584354442554e-05,
+      "loss": 4.084,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015004949674488e-05,
+      "loss": 4.1018,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014166354923436e-05,
+      "loss": 4.0734,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013327760172384e-05,
+      "loss": 4.0988,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012490803301705e-05,
+      "loss": 4.0963,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011652208550653e-05,
+      "loss": 4.0927,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0108136137996014e-05,
+      "loss": 4.0873,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0099750190485494e-05,
+      "loss": 4.0894,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0091364242974974e-05,
+      "loss": 4.0932,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0082978295464454e-05,
+      "loss": 4.1001,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007460872675766e-05,
+      "loss": 4.0776,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006622277924714e-05,
+      "loss": 4.0898,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005783683173662e-05,
+      "loss": 4.0814,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00494508842261e-05,
+      "loss": 4.0884,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004108131551931e-05,
+      "loss": 4.0844,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003269536800879e-05,
+      "loss": 4.0862,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002430942049827e-05,
+      "loss": 4.0905,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001592347298775e-05,
+      "loss": 4.0761,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000755390428097e-05,
+      "loss": 4.095,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.100830554962158,
+      "eval_runtime": 298.1924,
+      "eval_samples_per_second": 1279.68,
+      "eval_steps_per_second": 39.991,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999916795677045e-05,
+      "loss": 4.0862,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999078200925993e-05,
+      "loss": 4.0831,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99823960617494e-05,
+      "loss": 4.0924,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997401011423888e-05,
+      "loss": 4.0884,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996562416672836e-05,
+      "loss": 4.0909,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995723821921784e-05,
+      "loss": 4.0857,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994885227170732e-05,
+      "loss": 4.0842,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99404663241968e-05,
+      "loss": 4.0828,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993208037668628e-05,
+      "loss": 4.0885,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992369442917576e-05,
+      "loss": 4.0971,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991530848166524e-05,
+      "loss": 4.0912,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990692253415472e-05,
+      "loss": 4.0946,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989855296544793e-05,
+      "loss": 4.0799,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890167017937416e-05,
+      "loss": 4.0832,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9881781070426896e-05,
+      "loss": 4.092,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9873395122916376e-05,
+      "loss": 4.0723,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865009175405856e-05,
+      "loss": 4.0804,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9856623227895336e-05,
+      "loss": 4.0764,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848237280384816e-05,
+      "loss": 4.0836,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983985133287429e-05,
+      "loss": 4.087,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9831481764167505e-05,
+      "loss": 4.0945,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823095816656985e-05,
+      "loss": 4.0857,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981470986914646e-05,
+      "loss": 4.0752,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980632392163594e-05,
+      "loss": 4.0866,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9797954352929154e-05,
+      "loss": 4.0787,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9789568405418634e-05,
+      "loss": 4.0799,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781182457908114e-05,
+      "loss": 4.0709,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9772796510397594e-05,
+      "loss": 4.082,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9764410562887074e-05,
+      "loss": 4.0817,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9756024615376554e-05,
+      "loss": 4.073,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9747638667866034e-05,
+      "loss": 4.079,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9739252720355514e-05,
+      "loss": 4.0887,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973088315164872e-05,
+      "loss": 4.079,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972251358294193e-05,
+      "loss": 4.0803,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971412763543141e-05,
+      "loss": 4.0835,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970574168792089e-05,
+      "loss": 4.0861,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969735574041037e-05,
+      "loss": 4.0766,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968896979289985e-05,
+      "loss": 4.0655,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968058384538934e-05,
+      "loss": 4.0723,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967219789787882e-05,
+      "loss": 4.0729,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96638119503683e-05,
+      "loss": 4.0763,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965542600285778e-05,
+      "loss": 4.0787,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964705643415099e-05,
+      "loss": 4.0747,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963867048664047e-05,
+      "loss": 4.0761,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963028453912995e-05,
+      "loss": 4.084,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962189859161943e-05,
+      "loss": 4.0726,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9613529022912636e-05,
+      "loss": 4.0899,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9605143075402116e-05,
+      "loss": 4.0717,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959678988549906e-05,
+      "loss": 4.0519,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958840393798854e-05,
+      "loss": 4.096,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958001799047802e-05,
+      "loss": 4.0723,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95716320429675e-05,
+      "loss": 4.078,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956324609545698e-05,
+      "loss": 4.071,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955486014794646e-05,
+      "loss": 4.0711,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954647420043594e-05,
+      "loss": 4.0698,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953808825292542e-05,
+      "loss": 4.0702,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95297023054149e-05,
+      "loss": 4.0759,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952133273670811e-05,
+      "loss": 4.075,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951294678919759e-05,
+      "loss": 4.0853,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950456084168707e-05,
+      "loss": 4.0633,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949617489417655e-05,
+      "loss": 4.0628,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948778894666603e-05,
+      "loss": 4.0857,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947940299915551e-05,
+      "loss": 4.0574,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947101705164499e-05,
+      "loss": 4.0587,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9462647482938206e-05,
+      "loss": 4.0686,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454277914231415e-05,
+      "loss": 4.0739,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9445891966720895e-05,
+      "loss": 4.0618,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9437506019210375e-05,
+      "loss": 4.069,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429120071699855e-05,
+      "loss": 4.0602,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9420734124189335e-05,
+      "loss": 4.0804,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9412348176678815e-05,
+      "loss": 4.0753,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9403962229168295e-05,
+      "loss": 4.0743,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939557628165777e-05,
+      "loss": 4.0777,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9387206712950984e-05,
+      "loss": 4.0786,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93788371442442e-05,
+      "loss": 4.0885,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937045119673368e-05,
+      "loss": 4.0595,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936206524922316e-05,
+      "loss": 4.0809,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935367930171264e-05,
+      "loss": 4.0697,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934529335420212e-05,
+      "loss": 4.065,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933690740669159e-05,
+      "loss": 4.0743,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932852145918107e-05,
+      "loss": 4.0742,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932013551167055e-05,
+      "loss": 4.0779,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93117823217675e-05,
+      "loss": 4.0871,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930339637425698e-05,
+      "loss": 4.062,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929501042674646e-05,
+      "loss": 4.0572,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.928662447923594e-05,
+      "loss": 4.0687,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927823853172542e-05,
+      "loss": 4.0766,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92698525842149e-05,
+      "loss": 4.0725,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926148301550811e-05,
+      "loss": 4.0726,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925311344680132e-05,
+      "loss": 4.0618,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.92447274992908e-05,
+      "loss": 4.0656,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923634155178028e-05,
+      "loss": 4.0728,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922795560426976e-05,
+      "loss": 4.0646,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921956965675924e-05,
+      "loss": 4.0751,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211183709248715e-05,
+      "loss": 4.0775,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9202797761738195e-05,
+      "loss": 4.0701,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9194411814227675e-05,
+      "loss": 4.0638,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9186025866717155e-05,
+      "loss": 4.0669,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917765629801037e-05,
+      "loss": 4.0644,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916927035049985e-05,
+      "loss": 4.0681,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916088440298933e-05,
+      "loss": 4.0746,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915249845547881e-05,
+      "loss": 4.0667,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914411250796829e-05,
+      "loss": 4.0638,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913572656045777e-05,
+      "loss": 4.0516,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912735699175098e-05,
+      "loss": 4.065,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911897104424046e-05,
+      "loss": 4.0576,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911058509672994e-05,
+      "loss": 4.0443,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910219914921942e-05,
+      "loss": 4.069,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90938132017089e-05,
+      "loss": 4.0658,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908542725419838e-05,
+      "loss": 4.0656,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907705768549159e-05,
+      "loss": 4.0529,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906867173798107e-05,
+      "loss": 4.0688,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9060302169274284e-05,
+      "loss": 4.062,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9051916221763764e-05,
+      "loss": 4.0523,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9043530274253244e-05,
+      "loss": 4.0764,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9035144326742724e-05,
+      "loss": 4.0696,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9026758379232204e-05,
+      "loss": 4.0632,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9018372431721684e-05,
+      "loss": 4.0676,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9009986484211164e-05,
+      "loss": 4.0517,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9001600536700644e-05,
+      "loss": 4.0714,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8993214589190124e-05,
+      "loss": 4.0622,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898484502048333e-05,
+      "loss": 4.066,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897645907297281e-05,
+      "loss": 4.0578,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896807312546229e-05,
+      "loss": 4.0675,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895970355675551e-05,
+      "loss": 4.059,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895131760924499e-05,
+      "loss": 4.0554,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894293166173447e-05,
+      "loss": 4.0543,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893454571422395e-05,
+      "loss": 4.0651,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892615976671343e-05,
+      "loss": 4.0602,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89177738192029e-05,
+      "loss": 4.071,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890938787169238e-05,
+      "loss": 4.0593,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890100192418186e-05,
+      "loss": 4.0757,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889261597667134e-05,
+      "loss": 4.049,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888423002916082e-05,
+      "loss": 4.0708,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88758440816503e-05,
+      "loss": 4.0686,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886745813413978e-05,
+      "loss": 4.0702,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885908856543299e-05,
+      "loss": 4.0603,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885070261792248e-05,
+      "loss": 4.0604,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884233304921569e-05,
+      "loss": 4.0614,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8833947101705167e-05,
+      "loss": 4.0796,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8825561154194647e-05,
+      "loss": 4.0492,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8817175206684127e-05,
+      "loss": 4.0614,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8808789259173606e-05,
+      "loss": 4.0549,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8800403311663086e-05,
+      "loss": 4.0651,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8792017364152566e-05,
+      "loss": 4.0535,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8783631416642046e-05,
+      "loss": 4.0619,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8775261847935255e-05,
+      "loss": 4.0602,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8766892279228465e-05,
+      "loss": 4.0568,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8758506331717944e-05,
+      "loss": 4.065,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.081409454345703,
+      "eval_runtime": 289.9943,
+      "eval_samples_per_second": 1315.857,
+      "eval_steps_per_second": 41.121,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.875012038420743e-05,
+      "loss": 4.0537,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874173443669691e-05,
+      "loss": 4.0548,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.873336486799012e-05,
+      "loss": 4.0676,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.87249789204796e-05,
+      "loss": 4.0624,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871659297296908e-05,
+      "loss": 4.0647,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870820702545856e-05,
+      "loss": 4.0604,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869982107794804e-05,
+      "loss": 4.0559,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869143513043752e-05,
+      "loss": 4.0581,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.868306556173073e-05,
+      "loss": 4.0581,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.867467961422021e-05,
+      "loss": 4.0719,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.866629366670969e-05,
+      "loss": 4.0656,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865790771919917e-05,
+      "loss": 4.0637,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864953815049238e-05,
+      "loss": 4.0653,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8641152202981865e-05,
+      "loss": 4.0482,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8632766255471345e-05,
+      "loss": 4.0626,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8624380307960825e-05,
+      "loss": 4.049,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8615994360450305e-05,
+      "loss": 4.0538,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8607608412939785e-05,
+      "loss": 4.0533,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8599222465429265e-05,
+      "loss": 4.0537,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859083651791874e-05,
+      "loss": 4.06,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8582466949211954e-05,
+      "loss": 4.0713,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8574081001701434e-05,
+      "loss": 4.063,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8565695054190914e-05,
+      "loss": 4.0482,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855730910668039e-05,
+      "loss": 4.0566,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854892315916987e-05,
+      "loss": 4.0526,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854053721165935e-05,
+      "loss": 4.0544,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853216764295256e-05,
+      "loss": 4.0468,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852378169544204e-05,
+      "loss": 4.0587,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851539574793152e-05,
+      "loss": 4.0518,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8507009800421e-05,
+      "loss": 4.0506,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849862385291048e-05,
+      "loss": 4.0502,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849025428420369e-05,
+      "loss": 4.0651,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848186833669317e-05,
+      "loss": 4.0545,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847348238918265e-05,
+      "loss": 4.0556,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846509644167213e-05,
+      "loss": 4.0573,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.845671049416161e-05,
+      "loss": 4.0562,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.844832454665109e-05,
+      "loss": 4.0552,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843993859914057e-05,
+      "loss": 4.0418,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843155265163005e-05,
+      "loss": 4.0451,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.842318308292327e-05,
+      "loss": 4.0485,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.841479713541275e-05,
+      "loss": 4.0474,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8406427566705956e-05,
+      "loss": 4.0523,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8398041619195436e-05,
+      "loss": 4.0498,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8389655671684916e-05,
+      "loss": 4.0561,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8381269724174396e-05,
+      "loss": 4.053,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8372883776663876e-05,
+      "loss": 4.0498,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8364497829153356e-05,
+      "loss": 4.0619,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8356111881642836e-05,
+      "loss": 4.0491,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8347725934132316e-05,
+      "loss": 4.0286,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8339356365425525e-05,
+      "loss": 4.0639,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8330970417915005e-05,
+      "loss": 4.0479,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832260084920822e-05,
+      "loss": 4.0543,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83142149016977e-05,
+      "loss": 4.0489,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830582895418718e-05,
+      "loss": 4.0435,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.829744300667666e-05,
+      "loss": 4.0432,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828905705916614e-05,
+      "loss": 4.0456,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828067111165562e-05,
+      "loss": 4.0555,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.827230154294883e-05,
+      "loss": 4.0469,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826391559543831e-05,
+      "loss": 4.0595,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.825552964792779e-05,
+      "loss": 4.0451,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824714370041727e-05,
+      "loss": 4.0335,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823877413171048e-05,
+      "loss": 4.0613,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823038818419996e-05,
+      "loss": 4.0345,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822200223668944e-05,
+      "loss": 4.0337,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8213616289178925e-05,
+      "loss": 4.047,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.82052303416684e-05,
+      "loss": 4.047,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819684439415788e-05,
+      "loss": 4.0395,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818845844664736e-05,
+      "loss": 4.0455,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818007249913684e-05,
+      "loss": 4.039,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817170293043005e-05,
+      "loss": 4.0527,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816331698291953e-05,
+      "loss": 4.054,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815494741421274e-05,
+      "loss": 4.0515,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814656146670222e-05,
+      "loss": 4.0523,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8138175519191696e-05,
+      "loss": 4.0534,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8129789571681176e-05,
+      "loss": 4.0637,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812140362417066e-05,
+      "loss": 4.0364,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811303405546387e-05,
+      "loss": 4.0568,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810464810795335e-05,
+      "loss": 4.0421,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.809626216044283e-05,
+      "loss": 4.0436,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808787621293231e-05,
+      "loss": 4.0463,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807950664422552e-05,
+      "loss": 4.0505,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8071120696715e-05,
+      "loss": 4.0552,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806273474920448e-05,
+      "loss": 4.0638,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805434880169396e-05,
+      "loss": 4.0362,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804597923298717e-05,
+      "loss": 4.0379,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803759328547665e-05,
+      "loss": 4.0449,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802920733796613e-05,
+      "loss": 4.0516,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802082139045561e-05,
+      "loss": 4.0453,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8012435442945097e-05,
+      "loss": 4.0484,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8004049495434577e-05,
+      "loss": 4.0412,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7995663547924056e-05,
+      "loss": 4.044,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7987277600413536e-05,
+      "loss": 4.0441,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7978924410510475e-05,
+      "loss": 4.0401,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797055484180369e-05,
+      "loss": 4.052,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7962168894293164e-05,
+      "loss": 4.0537,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7953782946782644e-05,
+      "loss": 4.0503,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7945396999272124e-05,
+      "loss": 4.0405,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7937011051761604e-05,
+      "loss": 4.0407,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7928625104251084e-05,
+      "loss": 4.0423,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7920239156740564e-05,
+      "loss": 4.0446,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791185320923005e-05,
+      "loss": 4.0489,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.790346726171953e-05,
+      "loss": 4.0439,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789508131420901e-05,
+      "loss": 4.0414,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788669536669849e-05,
+      "loss": 4.029,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787830941918797e-05,
+      "loss": 4.045,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786992347167745e-05,
+      "loss": 4.0351,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786155390297066e-05,
+      "loss": 4.0225,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785316795546014e-05,
+      "loss": 4.0411,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.784478200794962e-05,
+      "loss": 4.051,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78363960604391e-05,
+      "loss": 4.0378,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782801011292858e-05,
+      "loss": 4.0333,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781962416541805e-05,
+      "loss": 4.0427,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781123821790753e-05,
+      "loss": 4.038,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780286864920075e-05,
+      "loss": 4.0313,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7794482701690235e-05,
+      "loss": 4.0561,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778609675417971e-05,
+      "loss": 4.048,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777771080666919e-05,
+      "loss": 4.0359,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776932485915867e-05,
+      "loss": 4.0454,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776093891164815e-05,
+      "loss": 4.0329,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.775255296413763e-05,
+      "loss": 4.046,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.774416701662711e-05,
+      "loss": 4.0412,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.773579744792032e-05,
+      "loss": 4.0412,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77274115004098e-05,
+      "loss": 4.0363,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771902555289928e-05,
+      "loss": 4.0466,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7710639605388757e-05,
+      "loss": 4.0379,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7702253657878237e-05,
+      "loss": 4.0341,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769388408917145e-05,
+      "loss": 4.0279,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768549814166093e-05,
+      "loss": 4.0423,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767711219415041e-05,
+      "loss": 4.0358,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766872624663989e-05,
+      "loss": 4.0482,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766034029912937e-05,
+      "loss": 4.0376,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765195435161885e-05,
+      "loss": 4.0489,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764358478291206e-05,
+      "loss": 4.0247,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763519883540154e-05,
+      "loss": 4.0527,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.762681288789102e-05,
+      "loss": 4.0457,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76184269403805e-05,
+      "loss": 4.0439,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761004099286998e-05,
+      "loss": 4.0414,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760165504535946e-05,
+      "loss": 4.0399,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759326909784894e-05,
+      "loss": 4.0392,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758488315033842e-05,
+      "loss": 4.0599,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.757651358163164e-05,
+      "loss": 4.0255,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756812763412112e-05,
+      "loss": 4.041,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75597416866106e-05,
+      "loss": 4.0354,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755135573910007e-05,
+      "loss": 4.0397,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7543002549197015e-05,
+      "loss": 4.03,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7534616601686495e-05,
+      "loss": 4.0411,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7526230654175975e-05,
+      "loss": 4.0362,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7517844706665455e-05,
+      "loss": 4.0378,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7509458759154935e-05,
+      "loss": 4.0414,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.066771507263184,
+      "eval_runtime": 290.5045,
+      "eval_samples_per_second": 1313.546,
+      "eval_steps_per_second": 41.049,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7501072811644415e-05,
+      "loss": 4.0324,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749268686413389e-05,
+      "loss": 4.0354,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7484300916623375e-05,
+      "loss": 4.0422,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7475914969112855e-05,
+      "loss": 4.043,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7467529021602335e-05,
+      "loss": 4.0443,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7459143074091815e-05,
+      "loss": 4.0399,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7450757126581294e-05,
+      "loss": 4.0345,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7442371179070774e-05,
+      "loss": 4.0352,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7433985231560254e-05,
+      "loss": 4.0383,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7425599284049734e-05,
+      "loss": 4.0474,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7417213336539214e-05,
+      "loss": 4.0428,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7408827389028694e-05,
+      "loss": 4.0438,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.74004578203219e-05,
+      "loss": 4.0448,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.739207187281138e-05,
+      "loss": 4.0228,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738368592530086e-05,
+      "loss": 4.0462,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737529997779034e-05,
+      "loss": 4.0233,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736691403027982e-05,
+      "loss": 4.0319,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73585280827693e-05,
+      "loss": 4.0329,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735014213525878e-05,
+      "loss": 4.03,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734175618774826e-05,
+      "loss": 4.0393,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733337024023774e-05,
+      "loss": 4.0467,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732500067153096e-05,
+      "loss": 4.0431,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731661472402043e-05,
+      "loss": 4.0308,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730822877650991e-05,
+      "loss": 4.0297,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729984282899939e-05,
+      "loss": 4.0363,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729145688148887e-05,
+      "loss": 4.0304,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728307093397835e-05,
+      "loss": 4.0264,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727468498646783e-05,
+      "loss": 4.0395,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726629903895731e-05,
+      "loss": 4.0269,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725792947025053e-05,
+      "loss": 4.0311,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724954352274001e-05,
+      "loss": 4.0254,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724115757522949e-05,
+      "loss": 4.0417,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723277162771897e-05,
+      "loss": 4.0363,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722440205901218e-05,
+      "loss": 4.031,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7216016111501657e-05,
+      "loss": 4.0377,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7207630163991137e-05,
+      "loss": 4.0305,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7199244216480616e-05,
+      "loss": 4.036,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7190874647773826e-05,
+      "loss": 4.0243,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7182488700263306e-05,
+      "loss": 4.0188,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7174102752752785e-05,
+      "loss": 4.024,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7165716805242265e-05,
+      "loss": 4.0301,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.715734723653548e-05,
+      "loss": 4.0285,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714896128902496e-05,
+      "loss": 4.0247,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714057534151444e-05,
+      "loss": 4.0359,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.713218939400392e-05,
+      "loss": 4.032,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71238034464934e-05,
+      "loss": 4.0296,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711543387778661e-05,
+      "loss": 4.0446,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.710704793027609e-05,
+      "loss": 4.0269,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709866198276557e-05,
+      "loss": 4.0053,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709027603525505e-05,
+      "loss": 4.0455,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.708189008774453e-05,
+      "loss": 4.0273,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.707352051903774e-05,
+      "loss": 4.0334,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.706515095033095e-05,
+      "loss": 4.0262,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7056765002820435e-05,
+      "loss": 4.0256,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7048379055309915e-05,
+      "loss": 4.0222,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7039993107799395e-05,
+      "loss": 4.0213,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7031623539092604e-05,
+      "loss": 4.0391,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7023237591582084e-05,
+      "loss": 4.0217,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7014851644071564e-05,
+      "loss": 4.0404,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7006465696561044e-05,
+      "loss": 4.024,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6998079749050524e-05,
+      "loss": 4.0104,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6989693801540004e-05,
+      "loss": 4.0371,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6981307854029484e-05,
+      "loss": 4.0204,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6972921906518964e-05,
+      "loss": 4.0096,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696453595900844e-05,
+      "loss": 4.0246,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695615001149792e-05,
+      "loss": 4.0245,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69477640639874e-05,
+      "loss": 4.0174,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6939378116476884e-05,
+      "loss": 4.0227,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6930992168966363e-05,
+      "loss": 4.0188,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6922606221455843e-05,
+      "loss": 4.0297,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691422027394532e-05,
+      "loss": 4.0355,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.690585070523853e-05,
+      "loss": 4.0263,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689746475772801e-05,
+      "loss": 4.0339,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688907881021749e-05,
+      "loss": 4.03,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688069286270697e-05,
+      "loss": 4.0432,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.687232329400018e-05,
+      "loss": 4.0171,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686393734648966e-05,
+      "loss": 4.0385,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685555139897914e-05,
+      "loss": 4.0228,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684716545146862e-05,
+      "loss": 4.0206,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68387795039581e-05,
+      "loss": 4.0308,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683039355644758e-05,
+      "loss": 4.0246,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682200760893707e-05,
+      "loss": 4.0363,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681362166142655e-05,
+      "loss": 4.0421,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680525209271976e-05,
+      "loss": 4.0182,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679686614520924e-05,
+      "loss": 4.0163,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6788496576502446e-05,
+      "loss": 4.0208,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6780110628991926e-05,
+      "loss": 4.0362,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6771724681481406e-05,
+      "loss": 4.0205,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6763355112774615e-05,
+      "loss": 4.0265,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6754969165264095e-05,
+      "loss": 4.0253,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6746583217753575e-05,
+      "loss": 4.0182,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6738197270243055e-05,
+      "loss": 4.0282,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6729811322732535e-05,
+      "loss": 4.0178,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672142537522202e-05,
+      "loss": 4.0315,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.67130394277115e-05,
+      "loss": 4.03,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.670465348020098e-05,
+      "loss": 4.033,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6696267532690455e-05,
+      "loss": 4.0178,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6687881585179935e-05,
+      "loss": 4.0241,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667951201647315e-05,
+      "loss": 4.0192,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6671126068962624e-05,
+      "loss": 4.0244,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6662740121452104e-05,
+      "loss": 4.0319,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6654354173941584e-05,
+      "loss": 4.0191,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66459846052348e-05,
+      "loss": 4.0238,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663759865772427e-05,
+      "loss": 4.0107,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662922908901749e-05,
+      "loss": 4.0224,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6620843141506975e-05,
+      "loss": 4.0164,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6612457193996455e-05,
+      "loss": 4.0012,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.660407124648593e-05,
+      "loss": 4.0222,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.659568529897541e-05,
+      "loss": 4.0266,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.658729935146489e-05,
+      "loss": 4.0186,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657891340395437e-05,
+      "loss": 4.0141,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657052745644385e-05,
+      "loss": 4.0205,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656215788773706e-05,
+      "loss": 4.0212,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655377194022654e-05,
+      "loss": 4.0088,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654538599271602e-05,
+      "loss": 4.0387,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65370000452055e-05,
+      "loss": 4.0233,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652861409769498e-05,
+      "loss": 4.0184,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652024452898819e-05,
+      "loss": 4.0263,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651185858147767e-05,
+      "loss": 4.0118,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650347263396715e-05,
+      "loss": 4.0253,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649508668645663e-05,
+      "loss": 4.0227,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648670073894611e-05,
+      "loss": 4.0178,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647831479143559e-05,
+      "loss": 4.0207,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646992884392507e-05,
+      "loss": 4.0219,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646154289641455e-05,
+      "loss": 4.0199,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.645317332770776e-05,
+      "loss": 4.0117,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644478738019724e-05,
+      "loss": 4.0147,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643640143268672e-05,
+      "loss": 4.0167,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64280154851762e-05,
+      "loss": 4.0158,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641964591646941e-05,
+      "loss": 4.0303,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641125996895889e-05,
+      "loss": 4.0196,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.640287402144838e-05,
+      "loss": 4.0296,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.639448807393786e-05,
+      "loss": 4.0023,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6386118505231067e-05,
+      "loss": 4.0358,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6377732557720546e-05,
+      "loss": 4.0216,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6369346610210026e-05,
+      "loss": 4.0247,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6360960662699506e-05,
+      "loss": 4.0222,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6352591093992715e-05,
+      "loss": 4.0145,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6344221525285925e-05,
+      "loss": 4.0201,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6335835577775405e-05,
+      "loss": 4.0352,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6327449630264884e-05,
+      "loss": 4.0075,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6319063682754364e-05,
+      "loss": 4.0206,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631069411404758e-05,
+      "loss": 4.0141,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630230816653706e-05,
+      "loss": 4.0194,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629392221902654e-05,
+      "loss": 4.0133,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628553627151602e-05,
+      "loss": 4.0183,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.62771503240055e-05,
+      "loss": 4.0172,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626878075529871e-05,
+      "loss": 4.0219,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626039480778819e-05,
+      "loss": 4.0161,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.054329872131348,
+      "eval_runtime": 297.6135,
+      "eval_samples_per_second": 1282.17,
+      "eval_steps_per_second": 40.069,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.625200886027767e-05,
+      "loss": 4.0122,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.624362291276715e-05,
+      "loss": 4.0152,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.623525334406036e-05,
+      "loss": 4.0234,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622686739654984e-05,
+      "loss": 4.022,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621848144903932e-05,
+      "loss": 4.0249,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.62100955015288e-05,
+      "loss": 4.0199,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6201725932822014e-05,
+      "loss": 4.0144,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6193339985311494e-05,
+      "loss": 4.0208,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6184954037800974e-05,
+      "loss": 4.0141,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6176568090290454e-05,
+      "loss": 4.0265,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6168182142779934e-05,
+      "loss": 4.0271,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615979619526941e-05,
+      "loss": 4.026,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615142662656262e-05,
+      "loss": 4.0239,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61430406790521e-05,
+      "loss": 4.0055,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.613465473154158e-05,
+      "loss": 4.0259,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6126268784031056e-05,
+      "loss": 4.0025,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6117882836520536e-05,
+      "loss": 4.0166,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610949688901002e-05,
+      "loss": 4.0133,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610112732030323e-05,
+      "loss": 4.0095,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609274137279271e-05,
+      "loss": 4.018,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608435542528219e-05,
+      "loss": 4.0295,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607596947777167e-05,
+      "loss": 4.0208,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606758353026115e-05,
+      "loss": 4.0158,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605919758275063e-05,
+      "loss": 4.0146,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605081163524011e-05,
+      "loss": 4.0149,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.604242568772959e-05,
+      "loss": 4.0111,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60340561190228e-05,
+      "loss": 4.0039,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602567017151228e-05,
+      "loss": 4.0201,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601728422400176e-05,
+      "loss": 4.0106,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600889827649124e-05,
+      "loss": 4.0143,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6000528707784456e-05,
+      "loss": 4.0079,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5992142760273936e-05,
+      "loss": 4.0202,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5983756812763416e-05,
+      "loss": 4.0125,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5975370865252896e-05,
+      "loss": 4.0169,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5966984917742376e-05,
+      "loss": 4.0189,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5958598970231856e-05,
+      "loss": 4.0117,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5950213022721336e-05,
+      "loss": 4.0197,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5941827075210816e-05,
+      "loss": 4.0003,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5933457506504025e-05,
+      "loss": 4.003,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5925071558993505e-05,
+      "loss": 4.0073,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5916701990286714e-05,
+      "loss": 4.0114,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590833242157993e-05,
+      "loss": 4.0102,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589994647406941e-05,
+      "loss": 4.0075,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589156052655889e-05,
+      "loss": 4.0132,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588317457904837e-05,
+      "loss": 4.0129,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587478863153785e-05,
+      "loss": 4.0108,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586640268402733e-05,
+      "loss": 4.0212,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585801673651681e-05,
+      "loss": 4.0108,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584963078900629e-05,
+      "loss": 3.9883,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58412612202995e-05,
+      "loss": 4.0269,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583287527278898e-05,
+      "loss": 4.0085,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582448932527846e-05,
+      "loss": 4.0108,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581610337776794e-05,
+      "loss": 4.0063,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580771743025742e-05,
+      "loss": 4.0048,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579933148274689e-05,
+      "loss": 4.0056,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579094553523638e-05,
+      "loss": 4.0021,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.578255958772586e-05,
+      "loss": 4.0239,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577419001901907e-05,
+      "loss": 4.0026,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.576580407150855e-05,
+      "loss": 4.0231,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.575741812399803e-05,
+      "loss": 4.0052,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574903217648751e-05,
+      "loss": 3.9898,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5740662607780716e-05,
+      "loss": 4.0191,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5732276660270196e-05,
+      "loss": 4.0022,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5723890712759676e-05,
+      "loss": 3.9934,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5715504765249156e-05,
+      "loss": 4.0058,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57071515753461e-05,
+      "loss": 4.0051,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569876562783558e-05,
+      "loss": 4.0016,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569037968032506e-05,
+      "loss": 4.0081,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568199373281454e-05,
+      "loss": 4.0013,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567360778530402e-05,
+      "loss": 4.0091,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566523821659724e-05,
+      "loss": 4.0143,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565685226908672e-05,
+      "loss": 4.01,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564846632157619e-05,
+      "loss": 4.0139,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564008037406567e-05,
+      "loss": 4.0114,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563169442655515e-05,
+      "loss": 4.0245,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562330847904463e-05,
+      "loss": 4.0013,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561493891033784e-05,
+      "loss": 4.0221,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.560655296282732e-05,
+      "loss": 4.0007,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55981670153168e-05,
+      "loss": 4.0043,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5589781067806286e-05,
+      "loss": 4.0097,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5581395120295766e-05,
+      "loss": 4.0072,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5573025551588975e-05,
+      "loss": 4.0151,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5564639604078455e-05,
+      "loss": 4.0248,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5556253656567935e-05,
+      "loss": 4.0051,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5547867709057415e-05,
+      "loss": 3.9943,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5539481761546895e-05,
+      "loss": 3.9998,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5531112192840104e-05,
+      "loss": 4.0196,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5522726245329584e-05,
+      "loss": 3.9987,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551435667662279e-05,
+      "loss": 4.0124,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550597072911227e-05,
+      "loss": 4.0077,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549758478160175e-05,
+      "loss": 3.9991,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548919883409124e-05,
+      "loss": 4.0074,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548081288658072e-05,
+      "loss": 4.0038,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54724269390702e-05,
+      "loss": 4.0089,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.546404099155968e-05,
+      "loss": 4.0155,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.545565504404916e-05,
+      "loss": 4.0139,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.544726909653864e-05,
+      "loss": 4.0008,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543888314902812e-05,
+      "loss": 4.0063,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54304972015176e-05,
+      "loss": 4.0011,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.542211125400708e-05,
+      "loss": 4.0075,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.541374168530029e-05,
+      "loss": 4.0121,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54053721165935e-05,
+      "loss": 4.0011,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539698616908298e-05,
+      "loss": 4.0067,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538860022157246e-05,
+      "loss": 3.9921,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538021427406194e-05,
+      "loss": 4.0053,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5371828326551424e-05,
+      "loss": 3.9978,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5363442379040904e-05,
+      "loss": 3.9838,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535505643153038e-05,
+      "loss": 4.0032,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534667048401986e-05,
+      "loss": 4.0075,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533830091531307e-05,
+      "loss": 4.0014,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532991496780255e-05,
+      "loss": 3.9973,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5321529020292026e-05,
+      "loss": 4.0028,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5313143072781506e-05,
+      "loss": 4.0016,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5304757125270986e-05,
+      "loss": 3.9928,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5296387556564195e-05,
+      "loss": 4.0169,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528801798785741e-05,
+      "loss": 4.0043,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527963204034689e-05,
+      "loss": 4.0022,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527126247164011e-05,
+      "loss": 4.0097,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526287652412959e-05,
+      "loss": 3.9968,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5254490576619067e-05,
+      "loss": 4.0068,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5246104629108547e-05,
+      "loss": 4.0091,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5237735060401756e-05,
+      "loss": 4.001,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5229349112891236e-05,
+      "loss": 4.002,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5220963165380716e-05,
+      "loss": 4.0027,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5212577217870196e-05,
+      "loss": 4.0044,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520419127035967e-05,
+      "loss": 3.9928,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519580532284915e-05,
+      "loss": 3.9943,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518741937533863e-05,
+      "loss": 4.0008,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517903342782811e-05,
+      "loss": 3.9963,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5170647480317595e-05,
+      "loss": 4.0135,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5162261532807075e-05,
+      "loss": 3.9979,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5153875585296555e-05,
+      "loss": 4.015,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5145489637786035e-05,
+      "loss": 3.9825,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5137120069079244e-05,
+      "loss": 4.0176,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512876687917619e-05,
+      "loss": 4.0009,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512038093166567e-05,
+      "loss": 4.0132,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.511199498415514e-05,
+      "loss": 4.0043,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.510360903664462e-05,
+      "loss": 3.9946,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.50952230891341e-05,
+      "loss": 4.0047,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.508683714162358e-05,
+      "loss": 4.0231,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507845119411306e-05,
+      "loss": 3.985,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507006524660255e-05,
+      "loss": 4.0045,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506167929909203e-05,
+      "loss": 3.9984,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505329335158151e-05,
+      "loss": 4.0013,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504492378287472e-05,
+      "loss": 3.9945,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.50365378353642e-05,
+      "loss": 4.0023,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502816826665741e-05,
+      "loss": 3.9967,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501978231914689e-05,
+      "loss": 4.0037,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501139637163637e-05,
+      "loss": 4.0017,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.043432712554932,
+      "eval_runtime": 295.9678,
+      "eval_samples_per_second": 1289.299,
+      "eval_steps_per_second": 40.292,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.500301042412585e-05,
+      "loss": 4.006,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499462447661533e-05,
+      "loss": 3.9938,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498623852910481e-05,
+      "loss": 4.0054,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497785258159429e-05,
+      "loss": 4.0073,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496946663408377e-05,
+      "loss": 4.0047,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496108068657325e-05,
+      "loss": 4.0019,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4952694739062733e-05,
+      "loss": 4.003,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494430879155221e-05,
+      "loss": 4.0008,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4935922844041687e-05,
+      "loss": 3.9967,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4927536896531166e-05,
+      "loss": 4.0048,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4919150949020646e-05,
+      "loss": 4.0132,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4910765001510126e-05,
+      "loss": 4.0081,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4902395432803336e-05,
+      "loss": 4.0093,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4894009485292815e-05,
+      "loss": 3.986,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4885623537782295e-05,
+      "loss": 4.0089,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4877237590271775e-05,
+      "loss": 3.9841,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4868851642761255e-05,
+      "loss": 4.0006,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486048207405447e-05,
+      "loss": 3.995,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.485209612654395e-05,
+      "loss": 3.9917,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484371017903343e-05,
+      "loss": 3.9987,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483532423152291e-05,
+      "loss": 4.0116,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482693828401239e-05,
+      "loss": 4.0023,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481855233650187e-05,
+      "loss": 3.9992,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481016638899135e-05,
+      "loss": 3.9977,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480178044148083e-05,
+      "loss": 4.0003,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479339449397031e-05,
+      "loss": 3.9937,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478500854645979e-05,
+      "loss": 3.9843,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4776638977753e-05,
+      "loss": 4.0004,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476825303024248e-05,
+      "loss": 3.9934,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.475986708273196e-05,
+      "loss": 3.9962,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.475148113522144e-05,
+      "loss": 3.9952,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.474309518771092e-05,
+      "loss": 4.0013,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.47347092402004e-05,
+      "loss": 3.9921,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.472632329268988e-05,
+      "loss": 3.9997,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4717953723983095e-05,
+      "loss": 4.0007,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4709567776472575e-05,
+      "loss": 3.9956,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470118182896205e-05,
+      "loss": 4.0053,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469279588145153e-05,
+      "loss": 3.9781,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468440993394101e-05,
+      "loss": 3.9891,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467602398643049e-05,
+      "loss": 3.9913,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466763803891997e-05,
+      "loss": 3.9902,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465926847021318e-05,
+      "loss": 3.9931,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465088252270266e-05,
+      "loss": 3.9941,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464249657519214e-05,
+      "loss": 3.9966,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463411062768162e-05,
+      "loss": 3.9971,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4625724680171104e-05,
+      "loss": 3.9915,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461735511146431e-05,
+      "loss": 4.0036,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460896916395379e-05,
+      "loss": 3.9933,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460058321644327e-05,
+      "loss": 3.9737,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459219726893275e-05,
+      "loss": 4.0058,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.458381132142223e-05,
+      "loss": 3.9962,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.457542537391171e-05,
+      "loss": 3.9924,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.456703942640119e-05,
+      "loss": 3.9912,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455865347889067e-05,
+      "loss": 3.9893,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455028391018388e-05,
+      "loss": 3.988,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.454189796267336e-05,
+      "loss": 3.9821,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.453351201516284e-05,
+      "loss": 4.0076,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.452512606765232e-05,
+      "loss": 3.9866,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.451675649894554e-05,
+      "loss": 4.0073,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450837055143502e-05,
+      "loss": 3.9922,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44999846039245e-05,
+      "loss": 3.971,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449161503521771e-05,
+      "loss": 4.001,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4483245466510916e-05,
+      "loss": 3.9847,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4474859519000396e-05,
+      "loss": 3.9796,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4466473571489876e-05,
+      "loss": 3.9854,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4458087623979356e-05,
+      "loss": 3.99,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4449701676468836e-05,
+      "loss": 3.9849,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4441315728958316e-05,
+      "loss": 3.9938,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4432929781447796e-05,
+      "loss": 3.9791,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4424543833937276e-05,
+      "loss": 3.9939,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4416157886426756e-05,
+      "loss": 3.9978,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440778831771997e-05,
+      "loss": 3.9917,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439940237020945e-05,
+      "loss": 3.9977,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439101642269893e-05,
+      "loss": 3.9962,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4382630475188404e-05,
+      "loss": 4.007,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4374244527677884e-05,
+      "loss": 3.9885,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.43658749589711e-05,
+      "loss": 3.9987,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435748901146058e-05,
+      "loss": 3.9895,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4349103063950053e-05,
+      "loss": 3.9875,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4340717116439533e-05,
+      "loss": 3.9957,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433233116892901e-05,
+      "loss": 3.9861,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432396160022223e-05,
+      "loss": 3.9963,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.431557565271171e-05,
+      "loss": 4.0091,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.430718970520119e-05,
+      "loss": 3.9915,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429880375769067e-05,
+      "loss": 3.9738,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429041781018015e-05,
+      "loss": 3.9842,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.428204824147336e-05,
+      "loss": 4.0006,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427366229396284e-05,
+      "loss": 3.9851,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.426527634645232e-05,
+      "loss": 3.9955,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.42568903989418e-05,
+      "loss": 3.993,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424852083023501e-05,
+      "loss": 3.9806,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424013488272449e-05,
+      "loss": 3.9948,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423174893521397e-05,
+      "loss": 3.984,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.422336298770345e-05,
+      "loss": 3.9991,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4214977040192934e-05,
+      "loss": 3.9959,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4206591092682414e-05,
+      "loss": 3.9961,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4198205145171894e-05,
+      "loss": 3.9833,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4189819197661374e-05,
+      "loss": 3.9879,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4181433250150854e-05,
+      "loss": 3.9911,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4173047302640333e-05,
+      "loss": 3.9873,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4164661355129813e-05,
+      "loss": 3.9966,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.415629178642302e-05,
+      "loss": 3.9839,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41479058389125e-05,
+      "loss": 3.9882,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413951989140198e-05,
+      "loss": 3.9782,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413113394389146e-05,
+      "loss": 3.9857,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.412274799638094e-05,
+      "loss": 3.982,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.411437842767415e-05,
+      "loss": 3.9708,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.410599248016363e-05,
+      "loss": 3.9847,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409760653265311e-05,
+      "loss": 3.9932,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40892205851426e-05,
+      "loss": 3.9804,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408083463763207e-05,
+      "loss": 3.9843,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407246506892529e-05,
+      "loss": 3.983,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406407912141477e-05,
+      "loss": 3.9875,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405569317390424e-05,
+      "loss": 3.9807,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404730722639372e-05,
+      "loss": 3.9944,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40389212788832e-05,
+      "loss": 3.988,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403053533137268e-05,
+      "loss": 3.9906,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402216576266589e-05,
+      "loss": 3.9908,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401377981515537e-05,
+      "loss": 3.9806,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400539386764485e-05,
+      "loss": 3.9901,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3997007920134336e-05,
+      "loss": 3.9918,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3988638351427545e-05,
+      "loss": 3.9864,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3980252403917025e-05,
+      "loss": 3.9879,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3971866456406505e-05,
+      "loss": 3.9842,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3963480508895985e-05,
+      "loss": 3.9898,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3955094561385465e-05,
+      "loss": 3.9791,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3946708613874945e-05,
+      "loss": 3.9806,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3938322666364425e-05,
+      "loss": 3.9803,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3929936718853905e-05,
+      "loss": 3.9793,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3921550771343385e-05,
+      "loss": 3.9986,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3913164823832865e-05,
+      "loss": 3.983,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3904795255126074e-05,
+      "loss": 3.9997,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3896409307615554e-05,
+      "loss": 3.9708,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3888023360105034e-05,
+      "loss": 3.9946,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387963741259452e-05,
+      "loss": 3.9925,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3871251465084e-05,
+      "loss": 3.9944,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.386286551757348e-05,
+      "loss": 3.99,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3854479570062953e-05,
+      "loss": 3.9805,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.384609362255243e-05,
+      "loss": 3.9886,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383772405384565e-05,
+      "loss": 4.0046,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382933810633513e-05,
+      "loss": 3.972,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38209521588246e-05,
+      "loss": 3.9859,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.381256621131408e-05,
+      "loss": 3.9852,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380418026380356e-05,
+      "loss": 3.9842,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379581069509678e-05,
+      "loss": 3.9838,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378742474758626e-05,
+      "loss": 3.9863,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377903880007574e-05,
+      "loss": 3.9791,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377065285256522e-05,
+      "loss": 3.9879,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.376228328385843e-05,
+      "loss": 3.9845,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.034024715423584,
+      "eval_runtime": 294.5753,
+      "eval_samples_per_second": 1295.394,
+      "eval_steps_per_second": 40.482,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.375389733634791e-05,
+      "loss": 3.9816,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.374551138883739e-05,
+      "loss": 3.9778,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.373712544132687e-05,
+      "loss": 3.9913,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3728755872620076e-05,
+      "loss": 3.9948,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3720369925109556e-05,
+      "loss": 3.9865,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.371200035640277e-05,
+      "loss": 3.9892,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.370361440889225e-05,
+      "loss": 3.9859,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3695228461381725e-05,
+      "loss": 3.9893,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.368684251387121e-05,
+      "loss": 3.9752,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367845656636069e-05,
+      "loss": 3.9942,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367007061885017e-05,
+      "loss": 3.995,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366168467133965e-05,
+      "loss": 3.9883,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365331510263286e-05,
+      "loss": 3.9992,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.364492915512234e-05,
+      "loss": 3.9678,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363654320761182e-05,
+      "loss": 3.9872,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.36281572601013e-05,
+      "loss": 3.9748,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361977131259078e-05,
+      "loss": 3.984,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361140174388399e-05,
+      "loss": 3.9795,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.360301579637347e-05,
+      "loss": 3.9758,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.359462984886295e-05,
+      "loss": 3.9772,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.358624390135243e-05,
+      "loss": 4.004,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357785795384191e-05,
+      "loss": 3.987,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356947200633139e-05,
+      "loss": 3.9828,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3561086058820876e-05,
+      "loss": 3.983,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3552700111310356e-05,
+      "loss": 3.9874,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3544314163799836e-05,
+      "loss": 3.9761,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3535928216289316e-05,
+      "loss": 3.9697,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.352754226877879e-05,
+      "loss": 3.9844,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351915632126827e-05,
+      "loss": 3.9811,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351077037375775e-05,
+      "loss": 3.9764,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3502400805050965e-05,
+      "loss": 3.9837,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3494031236344174e-05,
+      "loss": 3.9845,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3485645288833654e-05,
+      "loss": 3.9803,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3477259341323134e-05,
+      "loss": 3.9845,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3468873393812614e-05,
+      "loss": 3.9827,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3460487446302094e-05,
+      "loss": 3.9885,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3452101498791574e-05,
+      "loss": 3.9882,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.344373193008479e-05,
+      "loss": 3.9612,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.343534598257426e-05,
+      "loss": 3.974,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.342696003506374e-05,
+      "loss": 3.9755,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341857408755322e-05,
+      "loss": 3.9763,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34101881400427e-05,
+      "loss": 3.9814,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.340180219253218e-05,
+      "loss": 3.981,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339341624502166e-05,
+      "loss": 3.9794,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.338504667631487e-05,
+      "loss": 3.9833,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.337666072880435e-05,
+      "loss": 3.9781,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336827478129383e-05,
+      "loss": 3.9857,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335988883378331e-05,
+      "loss": 3.9802,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.33515028862728e-05,
+      "loss": 3.958,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334311693876228e-05,
+      "loss": 3.9889,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.333473099125176e-05,
+      "loss": 3.9812,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.332634504374124e-05,
+      "loss": 3.9807,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331797547503445e-05,
+      "loss": 3.9745,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3309605906327656e-05,
+      "loss": 3.975,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3301219958817136e-05,
+      "loss": 3.9744,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3292834011306616e-05,
+      "loss": 3.9697,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3284448063796096e-05,
+      "loss": 3.9906,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3276062116285576e-05,
+      "loss": 3.9689,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3267676168775056e-05,
+      "loss": 3.9925,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3259290221264536e-05,
+      "loss": 3.9831,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3250904273754016e-05,
+      "loss": 3.9574,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.324253470504723e-05,
+      "loss": 3.9856,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.323414875753671e-05,
+      "loss": 3.9693,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.322576281002619e-05,
+      "loss": 3.9659,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32173932413194e-05,
+      "loss": 3.9711,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320900729380888e-05,
+      "loss": 3.975,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320062134629836e-05,
+      "loss": 3.9741,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319223539878784e-05,
+      "loss": 3.9785,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318384945127732e-05,
+      "loss": 3.9639,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.31754635037668e-05,
+      "loss": 3.9789,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3167077556256274e-05,
+      "loss": 3.9842,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3158691608745754e-05,
+      "loss": 3.9801,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315032204003897e-05,
+      "loss": 3.9811,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.314193609252845e-05,
+      "loss": 3.9807,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.313355014501793e-05,
+      "loss": 3.9924,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.312516419750741e-05,
+      "loss": 3.9721,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3116794628800626e-05,
+      "loss": 3.9821,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.31084086812901e-05,
+      "loss": 3.9777,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310002273377958e-05,
+      "loss": 3.9753,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309163678626906e-05,
+      "loss": 3.9798,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.308325083875854e-05,
+      "loss": 3.9708,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.307486489124802e-05,
+      "loss": 3.9847,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.306649532254123e-05,
+      "loss": 3.9939,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.305810937503071e-05,
+      "loss": 3.9809,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304972342752019e-05,
+      "loss": 3.9625,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304133748000967e-05,
+      "loss": 3.9693,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3032967911302883e-05,
+      "loss": 3.9822,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.302458196379236e-05,
+      "loss": 3.9746,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.301619601628184e-05,
+      "loss": 3.9766,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.300781006877132e-05,
+      "loss": 3.9774,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29994241212608e-05,
+      "loss": 3.9712,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299103817375028e-05,
+      "loss": 3.9753,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.298265222623976e-05,
+      "loss": 3.973,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.297426627872924e-05,
+      "loss": 3.9809,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.296588033121872e-05,
+      "loss": 3.9813,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29574943837082e-05,
+      "loss": 3.9862,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294910843619768e-05,
+      "loss": 3.9669,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294072248868716e-05,
+      "loss": 3.9736,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.293235291998037e-05,
+      "loss": 3.9782,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.292396697246985e-05,
+      "loss": 3.9723,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.291558102495934e-05,
+      "loss": 3.985,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.290719507744881e-05,
+      "loss": 3.9692,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289884188754576e-05,
+      "loss": 3.9715,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289045594003524e-05,
+      "loss": 3.9654,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.288206999252472e-05,
+      "loss": 3.9781,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.28736840450142e-05,
+      "loss": 3.9677,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.286529809750368e-05,
+      "loss": 3.9542,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.285691214999316e-05,
+      "loss": 3.9718,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284852620248264e-05,
+      "loss": 3.9799,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284014025497211e-05,
+      "loss": 3.9708,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2831770686265326e-05,
+      "loss": 3.9676,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2823384738754806e-05,
+      "loss": 3.9697,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2814998791244286e-05,
+      "loss": 3.9742,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2806612843733766e-05,
+      "loss": 3.9667,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2798226896223246e-05,
+      "loss": 3.9782,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2789840948712725e-05,
+      "loss": 3.9787,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2781455001202205e-05,
+      "loss": 3.9763,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2773069053691685e-05,
+      "loss": 3.9757,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2764699484984894e-05,
+      "loss": 3.9669,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2756313537474374e-05,
+      "loss": 3.9758,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2747927589963854e-05,
+      "loss": 3.9781,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2739541642453334e-05,
+      "loss": 3.9705,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2731172073746543e-05,
+      "loss": 3.9766,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.272280250503976e-05,
+      "loss": 3.9694,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.271441655752924e-05,
+      "loss": 3.974,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.270603061001872e-05,
+      "loss": 3.9635,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26976446625082e-05,
+      "loss": 3.9697,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268927509380141e-05,
+      "loss": 3.9637,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268088914629089e-05,
+      "loss": 3.9692,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.267250319878037e-05,
+      "loss": 3.9839,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.266411725126985e-05,
+      "loss": 3.9689,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.265573130375933e-05,
+      "loss": 3.9835,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264734535624881e-05,
+      "loss": 3.9569,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263897578754202e-05,
+      "loss": 3.9818,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26305898400315e-05,
+      "loss": 3.9814,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2622203892520984e-05,
+      "loss": 3.9783,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2613817945010464e-05,
+      "loss": 3.976,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2605431997499944e-05,
+      "loss": 3.9644,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2597046049989424e-05,
+      "loss": 3.9782,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2588660102478904e-05,
+      "loss": 3.9901,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2580274154968384e-05,
+      "loss": 3.9613,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.257190458626159e-05,
+      "loss": 3.9654,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25635350175548e-05,
+      "loss": 3.9789,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.255514907004428e-05,
+      "loss": 3.9708,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.254676312253376e-05,
+      "loss": 3.9682,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253837717502324e-05,
+      "loss": 3.9741,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253000760631645e-05,
+      "loss": 3.9631,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252163803760967e-05,
+      "loss": 3.9805,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2513252090099147e-05,
+      "loss": 3.965,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.026326656341553,
+      "eval_runtime": 293.2175,
+      "eval_samples_per_second": 1301.392,
+      "eval_steps_per_second": 40.669,
+      "step": 1068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2504866142588627e-05,
+      "loss": 3.9689,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2496480195078107e-05,
+      "loss": 3.9655,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2488094247567586e-05,
+      "loss": 3.9756,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2479708300057066e-05,
+      "loss": 3.9821,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2471322352546546e-05,
+      "loss": 3.9758,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2462936405036026e-05,
+      "loss": 3.9727,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2454550457525506e-05,
+      "loss": 3.9724,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2446164510014986e-05,
+      "loss": 3.9739,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2437778562504466e-05,
+      "loss": 3.9623,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242939261499394e-05,
+      "loss": 3.9845,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242100666748342e-05,
+      "loss": 3.9795,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.24126207199729e-05,
+      "loss": 3.9771,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2404251151266115e-05,
+      "loss": 3.9866,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2395865203755595e-05,
+      "loss": 3.952,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2387479256245075e-05,
+      "loss": 3.9776,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2379093308734555e-05,
+      "loss": 3.9599,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2370707361224035e-05,
+      "loss": 3.9714,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2362321413713515e-05,
+      "loss": 3.963,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2353935466202995e-05,
+      "loss": 3.968,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2345549518692475e-05,
+      "loss": 3.9614,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2337163571181955e-05,
+      "loss": 3.9898,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2328777623671435e-05,
+      "loss": 3.9739,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2320391676160915e-05,
+      "loss": 3.9705,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2312022107454124e-05,
+      "loss": 3.9672,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2303636159943604e-05,
+      "loss": 3.9762,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.229526659123682e-05,
+      "loss": 3.9616,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.22868806437263e-05,
+      "loss": 3.96,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227849469621578e-05,
+      "loss": 3.9695,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227010874870526e-05,
+      "loss": 3.9675,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.226172280119474e-05,
+      "loss": 3.9629,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.225333685368422e-05,
+      "loss": 3.9669,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.22449509061737e-05,
+      "loss": 3.9702,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223656495866318e-05,
+      "loss": 3.972,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222817901115265e-05,
+      "loss": 3.9711,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221979306364213e-05,
+      "loss": 3.9655,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221142349493535e-05,
+      "loss": 3.9793,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220303754742483e-05,
+      "loss": 3.9749,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.219466797871804e-05,
+      "loss": 3.9476,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2186282031207524e-05,
+      "loss": 3.9634,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2177896083697004e-05,
+      "loss": 3.961,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2169510136186484e-05,
+      "loss": 3.9603,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216112418867596e-05,
+      "loss": 3.9685,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.215273824116544e-05,
+      "loss": 3.9697,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.214435229365492e-05,
+      "loss": 3.9603,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.21359663461444e-05,
+      "loss": 3.9678,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212758039863388e-05,
+      "loss": 3.9711,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211919445112336e-05,
+      "loss": 3.9691,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211080850361284e-05,
+      "loss": 3.9722,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.210242255610232e-05,
+      "loss": 3.9475,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2094052987395526e-05,
+      "loss": 3.9708,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.208568341868874e-05,
+      "loss": 3.9723,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207731384998196e-05,
+      "loss": 3.9661,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206892790247143e-05,
+      "loss": 3.9594,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206054195496091e-05,
+      "loss": 3.9656,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.205215600745039e-05,
+      "loss": 3.9521,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.204377005993987e-05,
+      "loss": 3.9599,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.203538411242935e-05,
+      "loss": 3.9762,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.202699816491883e-05,
+      "loss": 3.9587,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201861221740831e-05,
+      "loss": 3.9759,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201022626989779e-05,
+      "loss": 3.9705,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2001856701191e-05,
+      "loss": 3.9461,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.199347075368048e-05,
+      "loss": 3.973,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.198508480616996e-05,
+      "loss": 3.954,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.197669885865944e-05,
+      "loss": 3.9535,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1968312911148926e-05,
+      "loss": 3.9572,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1959926963638406e-05,
+      "loss": 3.9645,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1951541016127886e-05,
+      "loss": 3.959,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1943155068617366e-05,
+      "loss": 3.9595,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1934785499910575e-05,
+      "loss": 3.954,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1926399552400055e-05,
+      "loss": 3.9651,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1918013604889535e-05,
+      "loss": 3.9715,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1909627657379015e-05,
+      "loss": 3.9686,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1901258088672224e-05,
+      "loss": 3.9663,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1892872141161704e-05,
+      "loss": 3.9694,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1884486193651184e-05,
+      "loss": 3.9747,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1876100246140664e-05,
+      "loss": 3.9663,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.186774705623761e-05,
+      "loss": 3.9629,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185937748753082e-05,
+      "loss": 3.9648,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.18509915400203e-05,
+      "loss": 3.9637,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.184260559250978e-05,
+      "loss": 3.9666,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.183421964499926e-05,
+      "loss": 3.961,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.182583369748874e-05,
+      "loss": 3.9685,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181744774997822e-05,
+      "loss": 3.9844,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.18090618024677e-05,
+      "loss": 3.9674,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180067585495718e-05,
+      "loss": 3.9453,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.179228990744666e-05,
+      "loss": 3.9607,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.178390395993614e-05,
+      "loss": 3.9668,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.177551801242562e-05,
+      "loss": 3.9615,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1767148443718834e-05,
+      "loss": 3.9638,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1758762496208314e-05,
+      "loss": 3.9668,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1750376548697794e-05,
+      "loss": 3.9548,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.174199060118727e-05,
+      "loss": 3.962,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.173360465367675e-05,
+      "loss": 3.9602,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.172521870616623e-05,
+      "loss": 3.9706,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1716832758655707e-05,
+      "loss": 3.9639,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1708446811145187e-05,
+      "loss": 3.9752,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1700060863634667e-05,
+      "loss": 3.9508,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1691691294927876e-05,
+      "loss": 3.9619,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1683305347417356e-05,
+      "loss": 3.9651,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1674919399906836e-05,
+      "loss": 3.9617,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1666533452396315e-05,
+      "loss": 3.9747,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.165816388368953e-05,
+      "loss": 3.9525,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164977793617901e-05,
+      "loss": 3.9573,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164139198866849e-05,
+      "loss": 3.9512,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.163300604115797e-05,
+      "loss": 3.9648,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.162462009364745e-05,
+      "loss": 3.9532,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.161625052494066e-05,
+      "loss": 3.9458,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160786457743014e-05,
+      "loss": 3.9566,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159947862991962e-05,
+      "loss": 3.9598,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159110906121283e-05,
+      "loss": 3.959,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.158272311370231e-05,
+      "loss": 3.9624,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157433716619179e-05,
+      "loss": 3.9504,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.156595121868127e-05,
+      "loss": 3.9641,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1557581649974485e-05,
+      "loss": 3.9536,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1549195702463965e-05,
+      "loss": 3.9624,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1540809754953445e-05,
+      "loss": 3.9667,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1532423807442925e-05,
+      "loss": 3.9688,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1524054238736134e-05,
+      "loss": 3.9605,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1515668291225614e-05,
+      "loss": 3.9534,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1507282343715094e-05,
+      "loss": 3.9617,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1498896396204574e-05,
+      "loss": 3.9639,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1490510448694054e-05,
+      "loss": 3.9555,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1482124501183534e-05,
+      "loss": 3.9673,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.147375493247674e-05,
+      "loss": 3.9568,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.146536898496622e-05,
+      "loss": 3.9566,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.145699941625944e-05,
+      "loss": 3.9496,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144861346874892e-05,
+      "loss": 3.9562,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.14402275212384e-05,
+      "loss": 3.9534,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.143184157372788e-05,
+      "loss": 3.9547,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.142345562621736e-05,
+      "loss": 3.9717,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.141506967870684e-05,
+      "loss": 3.9597,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.140668373119632e-05,
+      "loss": 3.971,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.13982977836858e-05,
+      "loss": 3.949,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138992821497901e-05,
+      "loss": 3.9656,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138154226746849e-05,
+      "loss": 3.9688,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.137315631995797e-05,
+      "loss": 3.9674,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.136477037244745e-05,
+      "loss": 3.9599,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.135638442493693e-05,
+      "loss": 3.9562,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.134799847742641e-05,
+      "loss": 3.9652,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133961252991589e-05,
+      "loss": 3.9778,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133122658240537e-05,
+      "loss": 3.9538,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1322857013698576e-05,
+      "loss": 3.9525,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.131448744499179e-05,
+      "loss": 3.9626,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.130610149748127e-05,
+      "loss": 3.9564,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1297715549970745e-05,
+      "loss": 3.9555,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1289329602460225e-05,
+      "loss": 3.9617,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1280943654949705e-05,
+      "loss": 3.9495,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1272557707439185e-05,
+      "loss": 3.9688,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1264171759928665e-05,
+      "loss": 3.9529,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.019377708435059,
+      "eval_runtime": 292.3922,
+      "eval_samples_per_second": 1305.066,
+      "eval_steps_per_second": 40.784,
+      "step": 1144800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1255785812418145e-05,
+      "loss": 3.9539,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.124741624371136e-05,
+      "loss": 3.9565,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123903029620084e-05,
+      "loss": 3.9623,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123064434869032e-05,
+      "loss": 3.9681,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.12222584011798e-05,
+      "loss": 3.9617,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.121387245366928e-05,
+      "loss": 3.9618,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.120551926376622e-05,
+      "loss": 3.9601,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.11971333162557e-05,
+      "loss": 3.9569,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118874736874518e-05,
+      "loss": 3.9556,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118036142123466e-05,
+      "loss": 3.9675,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.117197547372414e-05,
+      "loss": 3.9662,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.116358952621362e-05,
+      "loss": 3.9656,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1155219957506835e-05,
+      "loss": 3.974,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1146834009996315e-05,
+      "loss": 3.944,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1138448062485795e-05,
+      "loss": 3.9637,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1130062114975274e-05,
+      "loss": 3.9472,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1121676167464754e-05,
+      "loss": 3.9599,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1113290219954234e-05,
+      "loss": 3.9485,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1104904272443714e-05,
+      "loss": 3.9551,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1096518324933194e-05,
+      "loss": 3.9517,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1088132377422674e-05,
+      "loss": 3.9763,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107976280871588e-05,
+      "loss": 3.963,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107137686120536e-05,
+      "loss": 3.9537,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.106299091369484e-05,
+      "loss": 3.9547,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.105462134498805e-05,
+      "loss": 3.9647,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.104625177628127e-05,
+      "loss": 3.9505,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.103786582877075e-05,
+      "loss": 3.9474,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102949626006396e-05,
+      "loss": 3.9603,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102111031255344e-05,
+      "loss": 3.9505,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.101272436504292e-05,
+      "loss": 3.9504,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.10043384175324e-05,
+      "loss": 3.957,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.099595247002188e-05,
+      "loss": 3.9563,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.098756652251136e-05,
+      "loss": 3.9609,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097918057500084e-05,
+      "loss": 3.953,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097079462749032e-05,
+      "loss": 3.9579,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09624086799798e-05,
+      "loss": 3.9666,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.095402273246928e-05,
+      "loss": 3.9616,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.094563678495876e-05,
+      "loss": 3.9346,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.093725083744824e-05,
+      "loss": 3.9538,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092888126874145e-05,
+      "loss": 3.9432,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092049532123093e-05,
+      "loss": 3.9496,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0912109373720406e-05,
+      "loss": 3.9565,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0903723426209886e-05,
+      "loss": 3.9584,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0895337478699366e-05,
+      "loss": 3.95,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0886951531188846e-05,
+      "loss": 3.9548,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0878565583678326e-05,
+      "loss": 3.9578,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0870179636167806e-05,
+      "loss": 3.958,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0861793688657286e-05,
+      "loss": 3.9553,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0853424119950495e-05,
+      "loss": 3.9374,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0845038172439975e-05,
+      "loss": 3.9569,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0836652224929455e-05,
+      "loss": 3.9619,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.082826627741894e-05,
+      "loss": 3.9542,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.081988032990842e-05,
+      "loss": 3.9457,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.081151076120163e-05,
+      "loss": 3.9524,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.080312481369111e-05,
+      "loss": 3.9481,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.079473886618059e-05,
+      "loss": 3.9393,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.07863692974738e-05,
+      "loss": 3.966,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.077798334996328e-05,
+      "loss": 3.9468,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076959740245276e-05,
+      "loss": 3.9615,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076121145494224e-05,
+      "loss": 3.96,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.075282550743172e-05,
+      "loss": 3.9348,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.074445593872493e-05,
+      "loss": 3.9533,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.073606999121441e-05,
+      "loss": 3.9474,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.072768404370389e-05,
+      "loss": 3.943,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0719298096193375e-05,
+      "loss": 3.9374,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0710912148682855e-05,
+      "loss": 3.9569,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0702526201172335e-05,
+      "loss": 3.9506,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0694140253661815e-05,
+      "loss": 3.9416,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0685754306151295e-05,
+      "loss": 3.9422,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.067736835864077e-05,
+      "loss": 3.9536,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0668998789933984e-05,
+      "loss": 3.9592,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0660612842423464e-05,
+      "loss": 3.9553,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0652226894912944e-05,
+      "loss": 3.955,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.064384094740242e-05,
+      "loss": 3.9556,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.063547137869563e-05,
+      "loss": 3.9616,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.062708543118511e-05,
+      "loss": 3.9572,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.061869948367459e-05,
+      "loss": 3.9541,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.061031353616407e-05,
+      "loss": 3.953,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.060192758865355e-05,
+      "loss": 3.9504,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.059354164114303e-05,
+      "loss": 3.9571,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.058517207243624e-05,
+      "loss": 3.9449,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.057678612492572e-05,
+      "loss": 3.9583,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.05684001774152e-05,
+      "loss": 3.9666,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056001422990468e-05,
+      "loss": 3.9586,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055162828239416e-05,
+      "loss": 3.9351,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.054324233488364e-05,
+      "loss": 3.9462,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.053485638737312e-05,
+      "loss": 3.9563,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.052648681866633e-05,
+      "loss": 3.9498,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.051810087115581e-05,
+      "loss": 3.9514,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0509714923645294e-05,
+      "loss": 3.9556,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0501328976134774e-05,
+      "loss": 3.9473,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0492943028624254e-05,
+      "loss": 3.9425,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0484573459917466e-05,
+      "loss": 3.9495,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0476187512406946e-05,
+      "loss": 3.9572,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0467801564896426e-05,
+      "loss": 3.9502,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0459415617385906e-05,
+      "loss": 3.9622,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0451029669875386e-05,
+      "loss": 3.9428,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0442660101168595e-05,
+      "loss": 3.9505,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0434274153658075e-05,
+      "loss": 3.953,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.042588820614756e-05,
+      "loss": 3.9487,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.041750225863704e-05,
+      "loss": 3.9584,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0409116311126518e-05,
+      "loss": 3.9466,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0400746742419727e-05,
+      "loss": 3.9457,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0392360794909207e-05,
+      "loss": 3.94,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0383974847398687e-05,
+      "loss": 3.9565,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0375588899888167e-05,
+      "loss": 3.936,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.036720295237765e-05,
+      "loss": 3.9334,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.035881700486713e-05,
+      "loss": 3.9488,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.035044743616034e-05,
+      "loss": 3.9479,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.034206148864982e-05,
+      "loss": 3.9436,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.03336755411393e-05,
+      "loss": 3.9498,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0325289593628776e-05,
+      "loss": 3.9413,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0316903646118256e-05,
+      "loss": 3.9541,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0308534077411472e-05,
+      "loss": 3.9362,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0300148129900952e-05,
+      "loss": 3.9492,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0291762182390425e-05,
+      "loss": 3.9547,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0283376234879905e-05,
+      "loss": 3.9562,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0275023044976853e-05,
+      "loss": 3.9481,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0266637097466333e-05,
+      "loss": 3.9413,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0258251149955813e-05,
+      "loss": 3.9518,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0249865202445293e-05,
+      "loss": 3.9507,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0241479254934773e-05,
+      "loss": 3.9409,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.023309330742425e-05,
+      "loss": 3.9552,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.022470735991373e-05,
+      "loss": 3.9451,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.021632141240321e-05,
+      "loss": 3.9466,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.020793546489269e-05,
+      "loss": 3.9426,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.019954951738217e-05,
+      "loss": 3.9427,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.019116356987165e-05,
+      "loss": 3.9475,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.018279400116486e-05,
+      "loss": 3.9432,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0174408053654342e-05,
+      "loss": 3.9504,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0166022106143822e-05,
+      "loss": 3.9542,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0157636158633302e-05,
+      "loss": 3.9557,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.014926658992651e-05,
+      "loss": 3.9373,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.014088064241599e-05,
+      "loss": 3.9527,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.013249469490547e-05,
+      "loss": 3.9568,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.012410874739495e-05,
+      "loss": 3.954,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0115722799884434e-05,
+      "loss": 3.9461,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0107336852373914e-05,
+      "loss": 3.9436,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0098950904863394e-05,
+      "loss": 3.9517,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0090581336156603e-05,
+      "loss": 3.9638,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0082195388646083e-05,
+      "loss": 3.9431,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0073809441135563e-05,
+      "loss": 3.9395,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0065423493625043e-05,
+      "loss": 3.9511,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0057037546114526e-05,
+      "loss": 3.9421,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0048667977407736e-05,
+      "loss": 3.9439,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0040282029897216e-05,
+      "loss": 3.9533,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0031896082386695e-05,
+      "loss": 3.9369,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0023510134876175e-05,
+      "loss": 3.9564,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0015124187365655e-05,
+      "loss": 3.9409,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.0006738239855135e-05,
+      "loss": 3.9447,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.012637615203857,
+      "eval_runtime": 327.4556,
+      "eval_samples_per_second": 1165.321,
+      "eval_steps_per_second": 36.417,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9998352292344612e-05,
+      "loss": 3.9403,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9989966344834092e-05,
+      "loss": 3.948,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9981580397323572e-05,
+      "loss": 3.9582,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9973194449813052e-05,
+      "loss": 3.9516,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9964808502302532e-05,
+      "loss": 3.9498,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9956422554792012e-05,
+      "loss": 3.9505,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.994803660728149e-05,
+      "loss": 3.946,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9939650659770975e-05,
+      "loss": 3.9351,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9931264712260455e-05,
+      "loss": 3.9631,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9922878764749935e-05,
+      "loss": 3.9484,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9914492817239415e-05,
+      "loss": 3.9588,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9906123248532624e-05,
+      "loss": 3.9618,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9897737301022104e-05,
+      "loss": 3.9353,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9889351353511584e-05,
+      "loss": 3.9481,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9880965406001067e-05,
+      "loss": 3.9384,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9872579458490547e-05,
+      "loss": 3.9433,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9864193510980027e-05,
+      "loss": 3.9443,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.98558075634695e-05,
+      "loss": 3.941,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.984742161595898e-05,
+      "loss": 3.9392,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.983903566844846e-05,
+      "loss": 3.9599,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9830666099741676e-05,
+      "loss": 3.9514,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9822280152231153e-05,
+      "loss": 3.944,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9813894204720633e-05,
+      "loss": 3.9437,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9805508257210113e-05,
+      "loss": 3.9561,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9797122309699593e-05,
+      "loss": 3.9386,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9788752740992805e-05,
+      "loss": 3.9338,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9780366793482285e-05,
+      "loss": 3.9444,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9771980845971765e-05,
+      "loss": 3.9398,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9763594898461245e-05,
+      "loss": 3.9391,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9755208950950725e-05,
+      "loss": 3.9425,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9746823003440205e-05,
+      "loss": 3.9453,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9738453434733414e-05,
+      "loss": 3.9509,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9730067487222897e-05,
+      "loss": 3.9402,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9721681539712377e-05,
+      "loss": 3.9457,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9713295592201857e-05,
+      "loss": 3.9565,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9704909644691337e-05,
+      "loss": 3.9486,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9696540075984546e-05,
+      "loss": 3.9278,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9688154128474026e-05,
+      "loss": 3.9429,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9679768180963506e-05,
+      "loss": 3.9302,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.967138223345299e-05,
+      "loss": 3.9441,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.966299628594247e-05,
+      "loss": 3.9417,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.965461033843195e-05,
+      "loss": 3.9461,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.964622439092143e-05,
+      "loss": 3.9381,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.963783844341091e-05,
+      "loss": 3.9479,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962945249590039e-05,
+      "loss": 3.9426,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.96210829271936e-05,
+      "loss": 3.9484,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.961269697968308e-05,
+      "loss": 3.9431,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.960431103217256e-05,
+      "loss": 3.9271,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.959592508466204e-05,
+      "loss": 3.9421,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.958755551595525e-05,
+      "loss": 3.9539,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.957916956844473e-05,
+      "loss": 3.9405,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.957078362093421e-05,
+      "loss": 3.9344,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9562397673423687e-05,
+      "loss": 3.9397,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9554028104716903e-05,
+      "loss": 3.9385,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9545642157206383e-05,
+      "loss": 3.9288,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9537272588499592e-05,
+      "loss": 3.956,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9528886640989072e-05,
+      "loss": 3.9364,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9520500693478552e-05,
+      "loss": 3.9515,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9512114745968032e-05,
+      "loss": 3.945,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9503728798457515e-05,
+      "loss": 3.9272,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.949534285094699e-05,
+      "loss": 3.9413,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.948695690343647e-05,
+      "loss": 3.9372,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.947857095592595e-05,
+      "loss": 3.928,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.947018500841543e-05,
+      "loss": 3.9285,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.946181543970864e-05,
+      "loss": 3.9471,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.945342949219812e-05,
+      "loss": 3.9373,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.94450435446876e-05,
+      "loss": 3.9357,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.943665759717708e-05,
+      "loss": 3.9262,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.942827164966656e-05,
+      "loss": 3.9438,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.941990208095977e-05,
+      "loss": 3.9501,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9411516133449253e-05,
+      "loss": 3.944,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9403146564742462e-05,
+      "loss": 3.9442,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9394760617231942e-05,
+      "loss": 3.9477,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9386374669721422e-05,
+      "loss": 3.9487,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9377988722210902e-05,
+      "loss": 3.9489,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9369619153504115e-05,
+      "loss": 3.9357,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9361233205993594e-05,
+      "loss": 3.9486,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9352847258483074e-05,
+      "loss": 3.9414,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9344461310972554e-05,
+      "loss": 3.9434,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9336091742265763e-05,
+      "loss": 3.9352,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9327705794755243e-05,
+      "loss": 3.944,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9319319847244723e-05,
+      "loss": 3.954,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9310933899734207e-05,
+      "loss": 3.9469,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9302547952223687e-05,
+      "loss": 3.9262,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9294162004713167e-05,
+      "loss": 3.9339,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9285776057202647e-05,
+      "loss": 3.9449,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9277390109692127e-05,
+      "loss": 3.9427,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9269004162181607e-05,
+      "loss": 3.9418,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9260650972278548e-05,
+      "loss": 3.9403,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9252265024768028e-05,
+      "loss": 3.9398,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9243879077257508e-05,
+      "loss": 3.9319,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9235509508550717e-05,
+      "loss": 3.9385,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9227123561040197e-05,
+      "loss": 3.9463,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9218737613529677e-05,
+      "loss": 3.9384,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.921035166601916e-05,
+      "loss": 3.9533,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.920196571850864e-05,
+      "loss": 3.9333,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.919357977099812e-05,
+      "loss": 3.9395,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.91851938234876e-05,
+      "loss": 3.9407,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.917680787597708e-05,
+      "loss": 3.9393,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.916842192846656e-05,
+      "loss": 3.9471,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.916003598095604e-05,
+      "loss": 3.9354,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.915165003344552e-05,
+      "loss": 3.9321,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9143264085934997e-05,
+      "loss": 3.9338,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9134894517228213e-05,
+      "loss": 3.9435,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9126508569717692e-05,
+      "loss": 3.9275,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9118122622207172e-05,
+      "loss": 3.9242,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9109736674696646e-05,
+      "loss": 3.9357,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9101383484793594e-05,
+      "loss": 3.9417,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9092997537283074e-05,
+      "loss": 3.9335,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9084611589772554e-05,
+      "loss": 3.9393,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9076225642262034e-05,
+      "loss": 3.9301,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9067839694751514e-05,
+      "loss": 3.9438,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9059453747240994e-05,
+      "loss": 3.9274,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.905106779973047e-05,
+      "loss": 3.9419,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9042698231023686e-05,
+      "loss": 3.9436,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9034328662316895e-05,
+      "loss": 3.9481,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9025942714806375e-05,
+      "loss": 3.9362,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9017556767295855e-05,
+      "loss": 3.9346,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9009170819785335e-05,
+      "loss": 3.9379,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9000784872274815e-05,
+      "loss": 3.9416,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8992398924764292e-05,
+      "loss": 3.9308,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.898401297725377e-05,
+      "loss": 3.9399,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.897562702974325e-05,
+      "loss": 3.9374,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.896724108223273e-05,
+      "loss": 3.9411,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.895885513472221e-05,
+      "loss": 3.9283,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8950485566015424e-05,
+      "loss": 3.9304,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8942099618504904e-05,
+      "loss": 3.9343,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8933713670994384e-05,
+      "loss": 3.9374,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8925327723483864e-05,
+      "loss": 3.934,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8916941775973344e-05,
+      "loss": 3.9458,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8908572207266553e-05,
+      "loss": 3.9434,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8900186259756036e-05,
+      "loss": 3.9304,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8891800312245516e-05,
+      "loss": 3.9397,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8883414364734996e-05,
+      "loss": 3.9476,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8875028417224476e-05,
+      "loss": 3.9458,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8866642469713956e-05,
+      "loss": 3.9386,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8858256522203436e-05,
+      "loss": 3.9333,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8849886953496645e-05,
+      "loss": 3.9455,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8841501005986125e-05,
+      "loss": 3.9479,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.883311505847561e-05,
+      "loss": 3.9348,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.882472911096509e-05,
+      "loss": 3.9317,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.881634316345457e-05,
+      "loss": 3.9381,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.880795721594405e-05,
+      "loss": 3.9341,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8799587647237257e-05,
+      "loss": 3.9312,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8791201699726737e-05,
+      "loss": 3.9442,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8782815752216217e-05,
+      "loss": 3.9331,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.87744298047057e-05,
+      "loss": 3.9415,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.876604385719518e-05,
+      "loss": 3.9349,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8757657909684654e-05,
+      "loss": 3.9305,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0083537101745605,
+      "eval_runtime": 292.4053,
+      "eval_samples_per_second": 1305.007,
+      "eval_steps_per_second": 40.782,
+      "step": 1297440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.87493047197816e-05,
+      "loss": 3.9318,
+      "step": 1297920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.874091877227108e-05,
+      "loss": 3.9409,
+      "step": 1298432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8732532824760562e-05,
+      "loss": 3.9462,
+      "step": 1298944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8724146877250042e-05,
+      "loss": 3.9388,
+      "step": 1299456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8715760929739522e-05,
+      "loss": 3.941,
+      "step": 1299968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8707374982229002e-05,
+      "loss": 3.9405,
+      "step": 1300480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8698989034718482e-05,
+      "loss": 3.9356,
+      "step": 1300992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8690603087207955e-05,
+      "loss": 3.9286,
+      "step": 1301504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.868221713969744e-05,
+      "loss": 3.9532,
+      "step": 1302016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.867383119218692e-05,
+      "loss": 3.9384,
+      "step": 1302528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.86654452446764e-05,
+      "loss": 3.9474,
+      "step": 1303040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.865705929716588e-05,
+      "loss": 3.9473,
+      "step": 1303552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8648689728459087e-05,
+      "loss": 3.9323,
+      "step": 1304064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8640303780948567e-05,
+      "loss": 3.9338,
+      "step": 1304576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8631917833438047e-05,
+      "loss": 3.9318,
+      "step": 1305088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.862353188592753e-05,
+      "loss": 3.9285,
+      "step": 1305600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.861514593841701e-05,
+      "loss": 3.9372,
+      "step": 1306112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.860675999090649e-05,
+      "loss": 3.9321,
+      "step": 1306624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.859837404339597e-05,
+      "loss": 3.9286,
+      "step": 1307136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.858998809588545e-05,
+      "loss": 3.9556,
+      "step": 1307648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.858161852717866e-05,
+      "loss": 3.9398,
+      "step": 1308160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.857323257966814e-05,
+      "loss": 3.9351,
+      "step": 1308672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8564846632157623e-05,
+      "loss": 3.9343,
+      "step": 1309184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8556460684647103e-05,
+      "loss": 3.9464,
+      "step": 1309696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8548091115940312e-05,
+      "loss": 3.9251,
+      "step": 1310208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8539705168429792e-05,
+      "loss": 3.9263,
+      "step": 1310720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8531319220919272e-05,
+      "loss": 3.9323,
+      "step": 1311232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8522933273408752e-05,
+      "loss": 3.9359,
+      "step": 1311744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8514563704701964e-05,
+      "loss": 3.931,
+      "step": 1312256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8506177757191444e-05,
+      "loss": 3.9346,
+      "step": 1312768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8497791809680924e-05,
+      "loss": 3.9363,
+      "step": 1313280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8489405862170404e-05,
+      "loss": 3.9404,
+      "step": 1313792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8481019914659884e-05,
+      "loss": 3.928,
+      "step": 1314304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8472666724756826e-05,
+      "loss": 3.9398,
+      "step": 1314816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8464280777246306e-05,
+      "loss": 3.9433,
+      "step": 1315328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8455894829735786e-05,
+      "loss": 3.9404,
+      "step": 1315840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8447508882225266e-05,
+      "loss": 3.9194,
+      "step": 1316352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8439122934714746e-05,
+      "loss": 3.9293,
+      "step": 1316864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8430736987204226e-05,
+      "loss": 3.9258,
+      "step": 1317376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8422351039693706e-05,
+      "loss": 3.9334,
+      "step": 1317888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8413965092183185e-05,
+      "loss": 3.9296,
+      "step": 1318400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8405595523476398e-05,
+      "loss": 3.9355,
+      "step": 1318912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8397209575965878e-05,
+      "loss": 3.9278,
+      "step": 1319424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8388840007259087e-05,
+      "loss": 3.9359,
+      "step": 1319936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8380454059748567e-05,
+      "loss": 3.9365,
+      "step": 1320448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8372068112238047e-05,
+      "loss": 3.9366,
+      "step": 1320960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.836368216472753e-05,
+      "loss": 3.9372,
+      "step": 1321472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.835529621721701e-05,
+      "loss": 3.9159,
+      "step": 1321984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.834691026970649e-05,
+      "loss": 3.933,
+      "step": 1322496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.83385407009997e-05,
+      "loss": 3.9396,
+      "step": 1323008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.833015475348918e-05,
+      "loss": 3.9344,
+      "step": 1323520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.832176880597866e-05,
+      "loss": 3.9288,
+      "step": 1324032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.831338285846814e-05,
+      "loss": 3.9239,
+      "step": 1324544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8304996910957616e-05,
+      "loss": 3.9319,
+      "step": 1325056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.829662734225083e-05,
+      "loss": 3.9155,
+      "step": 1325568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.828824139474031e-05,
+      "loss": 3.9446,
+      "step": 1326080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8279855447229785e-05,
+      "loss": 3.9285,
+      "step": 1326592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8271469499719268e-05,
+      "loss": 3.9412,
+      "step": 1327104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8263083552208748e-05,
+      "loss": 3.9379,
+      "step": 1327616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8254697604698228e-05,
+      "loss": 3.9165,
+      "step": 1328128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8246328035991437e-05,
+      "loss": 3.9307,
+      "step": 1328640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8237942088480917e-05,
+      "loss": 3.9299,
+      "step": 1329152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8229556140970397e-05,
+      "loss": 3.916,
+      "step": 1329664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8221170193459877e-05,
+      "loss": 3.9179,
+      "step": 1330176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8212784245949357e-05,
+      "loss": 3.9382,
+      "step": 1330688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.820439829843884e-05,
+      "loss": 3.9297,
+      "step": 1331200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.819601235092832e-05,
+      "loss": 3.9254,
+      "step": 1331712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.81876264034178e-05,
+      "loss": 3.9156,
+      "step": 1332224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.817925683471101e-05,
+      "loss": 3.9344,
+      "step": 1332736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.817087088720049e-05,
+      "loss": 3.9366,
+      "step": 1333248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.81625013184937e-05,
+      "loss": 3.9349,
+      "step": 1333760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.815411537098318e-05,
+      "loss": 3.934,
+      "step": 1334272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.814572942347266e-05,
+      "loss": 3.9422,
+      "step": 1334784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.813734347596214e-05,
+      "loss": 3.9375,
+      "step": 1335296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.812895752845162e-05,
+      "loss": 3.9418,
+      "step": 1335808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.812058795974483e-05,
+      "loss": 3.9238,
+      "step": 1336320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.811220201223431e-05,
+      "loss": 3.9383,
+      "step": 1336832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8103816064723794e-05,
+      "loss": 3.9322,
+      "step": 1337344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8095430117213274e-05,
+      "loss": 3.9319,
+      "step": 1337856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8087060548506483e-05,
+      "loss": 3.9289,
+      "step": 1338368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8078674600995963e-05,
+      "loss": 3.931,
+      "step": 1338880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8070288653485443e-05,
+      "loss": 3.9453,
+      "step": 1339392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8061902705974923e-05,
+      "loss": 3.9399,
+      "step": 1339904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8053516758464403e-05,
+      "loss": 3.9166,
+      "step": 1340416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8045130810953886e-05,
+      "loss": 3.9253,
+      "step": 1340928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8036744863443366e-05,
+      "loss": 3.933,
+      "step": 1341440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8028358915932846e-05,
+      "loss": 3.9335,
+      "step": 1341952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8019989347226055e-05,
+      "loss": 3.9344,
+      "step": 1342464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8011603399715535e-05,
+      "loss": 3.9313,
+      "step": 1342976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8003217452205015e-05,
+      "loss": 3.9288,
+      "step": 1343488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7994831504694495e-05,
+      "loss": 3.9221,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.798644555718397e-05,
+      "loss": 3.9318,
+      "step": 1344512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.797805960967345e-05,
+      "loss": 3.9308,
+      "step": 1345024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.796967366216293e-05,
+      "loss": 3.9318,
+      "step": 1345536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.796128771465241e-05,
+      "loss": 3.9405,
+      "step": 1346048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7952918145945624e-05,
+      "loss": 3.9312,
+      "step": 1346560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7944532198435104e-05,
+      "loss": 3.9268,
+      "step": 1347072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7936146250924584e-05,
+      "loss": 3.9307,
+      "step": 1347584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7927760303414064e-05,
+      "loss": 3.9317,
+      "step": 1348096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7919374355903544e-05,
+      "loss": 3.938,
+      "step": 1348608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7911004787196753e-05,
+      "loss": 3.9254,
+      "step": 1349120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.790263521848997e-05,
+      "loss": 3.9236,
+      "step": 1349632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7894249270979445e-05,
+      "loss": 3.9233,
+      "step": 1350144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7885863323468925e-05,
+      "loss": 3.9331,
+      "step": 1350656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7877477375958405e-05,
+      "loss": 3.9185,
+      "step": 1351168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7869091428447885e-05,
+      "loss": 3.9115,
+      "step": 1351680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7860705480937365e-05,
+      "loss": 3.9275,
+      "step": 1352192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7852319533426845e-05,
+      "loss": 3.9312,
+      "step": 1352704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7843933585916325e-05,
+      "loss": 3.9238,
+      "step": 1353216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.783554763840581e-05,
+      "loss": 3.9277,
+      "step": 1353728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7827178069699017e-05,
+      "loss": 3.9238,
+      "step": 1354240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7818808500992227e-05,
+      "loss": 3.9338,
+      "step": 1354752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7810422553481706e-05,
+      "loss": 3.9188,
+      "step": 1355264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7802036605971186e-05,
+      "loss": 3.9295,
+      "step": 1355776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.779365065846067e-05,
+      "loss": 3.9364,
+      "step": 1356288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.778526471095015e-05,
+      "loss": 3.9388,
+      "step": 1356800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.777689514224336e-05,
+      "loss": 3.9261,
+      "step": 1357312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.776850919473284e-05,
+      "loss": 3.9238,
+      "step": 1357824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.776012324722232e-05,
+      "loss": 3.9312,
+      "step": 1358336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.77517372997118e-05,
+      "loss": 3.9345,
+      "step": 1358848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.774335135220128e-05,
+      "loss": 3.9207,
+      "step": 1359360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7734965404690762e-05,
+      "loss": 3.9325,
+      "step": 1359872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7726579457180242e-05,
+      "loss": 3.9241,
+      "step": 1360384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7718193509669722e-05,
+      "loss": 3.9365,
+      "step": 1360896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.770982394096293e-05,
+      "loss": 3.9193,
+      "step": 1361408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.770143799345241e-05,
+      "loss": 3.9206,
+      "step": 1361920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7693068424745623e-05,
+      "loss": 3.9243,
+      "step": 1362432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7684682477235103e-05,
+      "loss": 3.9285,
+      "step": 1362944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7676296529724583e-05,
+      "loss": 3.9232,
+      "step": 1363456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7667910582214063e-05,
+      "loss": 3.9422,
+      "step": 1363968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7659524634703543e-05,
+      "loss": 3.9326,
+      "step": 1364480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7651138687193023e-05,
+      "loss": 3.9245,
+      "step": 1364992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7642752739682503e-05,
+      "loss": 3.9268,
+      "step": 1365504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7634366792171983e-05,
+      "loss": 3.9395,
+      "step": 1366016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.762598084466146e-05,
+      "loss": 3.9359,
+      "step": 1366528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7617611275954676e-05,
+      "loss": 3.9293,
+      "step": 1367040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7609241707247885e-05,
+      "loss": 3.9253,
+      "step": 1367552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7600855759737365e-05,
+      "loss": 3.9346,
+      "step": 1368064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7592469812226845e-05,
+      "loss": 3.9363,
+      "step": 1368576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7584083864716325e-05,
+      "loss": 3.9257,
+      "step": 1369088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7575697917205805e-05,
+      "loss": 3.9256,
+      "step": 1369600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.756731196969528e-05,
+      "loss": 3.9287,
+      "step": 1370112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.755892602218476e-05,
+      "loss": 3.9262,
+      "step": 1370624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.755054007467424e-05,
+      "loss": 3.9222,
+      "step": 1371136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.754215412716372e-05,
+      "loss": 3.9366,
+      "step": 1371648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.75337681796532e-05,
+      "loss": 3.924,
+      "step": 1372160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7525398610946413e-05,
+      "loss": 3.9309,
+      "step": 1372672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7517012663435893e-05,
+      "loss": 3.9248,
+      "step": 1373184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7508626715925373e-05,
+      "loss": 3.9213,
+      "step": 1373696
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.004011154174805,
+      "eval_runtime": 320.1238,
+      "eval_samples_per_second": 1192.011,
+      "eval_steps_per_second": 37.251,
+      "step": 1373760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7500240768414853e-05,
+      "loss": 3.9206,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7491854820904333e-05,
+      "loss": 3.9317,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7483468873393813e-05,
+      "loss": 3.9375,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7475082925883293e-05,
+      "loss": 3.9282,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7466696978372773e-05,
+      "loss": 3.9325,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7458311030862256e-05,
+      "loss": 3.9281,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7449925083351736e-05,
+      "loss": 3.9305,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7441539135841216e-05,
+      "loss": 3.916,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7433153188330696e-05,
+      "loss": 3.9401,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.742476724082017e-05,
+      "loss": 3.9355,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.741638129330965e-05,
+      "loss": 3.9328,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7407995345799133e-05,
+      "loss": 3.9402,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.739962577709235e-05,
+      "loss": 3.9212,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7391239829581822e-05,
+      "loss": 3.9233,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7382853882071302e-05,
+      "loss": 3.9318,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7374467934560782e-05,
+      "loss": 3.9159,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.736608198705026e-05,
+      "loss": 3.9258,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.735769603953974e-05,
+      "loss": 3.9213,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7349310092029225e-05,
+      "loss": 3.92,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7340924144518705e-05,
+      "loss": 3.9458,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7332538197008185e-05,
+      "loss": 3.9272,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7324152249497665e-05,
+      "loss": 3.9316,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7315766301987145e-05,
+      "loss": 3.9216,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7307380354476625e-05,
+      "loss": 3.9367,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7299027164573566e-05,
+      "loss": 3.912,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7290641217063046e-05,
+      "loss": 3.9207,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7282255269552526e-05,
+      "loss": 3.919,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7273869322042006e-05,
+      "loss": 3.9289,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7265483374531486e-05,
+      "loss": 3.9198,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7257097427020966e-05,
+      "loss": 3.9221,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7248711479510446e-05,
+      "loss": 3.9299,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7240325531999926e-05,
+      "loss": 3.9322,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.723195596329314e-05,
+      "loss": 3.9159,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.722357001578262e-05,
+      "loss": 3.9317,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.72151840682721e-05,
+      "loss": 3.9324,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.720679812076158e-05,
+      "loss": 3.9274,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.719841217325106e-05,
+      "loss": 3.9098,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7190026225740535e-05,
+      "loss": 3.9232,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7181640278230015e-05,
+      "loss": 3.9116,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7173254330719495e-05,
+      "loss": 3.9275,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.716488476201271e-05,
+      "loss": 3.9225,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7156498814502184e-05,
+      "loss": 3.9273,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7148112866991664e-05,
+      "loss": 3.9189,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.713974329828488e-05,
+      "loss": 3.9253,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7131357350774356e-05,
+      "loss": 3.9284,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7122971403263836e-05,
+      "loss": 3.9289,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7114585455753316e-05,
+      "loss": 3.9284,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7106199508242796e-05,
+      "loss": 3.9072,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7097813560732276e-05,
+      "loss": 3.9209,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7089427613221756e-05,
+      "loss": 3.936,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.708105804451497e-05,
+      "loss": 3.9206,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.707267209700445e-05,
+      "loss": 3.9189,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.706428614949393e-05,
+      "loss": 3.9185,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.705590020198341e-05,
+      "loss": 3.9206,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.704751425447289e-05,
+      "loss": 3.9095,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.703912830696237e-05,
+      "loss": 3.931,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.703075873825558e-05,
+      "loss": 3.9213,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.702237279074506e-05,
+      "loss": 3.9261,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.701398684323454e-05,
+      "loss": 3.9335,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.700560089572402e-05,
+      "loss": 3.9059,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.69972149482135e-05,
+      "loss": 3.9224,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698882900070298e-05,
+      "loss": 3.9168,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698044305319246e-05,
+      "loss": 3.9111,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.697205710568194e-05,
+      "loss": 3.9094,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6963687536975153e-05,
+      "loss": 3.9254,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6955301589464633e-05,
+      "loss": 3.923,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6946915641954113e-05,
+      "loss": 3.9159,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6938546073247322e-05,
+      "loss": 3.9051,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6930160125736802e-05,
+      "loss": 3.9276,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6921774178226282e-05,
+      "loss": 3.9323,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6913388230715765e-05,
+      "loss": 3.9227,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6905002283205245e-05,
+      "loss": 3.9262,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.689661633569472e-05,
+      "loss": 3.9347,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6888246766987934e-05,
+      "loss": 3.9281,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6879860819477414e-05,
+      "loss": 3.9331,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6871491250770627e-05,
+      "loss": 3.9119,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6863105303260107e-05,
+      "loss": 3.9303,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6854719355749587e-05,
+      "loss": 3.9268,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6846333408239067e-05,
+      "loss": 3.9202,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.683794746072854e-05,
+      "loss": 3.9243,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.682956151321802e-05,
+      "loss": 3.923,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6821175565707503e-05,
+      "loss": 3.9381,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6812789618196983e-05,
+      "loss": 3.928,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6804403670686463e-05,
+      "loss": 3.9095,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6796034101979672e-05,
+      "loss": 3.9145,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6787648154469152e-05,
+      "loss": 3.9216,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6779278585762368e-05,
+      "loss": 3.9256,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6770892638251844e-05,
+      "loss": 3.9267,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6762506690741324e-05,
+      "loss": 3.9206,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6754120743230804e-05,
+      "loss": 3.9184,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6745734795720284e-05,
+      "loss": 3.9134,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6737348848209764e-05,
+      "loss": 3.925,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6728962900699244e-05,
+      "loss": 3.9169,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6720576953188724e-05,
+      "loss": 3.9247,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6712191005678204e-05,
+      "loss": 3.9318,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6703805058167684e-05,
+      "loss": 3.9259,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6695419110657167e-05,
+      "loss": 3.9166,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6687049541950377e-05,
+      "loss": 3.9236,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6678663594439856e-05,
+      "loss": 3.9212,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6670277646929336e-05,
+      "loss": 3.9278,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6661891699418816e-05,
+      "loss": 3.9169,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6653505751908296e-05,
+      "loss": 3.9162,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.664513618320151e-05,
+      "loss": 3.9142,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.663675023569099e-05,
+      "loss": 3.9244,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.662836428818047e-05,
+      "loss": 3.9117,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.661997834066995e-05,
+      "loss": 3.9058,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.661159239315943e-05,
+      "loss": 3.914,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6603206445648905e-05,
+      "loss": 3.9231,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6594820498138385e-05,
+      "loss": 3.912,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6586434550627865e-05,
+      "loss": 3.9224,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.657806498192108e-05,
+      "loss": 3.912,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6569679034410554e-05,
+      "loss": 3.9235,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.656130946570377e-05,
+      "loss": 3.9104,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.655292351819325e-05,
+      "loss": 3.9155,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.654453757068273e-05,
+      "loss": 3.9288,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6536151623172207e-05,
+      "loss": 3.9284,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6527765675661687e-05,
+      "loss": 3.9191,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6519379728151166e-05,
+      "loss": 3.9164,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6510993780640646e-05,
+      "loss": 3.9169,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6502607833130126e-05,
+      "loss": 3.9253,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6494221885619606e-05,
+      "loss": 3.9126,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.648583593810909e-05,
+      "loss": 3.9266,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6477482748206028e-05,
+      "loss": 3.9153,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6469096800695508e-05,
+      "loss": 3.926,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6460710853184988e-05,
+      "loss": 3.9068,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6452324905674468e-05,
+      "loss": 3.9158,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.644395533696768e-05,
+      "loss": 3.9125,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.643556938945716e-05,
+      "loss": 3.918,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.642718344194664e-05,
+      "loss": 3.914,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.641879749443612e-05,
+      "loss": 3.9336,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.64104115469256e-05,
+      "loss": 3.9218,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.640202559941508e-05,
+      "loss": 3.9192,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.639363965190456e-05,
+      "loss": 3.9114,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6385270083197772e-05,
+      "loss": 3.9287,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6376884135687252e-05,
+      "loss": 3.9262,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6368498188176732e-05,
+      "loss": 3.9229,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6360112240666212e-05,
+      "loss": 3.9169,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6351726293155692e-05,
+      "loss": 3.9233,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6343340345645172e-05,
+      "loss": 3.9268,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6334954398134652e-05,
+      "loss": 3.9167,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6326568450624136e-05,
+      "loss": 3.9151,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6318198881917345e-05,
+      "loss": 3.9212,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6309812934406825e-05,
+      "loss": 3.9162,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6301426986896305e-05,
+      "loss": 3.9127,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6293041039385785e-05,
+      "loss": 3.9194,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6284655091875264e-05,
+      "loss": 3.9223,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6276301901972206e-05,
+      "loss": 3.9185,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6267915954461686e-05,
+      "loss": 3.9153,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6259530006951166e-05,
+      "loss": 3.9105,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.000067234039307,
+      "eval_runtime": 291.4889,
+      "eval_samples_per_second": 1309.11,
+      "eval_steps_per_second": 40.911,
+      "step": 1450080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6251144059440646e-05,
+      "loss": 3.9189,
+      "step": 1450496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6242758111930126e-05,
+      "loss": 3.9202,
+      "step": 1451008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6234372164419606e-05,
+      "loss": 3.9285,
+      "step": 1451520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6225986216909086e-05,
+      "loss": 3.92,
+      "step": 1452032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6217600269398562e-05,
+      "loss": 3.9268,
+      "step": 1452544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6209230700691778e-05,
+      "loss": 3.9171,
+      "step": 1453056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6200844753181258e-05,
+      "loss": 3.9211,
+      "step": 1453568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6192458805670738e-05,
+      "loss": 3.9091,
+      "step": 1454080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6184072858160215e-05,
+      "loss": 3.9314,
+      "step": 1454592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.617570328945343e-05,
+      "loss": 3.9252,
+      "step": 1455104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.616731734194291e-05,
+      "loss": 3.9247,
+      "step": 1455616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.615894777323612e-05,
+      "loss": 3.9316,
+      "step": 1456128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.615057820452933e-05,
+      "loss": 3.9092,
+      "step": 1456640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6142192257018812e-05,
+      "loss": 3.9156,
+      "step": 1457152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6133806309508292e-05,
+      "loss": 3.9234,
+      "step": 1457664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6125420361997772e-05,
+      "loss": 3.9102,
+      "step": 1458176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6117034414487252e-05,
+      "loss": 3.9121,
+      "step": 1458688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6108648466976732e-05,
+      "loss": 3.9111,
+      "step": 1459200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6100262519466212e-05,
+      "loss": 3.9119,
+      "step": 1459712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.609187657195569e-05,
+      "loss": 3.939,
+      "step": 1460224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.608349062444517e-05,
+      "loss": 3.9185,
+      "step": 1460736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.607510467693465e-05,
+      "loss": 3.9201,
+      "step": 1461248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.606671872942413e-05,
+      "loss": 3.916,
+      "step": 1461760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6058332781913608e-05,
+      "loss": 3.9243,
+      "step": 1462272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6049946834403088e-05,
+      "loss": 3.9063,
+      "step": 1462784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6041577265696297e-05,
+      "loss": 3.909,
+      "step": 1463296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6033191318185777e-05,
+      "loss": 3.9116,
+      "step": 1463808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.602480537067526e-05,
+      "loss": 3.92,
+      "step": 1464320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.601641942316474e-05,
+      "loss": 3.9085,
+      "step": 1464832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.600803347565422e-05,
+      "loss": 3.9153,
+      "step": 1465344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.599966390694743e-05,
+      "loss": 3.9164,
+      "step": 1465856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.599127795943691e-05,
+      "loss": 3.9227,
+      "step": 1466368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.598289201192639e-05,
+      "loss": 3.9138,
+      "step": 1466880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.597450606441587e-05,
+      "loss": 3.9213,
+      "step": 1467392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5966120116905353e-05,
+      "loss": 3.9228,
+      "step": 1467904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5957734169394833e-05,
+      "loss": 3.9195,
+      "step": 1468416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5949364600688042e-05,
+      "loss": 3.9011,
+      "step": 1468928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5940978653177522e-05,
+      "loss": 3.9171,
+      "step": 1469440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5932592705667002e-05,
+      "loss": 3.9018,
+      "step": 1469952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5924206758156482e-05,
+      "loss": 3.918,
+      "step": 1470464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5915820810645962e-05,
+      "loss": 3.9086,
+      "step": 1470976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5907434863135445e-05,
+      "loss": 3.9198,
+      "step": 1471488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5899065294428654e-05,
+      "loss": 3.9123,
+      "step": 1472000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5890679346918134e-05,
+      "loss": 3.913,
+      "step": 1472512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5882293399407614e-05,
+      "loss": 3.9196,
+      "step": 1473024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5873907451897094e-05,
+      "loss": 3.9169,
+      "step": 1473536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5865521504386574e-05,
+      "loss": 3.9194,
+      "step": 1474048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.585713555687605e-05,
+      "loss": 3.9057,
+      "step": 1474560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.584874960936553e-05,
+      "loss": 3.9071,
+      "step": 1475072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5840380040658746e-05,
+      "loss": 3.9268,
+      "step": 1475584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5831994093148226e-05,
+      "loss": 3.9132,
+      "step": 1476096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.58236081456377e-05,
+      "loss": 3.9076,
+      "step": 1476608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5815222198127183e-05,
+      "loss": 3.9112,
+      "step": 1477120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5806836250616663e-05,
+      "loss": 3.9163,
+      "step": 1477632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5798450303106143e-05,
+      "loss": 3.9007,
+      "step": 1478144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5790064355595623e-05,
+      "loss": 3.9179,
+      "step": 1478656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5781694786888832e-05,
+      "loss": 3.9127,
+      "step": 1479168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5773308839378312e-05,
+      "loss": 3.9187,
+      "step": 1479680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5764922891867792e-05,
+      "loss": 3.923,
+      "step": 1480192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5756536944357275e-05,
+      "loss": 3.9042,
+      "step": 1480704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5748150996846755e-05,
+      "loss": 3.9052,
+      "step": 1481216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5739765049336235e-05,
+      "loss": 3.9134,
+      "step": 1481728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5731379101825715e-05,
+      "loss": 3.9011,
+      "step": 1482240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5722993154315195e-05,
+      "loss": 3.9033,
+      "step": 1482752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5714623585608404e-05,
+      "loss": 3.9139,
+      "step": 1483264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5706237638097884e-05,
+      "loss": 3.9178,
+      "step": 1483776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5697851690587367e-05,
+      "loss": 3.9047,
+      "step": 1484288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5689465743076847e-05,
+      "loss": 3.8983,
+      "step": 1484800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5681096174370056e-05,
+      "loss": 3.914,
+      "step": 1485312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5672710226859536e-05,
+      "loss": 3.922,
+      "step": 1485824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5664340658152745e-05,
+      "loss": 3.9112,
+      "step": 1486336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.565595471064223e-05,
+      "loss": 3.9199,
+      "step": 1486848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.564756876313171e-05,
+      "loss": 3.9215,
+      "step": 1487360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.563918281562119e-05,
+      "loss": 3.9216,
+      "step": 1487872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.563079686811067e-05,
+      "loss": 3.9276,
+      "step": 1488384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5622427299403878e-05,
+      "loss": 3.9012,
+      "step": 1488896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5614041351893358e-05,
+      "loss": 3.9192,
+      "step": 1489408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5605655404382838e-05,
+      "loss": 3.9198,
+      "step": 1489920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.559728583567605e-05,
+      "loss": 3.9074,
+      "step": 1490432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.558889988816553e-05,
+      "loss": 3.92,
+      "step": 1490944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.558051394065501e-05,
+      "loss": 3.9114,
+      "step": 1491456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.557212799314449e-05,
+      "loss": 3.9259,
+      "step": 1491968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.556374204563397e-05,
+      "loss": 3.9219,
+      "step": 1492480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.555535609812345e-05,
+      "loss": 3.898,
+      "step": 1492992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.554697015061293e-05,
+      "loss": 3.9094,
+      "step": 1493504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.553858420310241e-05,
+      "loss": 3.9105,
+      "step": 1494016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5530198255591886e-05,
+      "loss": 3.9164,
+      "step": 1494528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5521828686885102e-05,
+      "loss": 3.9178,
+      "step": 1495040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5513442739374582e-05,
+      "loss": 3.91,
+      "step": 1495552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5505056791864055e-05,
+      "loss": 3.9109,
+      "step": 1496064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.549667084435354e-05,
+      "loss": 3.9049,
+      "step": 1496576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.548828489684302e-05,
+      "loss": 3.9159,
+      "step": 1497088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.54798989493325e-05,
+      "loss": 3.905,
+      "step": 1497600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.547151300182198e-05,
+      "loss": 3.9202,
+      "step": 1498112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.546312705431146e-05,
+      "loss": 3.9239,
+      "step": 1498624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5454757485604668e-05,
+      "loss": 3.9156,
+      "step": 1499136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5446371538094148e-05,
+      "loss": 3.9079,
+      "step": 1499648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.543800196938736e-05,
+      "loss": 3.9144,
+      "step": 1500160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.542961602187684e-05,
+      "loss": 3.9102,
+      "step": 1500672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.542123007436632e-05,
+      "loss": 3.9205,
+      "step": 1501184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.54128441268558e-05,
+      "loss": 3.9127,
+      "step": 1501696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.540445817934528e-05,
+      "loss": 3.9083,
+      "step": 1502208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.539607223183476e-05,
+      "loss": 3.9041,
+      "step": 1502720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.538768628432424e-05,
+      "loss": 3.918,
+      "step": 1503232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5379300336813723e-05,
+      "loss": 3.8984,
+      "step": 1503744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5370930768106932e-05,
+      "loss": 3.8984,
+      "step": 1504256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.536256119940014e-05,
+      "loss": 3.9043,
+      "step": 1504768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.535417525188962e-05,
+      "loss": 3.9119,
+      "step": 1505280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5345805683182834e-05,
+      "loss": 3.9066,
+      "step": 1505792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5337419735672314e-05,
+      "loss": 3.9198,
+      "step": 1506304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5329033788161794e-05,
+      "loss": 3.9051,
+      "step": 1506816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5320647840651274e-05,
+      "loss": 3.9084,
+      "step": 1507328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5312261893140754e-05,
+      "loss": 3.9063,
+      "step": 1507840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5303875945630234e-05,
+      "loss": 3.9017,
+      "step": 1508352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5295489998119714e-05,
+      "loss": 3.925,
+      "step": 1508864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5287104050609193e-05,
+      "loss": 3.9188,
+      "step": 1509376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5278734481902406e-05,
+      "loss": 3.9114,
+      "step": 1509888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5270348534391886e-05,
+      "loss": 3.9109,
+      "step": 1510400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5261962586881366e-05,
+      "loss": 3.9026,
+      "step": 1510912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5253576639370846e-05,
+      "loss": 3.9159,
+      "step": 1511424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5245190691860326e-05,
+      "loss": 3.9052,
+      "step": 1511936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5236804744349806e-05,
+      "loss": 3.9186,
+      "step": 1512448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5228418796839286e-05,
+      "loss": 3.9049,
+      "step": 1512960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.522003284932877e-05,
+      "loss": 3.9149,
+      "step": 1513472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.521164690181825e-05,
+      "loss": 3.904,
+      "step": 1513984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5203260954307722e-05,
+      "loss": 3.9085,
+      "step": 1514496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5194891385600938e-05,
+      "loss": 3.9025,
+      "step": 1515008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5186521816894147e-05,
+      "loss": 3.9128,
+      "step": 1515520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.517813586938363e-05,
+      "loss": 3.9049,
+      "step": 1516032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.516974992187311e-05,
+      "loss": 3.9242,
+      "step": 1516544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.516136397436259e-05,
+      "loss": 3.9167,
+      "step": 1517056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.515297802685207e-05,
+      "loss": 3.9116,
+      "step": 1517568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.514460845814528e-05,
+      "loss": 3.9002,
+      "step": 1518080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.513622251063476e-05,
+      "loss": 3.9216,
+      "step": 1518592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5127852941927972e-05,
+      "loss": 3.9163,
+      "step": 1519104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5119466994417452e-05,
+      "loss": 3.9206,
+      "step": 1519616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5111081046906932e-05,
+      "loss": 3.91,
+      "step": 1520128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5102695099396412e-05,
+      "loss": 3.9106,
+      "step": 1520640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5094309151885892e-05,
+      "loss": 3.9185,
+      "step": 1521152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5085923204375368e-05,
+      "loss": 3.9137,
+      "step": 1521664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5077537256864848e-05,
+      "loss": 3.906,
+      "step": 1522176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5069151309354328e-05,
+      "loss": 3.9144,
+      "step": 1522688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5060765361843808e-05,
+      "loss": 3.9079,
+      "step": 1523200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5052379414333288e-05,
+      "loss": 3.9048,
+      "step": 1523712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5043993466822768e-05,
+      "loss": 3.9126,
+      "step": 1524224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5035607519312248e-05,
+      "loss": 3.917,
+      "step": 1524736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.502723795060546e-05,
+      "loss": 3.9077,
+      "step": 1525248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.501885200309494e-05,
+      "loss": 3.9052,
+      "step": 1525760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.501046605558442e-05,
+      "loss": 3.9091,
+      "step": 1526272
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9959750175476074,
+      "eval_runtime": 291.7693,
+      "eval_samples_per_second": 1307.852,
+      "eval_steps_per_second": 40.871,
+      "step": 1526400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.50020801080739e-05,
+      "loss": 3.9076,
+      "step": 1526784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.499369416056338e-05,
+      "loss": 3.9122,
+      "step": 1527296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.498530821305286e-05,
+      "loss": 3.9203,
+      "step": 1527808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.497692226554234e-05,
+      "loss": 3.9095,
+      "step": 1528320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4968536318031817e-05,
+      "loss": 3.9216,
+      "step": 1528832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.49601503705213e-05,
+      "loss": 3.9093,
+      "step": 1529344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.495176442301078e-05,
+      "loss": 3.9119,
+      "step": 1529856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.494337847550026e-05,
+      "loss": 3.9012,
+      "step": 1530368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.493499252798974e-05,
+      "loss": 3.9239,
+      "step": 1530880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.492660658047922e-05,
+      "loss": 3.9161,
+      "step": 1531392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4918220632968697e-05,
+      "loss": 3.9124,
+      "step": 1531904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4909834685458176e-05,
+      "loss": 3.9236,
+      "step": 1532416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4901465116751392e-05,
+      "loss": 3.9016,
+      "step": 1532928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4893079169240872e-05,
+      "loss": 3.9118,
+      "step": 1533440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.488469322173035e-05,
+      "loss": 3.9125,
+      "step": 1533952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.487630727421983e-05,
+      "loss": 3.903,
+      "step": 1534464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.486792132670931e-05,
+      "loss": 3.9054,
+      "step": 1534976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.485953537919879e-05,
+      "loss": 3.9036,
+      "step": 1535488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.485114943168827e-05,
+      "loss": 3.9083,
+      "step": 1536000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.484276348417775e-05,
+      "loss": 3.9245,
+      "step": 1536512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4834377536667232e-05,
+      "loss": 3.9118,
+      "step": 1537024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.482600796796044e-05,
+      "loss": 3.9109,
+      "step": 1537536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.481762202044992e-05,
+      "loss": 3.9088,
+      "step": 1538048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.48092360729394e-05,
+      "loss": 3.9175,
+      "step": 1538560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4800850125428878e-05,
+      "loss": 3.9017,
+      "step": 1539072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4792464177918358e-05,
+      "loss": 3.9038,
+      "step": 1539584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4784094609211573e-05,
+      "loss": 3.9013,
+      "step": 1540096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4775708661701053e-05,
+      "loss": 3.9131,
+      "step": 1540608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.476732271419053e-05,
+      "loss": 3.8982,
+      "step": 1541120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.475893676668001e-05,
+      "loss": 3.9102,
+      "step": 1541632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.475055081916949e-05,
+      "loss": 3.909,
+      "step": 1542144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.474216487165897e-05,
+      "loss": 3.9138,
+      "step": 1542656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.473377892414845e-05,
+      "loss": 3.9064,
+      "step": 1543168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4725392976637933e-05,
+      "loss": 3.9111,
+      "step": 1543680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4717007029127413e-05,
+      "loss": 3.9137,
+      "step": 1544192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4708637460420622e-05,
+      "loss": 3.9126,
+      "step": 1544704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4700251512910102e-05,
+      "loss": 3.8974,
+      "step": 1545216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.469188194420331e-05,
+      "loss": 3.9041,
+      "step": 1545728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4683495996692795e-05,
+      "loss": 3.8958,
+      "step": 1546240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4675110049182275e-05,
+      "loss": 3.9082,
+      "step": 1546752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4666724101671754e-05,
+      "loss": 3.903,
+      "step": 1547264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4658354532964964e-05,
+      "loss": 3.9135,
+      "step": 1547776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4649968585454444e-05,
+      "loss": 3.9048,
+      "step": 1548288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4641582637943923e-05,
+      "loss": 3.9095,
+      "step": 1548800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4633196690433403e-05,
+      "loss": 3.9123,
+      "step": 1549312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4624810742922887e-05,
+      "loss": 3.9076,
+      "step": 1549824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4616424795412363e-05,
+      "loss": 3.9125,
+      "step": 1550336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4608038847901843e-05,
+      "loss": 3.9005,
+      "step": 1550848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4599652900391323e-05,
+      "loss": 3.8934,
+      "step": 1551360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4591299710488265e-05,
+      "loss": 3.9176,
+      "step": 1551872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4582913762977748e-05,
+      "loss": 3.906,
+      "step": 1552384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4574527815467228e-05,
+      "loss": 3.8997,
+      "step": 1552896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4566141867956708e-05,
+      "loss": 3.903,
+      "step": 1553408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4557755920446185e-05,
+      "loss": 3.9086,
+      "step": 1553920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4549369972935665e-05,
+      "loss": 3.8938,
+      "step": 1554432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4540984025425145e-05,
+      "loss": 3.9094,
+      "step": 1554944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4532598077914625e-05,
+      "loss": 3.9039,
+      "step": 1555456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4524212130404105e-05,
+      "loss": 3.9083,
+      "step": 1555968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4515826182893588e-05,
+      "loss": 3.9162,
+      "step": 1556480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4507440235383064e-05,
+      "loss": 3.899,
+      "step": 1556992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4499054287872544e-05,
+      "loss": 3.8959,
+      "step": 1557504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4490684719165757e-05,
+      "loss": 3.9111,
+      "step": 1558016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4482315150458966e-05,
+      "loss": 3.8898,
+      "step": 1558528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.447392920294845e-05,
+      "loss": 3.8953,
+      "step": 1559040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.446554325543793e-05,
+      "loss": 3.9075,
+      "step": 1559552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.445715730792741e-05,
+      "loss": 3.9086,
+      "step": 1560064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.444877136041689e-05,
+      "loss": 3.8973,
+      "step": 1560576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4440385412906366e-05,
+      "loss": 3.8924,
+      "step": 1561088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4431999465395846e-05,
+      "loss": 3.9038,
+      "step": 1561600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4423613517885326e-05,
+      "loss": 3.9114,
+      "step": 1562112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4415243949178538e-05,
+      "loss": 3.9074,
+      "step": 1562624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.440687438047175e-05,
+      "loss": 3.9095,
+      "step": 1563136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.439850481176496e-05,
+      "loss": 3.915,
+      "step": 1563648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.439011886425444e-05,
+      "loss": 3.9103,
+      "step": 1564160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.438173291674392e-05,
+      "loss": 3.9227,
+      "step": 1564672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4373346969233403e-05,
+      "loss": 3.8931,
+      "step": 1565184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4364961021722883e-05,
+      "loss": 3.9121,
+      "step": 1565696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4356575074212363e-05,
+      "loss": 3.9115,
+      "step": 1566208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.434818912670184e-05,
+      "loss": 3.8992,
+      "step": 1566720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.433980317919132e-05,
+      "loss": 3.9145,
+      "step": 1567232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.43314172316808e-05,
+      "loss": 3.9037,
+      "step": 1567744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.432303128417028e-05,
+      "loss": 3.9163,
+      "step": 1568256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.431464533665976e-05,
+      "loss": 3.9174,
+      "step": 1568768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4306259389149243e-05,
+      "loss": 3.893,
+      "step": 1569280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4297889820442452e-05,
+      "loss": 3.9034,
+      "step": 1569792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.428950387293193e-05,
+      "loss": 3.8992,
+      "step": 1570304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.428111792542141e-05,
+      "loss": 3.9107,
+      "step": 1570816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.427273197791089e-05,
+      "loss": 3.9109,
+      "step": 1571328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4264362409204104e-05,
+      "loss": 3.8996,
+      "step": 1571840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4255976461693584e-05,
+      "loss": 3.9045,
+      "step": 1572352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4247590514183064e-05,
+      "loss": 3.9015,
+      "step": 1572864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4239204566672544e-05,
+      "loss": 3.9041,
+      "step": 1573376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.423081861916202e-05,
+      "loss": 3.8991,
+      "step": 1573888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4222465429258966e-05,
+      "loss": 3.9134,
+      "step": 1574400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4214079481748445e-05,
+      "loss": 3.9151,
+      "step": 1574912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4205693534237925e-05,
+      "loss": 3.908,
+      "step": 1575424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4197307586727405e-05,
+      "loss": 3.8998,
+      "step": 1575936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4188921639216885e-05,
+      "loss": 3.9043,
+      "step": 1576448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4180552070510094e-05,
+      "loss": 3.906,
+      "step": 1576960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4172166122999574e-05,
+      "loss": 3.9151,
+      "step": 1577472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4163780175489058e-05,
+      "loss": 3.9057,
+      "step": 1577984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4155394227978538e-05,
+      "loss": 3.9009,
+      "step": 1578496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4147008280468018e-05,
+      "loss": 3.8961,
+      "step": 1579008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4138622332957494e-05,
+      "loss": 3.9079,
+      "step": 1579520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4130252764250707e-05,
+      "loss": 3.8913,
+      "step": 1580032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4121866816740187e-05,
+      "loss": 3.8956,
+      "step": 1580544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4113480869229667e-05,
+      "loss": 3.8925,
+      "step": 1581056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4105094921719147e-05,
+      "loss": 3.9052,
+      "step": 1581568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4096708974208627e-05,
+      "loss": 3.8945,
+      "step": 1582080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4088323026698106e-05,
+      "loss": 3.9165,
+      "step": 1582592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4079937079187586e-05,
+      "loss": 3.8959,
+      "step": 1583104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4071551131677066e-05,
+      "loss": 3.9036,
+      "step": 1583616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.406318156297028e-05,
+      "loss": 3.9015,
+      "step": 1584128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.405479561545976e-05,
+      "loss": 3.8945,
+      "step": 1584640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.404640966794924e-05,
+      "loss": 3.9157,
+      "step": 1585152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.403802372043872e-05,
+      "loss": 3.9114,
+      "step": 1585664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4029654151731928e-05,
+      "loss": 3.9042,
+      "step": 1586176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4021268204221408e-05,
+      "loss": 3.9023,
+      "step": 1586688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4012882256710888e-05,
+      "loss": 3.8972,
+      "step": 1587200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.40045126880041e-05,
+      "loss": 3.9097,
+      "step": 1587712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.399612674049358e-05,
+      "loss": 3.8916,
+      "step": 1588224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.398774079298306e-05,
+      "loss": 3.9152,
+      "step": 1588736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.397935484547254e-05,
+      "loss": 3.9022,
+      "step": 1589248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.397096889796202e-05,
+      "loss": 3.9043,
+      "step": 1589760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3962582950451497e-05,
+      "loss": 3.898,
+      "step": 1590272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.395419700294098e-05,
+      "loss": 3.8999,
+      "step": 1590784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.394581105543046e-05,
+      "loss": 3.8937,
+      "step": 1591296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.393744148672367e-05,
+      "loss": 3.9064,
+      "step": 1591808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.392905553921315e-05,
+      "loss": 3.8968,
+      "step": 1592320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.392066959170263e-05,
+      "loss": 3.9143,
+      "step": 1592832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.391228364419211e-05,
+      "loss": 3.9101,
+      "step": 1593344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.390389769668159e-05,
+      "loss": 3.9067,
+      "step": 1593856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3895511749171072e-05,
+      "loss": 3.8932,
+      "step": 1594368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.388714218046428e-05,
+      "loss": 3.9134,
+      "step": 1594880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.387875623295376e-05,
+      "loss": 3.9092,
+      "step": 1595392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.387037028544324e-05,
+      "loss": 3.9132,
+      "step": 1595904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.386198433793272e-05,
+      "loss": 3.9022,
+      "step": 1596416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.38535983904222e-05,
+      "loss": 3.9024,
+      "step": 1596928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.384521244291168e-05,
+      "loss": 3.9094,
+      "step": 1597440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.383682649540116e-05,
+      "loss": 3.9101,
+      "step": 1597952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.382844054789064e-05,
+      "loss": 3.8972,
+      "step": 1598464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.382007097918385e-05,
+      "loss": 3.9089,
+      "step": 1598976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.381168503167333e-05,
+      "loss": 3.8958,
+      "step": 1599488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.380329908416281e-05,
+      "loss": 3.8992,
+      "step": 1600000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3794929515456026e-05,
+      "loss": 3.9092,
+      "step": 1600512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3786543567945502e-05,
+      "loss": 3.9063,
+      "step": 1601024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3778157620434982e-05,
+      "loss": 3.9024,
+      "step": 1601536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3769771672924462e-05,
+      "loss": 3.8983,
+      "step": 1602048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3761385725413942e-05,
+      "loss": 3.9015,
+      "step": 1602560
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9934589862823486,
+      "eval_runtime": 289.6382,
+      "eval_samples_per_second": 1317.475,
+      "eval_steps_per_second": 41.172,
+      "step": 1602720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.375301615670715e-05,
+      "loss": 3.9033,
+      "step": 1603072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3744646588000367e-05,
+      "loss": 3.9023,
+      "step": 1603584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3736277019293576e-05,
+      "loss": 3.9144,
+      "step": 1604096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3727891071783056e-05,
+      "loss": 3.9057,
+      "step": 1604608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3719505124272536e-05,
+      "loss": 3.9089,
+      "step": 1605120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3711119176762016e-05,
+      "loss": 3.9048,
+      "step": 1605632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3702733229251496e-05,
+      "loss": 3.9017,
+      "step": 1606144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3694347281740976e-05,
+      "loss": 3.8968,
+      "step": 1606656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3685961334230456e-05,
+      "loss": 3.9113,
+      "step": 1607168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3677575386719936e-05,
+      "loss": 3.9131,
+      "step": 1607680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3669189439209416e-05,
+      "loss": 3.9072,
+      "step": 1608192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3660803491698896e-05,
+      "loss": 3.9163,
+      "step": 1608704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3652433922992105e-05,
+      "loss": 3.8943,
+      "step": 1609216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.364404797548159e-05,
+      "loss": 3.9038,
+      "step": 1609728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.363566202797107e-05,
+      "loss": 3.9114,
+      "step": 1610240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.362727608046055e-05,
+      "loss": 3.8923,
+      "step": 1610752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3618890132950028e-05,
+      "loss": 3.8989,
+      "step": 1611264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3610520564243237e-05,
+      "loss": 3.8962,
+      "step": 1611776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3602134616732717e-05,
+      "loss": 3.9006,
+      "step": 1612288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3593748669222197e-05,
+      "loss": 3.915,
+      "step": 1612800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.358536272171168e-05,
+      "loss": 3.9058,
+      "step": 1613312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3576976774201157e-05,
+      "loss": 3.9065,
+      "step": 1613824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.356860720549437e-05,
+      "loss": 3.8968,
+      "step": 1614336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.356022125798385e-05,
+      "loss": 3.912,
+      "step": 1614848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.355183531047333e-05,
+      "loss": 3.8932,
+      "step": 1615360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3543449362962806e-05,
+      "loss": 3.897,
+      "step": 1615872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.353506341545229e-05,
+      "loss": 3.8982,
+      "step": 1616384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.352667746794177e-05,
+      "loss": 3.9023,
+      "step": 1616896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.351829152043125e-05,
+      "loss": 3.8924,
+      "step": 1617408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.350990557292073e-05,
+      "loss": 3.9023,
+      "step": 1617920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.350153600421394e-05,
+      "loss": 3.9061,
+      "step": 1618432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.349315005670342e-05,
+      "loss": 3.9065,
+      "step": 1618944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.34847641091929e-05,
+      "loss": 3.9022,
+      "step": 1619456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3476378161682382e-05,
+      "loss": 3.901,
+      "step": 1619968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.346800859297559e-05,
+      "loss": 3.9079,
+      "step": 1620480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3459639024268803e-05,
+      "loss": 3.9003,
+      "step": 1620992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.345125307675828e-05,
+      "loss": 3.8942,
+      "step": 1621504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.344286712924776e-05,
+      "loss": 3.8976,
+      "step": 1622016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3434481181737243e-05,
+      "loss": 3.8903,
+      "step": 1622528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3426095234226723e-05,
+      "loss": 3.8999,
+      "step": 1623040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3417725665519932e-05,
+      "loss": 3.8966,
+      "step": 1623552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3409339718009412e-05,
+      "loss": 3.9059,
+      "step": 1624064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3400953770498892e-05,
+      "loss": 3.8979,
+      "step": 1624576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3392567822988372e-05,
+      "loss": 3.8993,
+      "step": 1625088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3384181875477852e-05,
+      "loss": 3.9107,
+      "step": 1625600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3375795927967335e-05,
+      "loss": 3.8993,
+      "step": 1626112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3367409980456812e-05,
+      "loss": 3.9068,
+      "step": 1626624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3359024032946292e-05,
+      "loss": 3.8916,
+      "step": 1627136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3350654464239504e-05,
+      "loss": 3.8878,
+      "step": 1627648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.334226851672898e-05,
+      "loss": 3.9148,
+      "step": 1628160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3333898948022197e-05,
+      "loss": 3.8974,
+      "step": 1628672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3325513000511677e-05,
+      "loss": 3.8945,
+      "step": 1629184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3317127053001157e-05,
+      "loss": 3.8952,
+      "step": 1629696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3308741105490633e-05,
+      "loss": 3.9023,
+      "step": 1630208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3300355157980113e-05,
+      "loss": 3.8847,
+      "step": 1630720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3291969210469593e-05,
+      "loss": 3.9005,
+      "step": 1631232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3283599641762806e-05,
+      "loss": 3.899,
+      "step": 1631744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3275213694252286e-05,
+      "loss": 3.8969,
+      "step": 1632256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3266827746741766e-05,
+      "loss": 3.9135,
+      "step": 1632768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3258441799231246e-05,
+      "loss": 3.8904,
+      "step": 1633280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3250055851720726e-05,
+      "loss": 3.8916,
+      "step": 1633792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3241686283013935e-05,
+      "loss": 3.9043,
+      "step": 1634304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3233300335503415e-05,
+      "loss": 3.8849,
+      "step": 1634816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3224914387992898e-05,
+      "loss": 3.8883,
+      "step": 1635328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3216528440482378e-05,
+      "loss": 3.8966,
+      "step": 1635840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3208142492971858e-05,
+      "loss": 3.9036,
+      "step": 1636352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3199756545461338e-05,
+      "loss": 3.8913,
+      "step": 1636864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3191370597950814e-05,
+      "loss": 3.889,
+      "step": 1637376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3182984650440294e-05,
+      "loss": 3.8921,
+      "step": 1637888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3174615081733507e-05,
+      "loss": 3.902,
+      "step": 1638400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.316622913422299e-05,
+      "loss": 3.9002,
+      "step": 1638912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3157843186712467e-05,
+      "loss": 3.905,
+      "step": 1639424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3149457239201947e-05,
+      "loss": 3.9041,
+      "step": 1639936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3141104049298888e-05,
+      "loss": 3.9054,
+      "step": 1640448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3132734480592104e-05,
+      "loss": 3.9167,
+      "step": 1640960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.312434853308158e-05,
+      "loss": 3.8859,
+      "step": 1641472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.311596258557106e-05,
+      "loss": 3.9085,
+      "step": 1641984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.310757663806054e-05,
+      "loss": 3.9064,
+      "step": 1642496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.309919069055002e-05,
+      "loss": 3.8872,
+      "step": 1643008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.30908047430395e-05,
+      "loss": 3.9115,
+      "step": 1643520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.308241879552898e-05,
+      "loss": 3.896,
+      "step": 1644032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.307403284801846e-05,
+      "loss": 3.911,
+      "step": 1644544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.306564690050794e-05,
+      "loss": 3.9112,
+      "step": 1645056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.305726095299742e-05,
+      "loss": 3.8883,
+      "step": 1645568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.30488750054869e-05,
+      "loss": 3.891,
+      "step": 1646080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.304048905797638e-05,
+      "loss": 3.8928,
+      "step": 1646592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.303211948926959e-05,
+      "loss": 3.9058,
+      "step": 1647104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3023733541759073e-05,
+      "loss": 3.903,
+      "step": 1647616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3015347594248553e-05,
+      "loss": 3.8967,
+      "step": 1648128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3006961646738033e-05,
+      "loss": 3.8964,
+      "step": 1648640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2998592078031242e-05,
+      "loss": 3.8957,
+      "step": 1649152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.299020613052072e-05,
+      "loss": 3.897,
+      "step": 1649664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.29818201830102e-05,
+      "loss": 3.8909,
+      "step": 1650176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.297343423549968e-05,
+      "loss": 3.9096,
+      "step": 1650688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.296504828798916e-05,
+      "loss": 3.9067,
+      "step": 1651200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2956662340478645e-05,
+      "loss": 3.9046,
+      "step": 1651712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.294827639296812e-05,
+      "loss": 3.889,
+      "step": 1652224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.29398904454576e-05,
+      "loss": 3.9005,
+      "step": 1652736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2931520876750814e-05,
+      "loss": 3.8956,
+      "step": 1653248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.292313492924029e-05,
+      "loss": 3.9039,
+      "step": 1653760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2914748981729774e-05,
+      "loss": 3.9022,
+      "step": 1654272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2906363034219254e-05,
+      "loss": 3.8959,
+      "step": 1654784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2897977086708734e-05,
+      "loss": 3.8904,
+      "step": 1655296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2889607518001943e-05,
+      "loss": 3.9008,
+      "step": 1655808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2881221570491423e-05,
+      "loss": 3.8897,
+      "step": 1656320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2872835622980903e-05,
+      "loss": 3.8861,
+      "step": 1656832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2864449675470383e-05,
+      "loss": 3.8813,
+      "step": 1657344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2856080106763595e-05,
+      "loss": 3.8987,
+      "step": 1657856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2847710538056808e-05,
+      "loss": 3.8939,
+      "step": 1658368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2839324590546288e-05,
+      "loss": 3.9057,
+      "step": 1658880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2830938643035764e-05,
+      "loss": 3.8918,
+      "step": 1659392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2822552695525244e-05,
+      "loss": 3.8956,
+      "step": 1659904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2814166748014727e-05,
+      "loss": 3.8974,
+      "step": 1660416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2805780800504207e-05,
+      "loss": 3.8868,
+      "step": 1660928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2797394852993687e-05,
+      "loss": 3.9081,
+      "step": 1661440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2789008905483167e-05,
+      "loss": 3.9059,
+      "step": 1661952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2780639336776376e-05,
+      "loss": 3.8961,
+      "step": 1662464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2772253389265856e-05,
+      "loss": 3.8964,
+      "step": 1662976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.276388382055907e-05,
+      "loss": 3.8933,
+      "step": 1663488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.275549787304855e-05,
+      "loss": 3.9031,
+      "step": 1664000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.274711192553803e-05,
+      "loss": 3.8818,
+      "step": 1664512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.273872597802751e-05,
+      "loss": 3.9106,
+      "step": 1665024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.273034003051699e-05,
+      "loss": 3.892,
+      "step": 1665536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.272195408300647e-05,
+      "loss": 3.8952,
+      "step": 1666048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2713568135495945e-05,
+      "loss": 3.8949,
+      "step": 1666560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.270518218798543e-05,
+      "loss": 3.8904,
+      "step": 1667072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.269679624047491e-05,
+      "loss": 3.8932,
+      "step": 1667584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.268841029296439e-05,
+      "loss": 3.8969,
+      "step": 1668096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2680040724257598e-05,
+      "loss": 3.8923,
+      "step": 1668608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2671654776747078e-05,
+      "loss": 3.9053,
+      "step": 1669120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2663268829236557e-05,
+      "loss": 3.8992,
+      "step": 1669632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2654882881726037e-05,
+      "loss": 3.904,
+      "step": 1670144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.264651331301925e-05,
+      "loss": 3.884,
+      "step": 1670656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.263812736550873e-05,
+      "loss": 3.9078,
+      "step": 1671168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.262974141799821e-05,
+      "loss": 3.8998,
+      "step": 1671680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.262135547048769e-05,
+      "loss": 3.9042,
+      "step": 1672192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.261296952297717e-05,
+      "loss": 3.897,
+      "step": 1672704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2604599954270382e-05,
+      "loss": 3.8933,
+      "step": 1673216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2596214006759862e-05,
+      "loss": 3.9062,
+      "step": 1673728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2587828059249342e-05,
+      "loss": 3.905,
+      "step": 1674240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2579442111738822e-05,
+      "loss": 3.8888,
+      "step": 1674752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2571056164228302e-05,
+      "loss": 3.8973,
+      "step": 1675264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.256267021671778e-05,
+      "loss": 3.8949,
+      "step": 1675776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.255428426920726e-05,
+      "loss": 3.8914,
+      "step": 1676288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.254589832169674e-05,
+      "loss": 3.9015,
+      "step": 1676800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.253752875298995e-05,
+      "loss": 3.8954,
+      "step": 1677312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2529159184283164e-05,
+      "loss": 3.8995,
+      "step": 1677824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2520773236772643e-05,
+      "loss": 3.8871,
+      "step": 1678336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2512387289262123e-05,
+      "loss": 3.8989,
+      "step": 1678848
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9908547401428223,
+      "eval_runtime": 290.3419,
+      "eval_samples_per_second": 1314.282,
+      "eval_steps_per_second": 41.072,
+      "step": 1679040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.25040013417516e-05,
+      "loss": 3.9006,
+      "step": 1679360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2495615394241083e-05,
+      "loss": 3.8924,
+      "step": 1679872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2487229446730563e-05,
+      "loss": 3.9066,
+      "step": 1680384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2478843499220043e-05,
+      "loss": 3.902,
+      "step": 1680896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2470457551709523e-05,
+      "loss": 3.9012,
+      "step": 1681408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2462071604199003e-05,
+      "loss": 3.899,
+      "step": 1681920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2453685656688483e-05,
+      "loss": 3.8898,
+      "step": 1682432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.244529970917796e-05,
+      "loss": 3.8942,
+      "step": 1682944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.243691376166744e-05,
+      "loss": 3.9016,
+      "step": 1683456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2428527814156923e-05,
+      "loss": 3.9077,
+      "step": 1683968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2420141866646403e-05,
+      "loss": 3.8996,
+      "step": 1684480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2411755919135883e-05,
+      "loss": 3.9101,
+      "step": 1684992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2403386350429092e-05,
+      "loss": 3.8921,
+      "step": 1685504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2395000402918572e-05,
+      "loss": 3.8929,
+      "step": 1686016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2386614455408052e-05,
+      "loss": 3.9061,
+      "step": 1686528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2378228507897532e-05,
+      "loss": 3.8859,
+      "step": 1687040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2369842560387015e-05,
+      "loss": 3.8932,
+      "step": 1687552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2361456612876492e-05,
+      "loss": 3.8889,
+      "step": 1688064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2353070665365972e-05,
+      "loss": 3.8949,
+      "step": 1688576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.234468471785545e-05,
+      "loss": 3.9045,
+      "step": 1689088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.233629877034493e-05,
+      "loss": 3.9026,
+      "step": 1689600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.232791282283441e-05,
+      "loss": 3.9017,
+      "step": 1690112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2319543254127624e-05,
+      "loss": 3.8886,
+      "step": 1690624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2311157306617104e-05,
+      "loss": 3.9036,
+      "step": 1691136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2302771359106584e-05,
+      "loss": 3.8908,
+      "step": 1691648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2294385411596064e-05,
+      "loss": 3.8909,
+      "step": 1692160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2285999464085544e-05,
+      "loss": 3.8873,
+      "step": 1692672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.227761351657502e-05,
+      "loss": 3.894,
+      "step": 1693184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.22692275690645e-05,
+      "loss": 3.8948,
+      "step": 1693696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2260858000357716e-05,
+      "loss": 3.8894,
+      "step": 1694208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2252472052847196e-05,
+      "loss": 3.8995,
+      "step": 1694720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2244086105336673e-05,
+      "loss": 3.9002,
+      "step": 1695232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2235700157826153e-05,
+      "loss": 3.895,
+      "step": 1695744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2227314210315633e-05,
+      "loss": 3.9005,
+      "step": 1696256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2218928262805113e-05,
+      "loss": 3.8986,
+      "step": 1696768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2210542315294593e-05,
+      "loss": 3.896,
+      "step": 1697280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2202172746587805e-05,
+      "loss": 3.8911,
+      "step": 1697792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2193786799077285e-05,
+      "loss": 3.8843,
+      "step": 1698304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2185400851566765e-05,
+      "loss": 3.8834,
+      "step": 1698816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2177014904056245e-05,
+      "loss": 3.893,
+      "step": 1699328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2168628956545725e-05,
+      "loss": 3.8926,
+      "step": 1699840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2160259387838937e-05,
+      "loss": 3.8964,
+      "step": 1700352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2151873440328417e-05,
+      "loss": 3.891,
+      "step": 1700864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2143487492817897e-05,
+      "loss": 3.8918,
+      "step": 1701376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2135101545307377e-05,
+      "loss": 3.9042,
+      "step": 1701888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2126715597796854e-05,
+      "loss": 3.8943,
+      "step": 1702400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2118346029090066e-05,
+      "loss": 3.9025,
+      "step": 1702912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2109960081579546e-05,
+      "loss": 3.8862,
+      "step": 1703424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2101574134069026e-05,
+      "loss": 3.8758,
+      "step": 1703936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2093188186558506e-05,
+      "loss": 3.9088,
+      "step": 1704448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.208481861785172e-05,
+      "loss": 3.8943,
+      "step": 1704960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2076449049144928e-05,
+      "loss": 3.8852,
+      "step": 1705472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2068063101634408e-05,
+      "loss": 3.8911,
+      "step": 1705984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.205967715412389e-05,
+      "loss": 3.8931,
+      "step": 1706496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.205129120661337e-05,
+      "loss": 3.8818,
+      "step": 1707008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.204290525910285e-05,
+      "loss": 3.893,
+      "step": 1707520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2034519311592328e-05,
+      "loss": 3.8927,
+      "step": 1708032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2026133364081808e-05,
+      "loss": 3.8936,
+      "step": 1708544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2017747416571287e-05,
+      "loss": 3.9044,
+      "step": 1709056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.20093778478645e-05,
+      "loss": 3.8831,
+      "step": 1709568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.200099190035398e-05,
+      "loss": 3.8851,
+      "step": 1710080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.199260595284346e-05,
+      "loss": 3.9008,
+      "step": 1710592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.198422000533294e-05,
+      "loss": 3.8788,
+      "step": 1711104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.197583405782242e-05,
+      "loss": 3.8791,
+      "step": 1711616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.19674481103119e-05,
+      "loss": 3.891,
+      "step": 1712128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.195906216280138e-05,
+      "loss": 3.8964,
+      "step": 1712640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1950676215290856e-05,
+      "loss": 3.8837,
+      "step": 1713152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1942306646584072e-05,
+      "loss": 3.8848,
+      "step": 1713664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.193393707787728e-05,
+      "loss": 3.8821,
+      "step": 1714176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.192555113036676e-05,
+      "loss": 3.9016,
+      "step": 1714688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.191716518285624e-05,
+      "loss": 3.8928,
+      "step": 1715200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.190877923534572e-05,
+      "loss": 3.8977,
+      "step": 1715712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.19003932878352e-05,
+      "loss": 3.8969,
+      "step": 1716224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1892023719128414e-05,
+      "loss": 3.8967,
+      "step": 1716736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1883637771617893e-05,
+      "loss": 3.9105,
+      "step": 1717248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1875268202911103e-05,
+      "loss": 3.8819,
+      "step": 1717760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1866882255400583e-05,
+      "loss": 3.9012,
+      "step": 1718272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1858496307890063e-05,
+      "loss": 3.9007,
+      "step": 1718784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1850110360379546e-05,
+      "loss": 3.8833,
+      "step": 1719296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1841724412869026e-05,
+      "loss": 3.8997,
+      "step": 1719808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1833338465358506e-05,
+      "loss": 3.8914,
+      "step": 1720320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1824952517847982e-05,
+      "loss": 3.9028,
+      "step": 1720832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1816566570337462e-05,
+      "loss": 3.9053,
+      "step": 1721344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1808180622826942e-05,
+      "loss": 3.8881,
+      "step": 1721856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1799794675316422e-05,
+      "loss": 3.8819,
+      "step": 1722368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1791408727805902e-05,
+      "loss": 3.8867,
+      "step": 1722880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1783022780295385e-05,
+      "loss": 3.8997,
+      "step": 1723392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1774653211588595e-05,
+      "loss": 3.8924,
+      "step": 1723904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1766267264078075e-05,
+      "loss": 3.8909,
+      "step": 1724416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1757897695371284e-05,
+      "loss": 3.8912,
+      "step": 1724928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.17495281266645e-05,
+      "loss": 3.8872,
+      "step": 1725440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.174114217915398e-05,
+      "loss": 3.8923,
+      "step": 1725952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1732756231643456e-05,
+      "loss": 3.8861,
+      "step": 1726464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1724370284132936e-05,
+      "loss": 3.9024,
+      "step": 1726976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1715984336622416e-05,
+      "loss": 3.9,
+      "step": 1727488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1707598389111896e-05,
+      "loss": 3.8996,
+      "step": 1728000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1699212441601376e-05,
+      "loss": 3.8831,
+      "step": 1728512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.169084287289459e-05,
+      "loss": 3.8923,
+      "step": 1729024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1682456925384068e-05,
+      "loss": 3.8948,
+      "step": 1729536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1674070977873548e-05,
+      "loss": 3.8942,
+      "step": 1730048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1665685030363028e-05,
+      "loss": 3.8956,
+      "step": 1730560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1657299082852508e-05,
+      "loss": 3.8859,
+      "step": 1731072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1648913135341985e-05,
+      "loss": 3.8886,
+      "step": 1731584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1640527187831465e-05,
+      "loss": 3.8908,
+      "step": 1732096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1632141240320948e-05,
+      "loss": 3.8821,
+      "step": 1732608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.162377167161416e-05,
+      "loss": 3.8897,
+      "step": 1733120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1615385724103637e-05,
+      "loss": 3.8697,
+      "step": 1733632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1606999776593117e-05,
+      "loss": 3.8923,
+      "step": 1734144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1598613829082597e-05,
+      "loss": 3.8896,
+      "step": 1734656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1590227881572077e-05,
+      "loss": 3.8953,
+      "step": 1735168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1581841934061557e-05,
+      "loss": 3.8844,
+      "step": 1735680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.157345598655104e-05,
+      "loss": 3.8908,
+      "step": 1736192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1565070039040517e-05,
+      "loss": 3.8912,
+      "step": 1736704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.155670047033373e-05,
+      "loss": 3.8784,
+      "step": 1737216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.154831452282321e-05,
+      "loss": 3.9042,
+      "step": 1737728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.153992857531269e-05,
+      "loss": 3.8936,
+      "step": 1738240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1531559006605902e-05,
+      "loss": 3.8907,
+      "step": 1738752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.152317305909538e-05,
+      "loss": 3.8931,
+      "step": 1739264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.151478711158486e-05,
+      "loss": 3.879,
+      "step": 1739776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1506401164074338e-05,
+      "loss": 3.9011,
+      "step": 1740288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1498015216563818e-05,
+      "loss": 3.8817,
+      "step": 1740800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1489629269053298e-05,
+      "loss": 3.901,
+      "step": 1741312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1481243321542778e-05,
+      "loss": 3.8839,
+      "step": 1741824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.147287375283599e-05,
+      "loss": 3.8914,
+      "step": 1742336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.146448780532547e-05,
+      "loss": 3.8855,
+      "step": 1742848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.145610185781495e-05,
+      "loss": 3.886,
+      "step": 1743360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.144771591030443e-05,
+      "loss": 3.8873,
+      "step": 1743872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.143932996279391e-05,
+      "loss": 3.8885,
+      "step": 1744384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1430960394087123e-05,
+      "loss": 3.8886,
+      "step": 1744896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1422574446576603e-05,
+      "loss": 3.8979,
+      "step": 1745408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1414188499066083e-05,
+      "loss": 3.8909,
+      "step": 1745920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1405802551555563e-05,
+      "loss": 3.9003,
+      "step": 1746432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1397416604045043e-05,
+      "loss": 3.8778,
+      "step": 1746944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.138903065653452e-05,
+      "loss": 3.9022,
+      "step": 1747456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1380644709024e-05,
+      "loss": 3.896,
+      "step": 1747968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.137225876151348e-05,
+      "loss": 3.8963,
+      "step": 1748480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1363905571610424e-05,
+      "loss": 3.8912,
+      "step": 1748992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1355519624099904e-05,
+      "loss": 3.8915,
+      "step": 1749504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1347133676589384e-05,
+      "loss": 3.8944,
+      "step": 1750016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1338747729078864e-05,
+      "loss": 3.9007,
+      "step": 1750528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1330361781568344e-05,
+      "loss": 3.8819,
+      "step": 1751040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1321975834057824e-05,
+      "loss": 3.8931,
+      "step": 1751552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1313589886547304e-05,
+      "loss": 3.8872,
+      "step": 1752064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1305203939036784e-05,
+      "loss": 3.887,
+      "step": 1752576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1296817991526264e-05,
+      "loss": 3.8917,
+      "step": 1753088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1288448422819473e-05,
+      "loss": 3.8907,
+      "step": 1753600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1280078854112685e-05,
+      "loss": 3.8924,
+      "step": 1754112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1271692906602165e-05,
+      "loss": 3.8815,
+      "step": 1754624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1263306959091645e-05,
+      "loss": 3.8955,
+      "step": 1755136
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.988119125366211,
+      "eval_runtime": 297.5358,
+      "eval_samples_per_second": 1282.505,
+      "eval_steps_per_second": 40.079,
+      "step": 1755360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1254921011581125e-05,
+      "loss": 3.8905,
+      "step": 1755648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1246535064070605e-05,
+      "loss": 3.8874,
+      "step": 1756160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1238165495363818e-05,
+      "loss": 3.8974,
+      "step": 1756672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1229779547853294e-05,
+      "loss": 3.8934,
+      "step": 1757184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1221393600342778e-05,
+      "loss": 3.8942,
+      "step": 1757696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1213007652832258e-05,
+      "loss": 3.8947,
+      "step": 1758208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1204621705321738e-05,
+      "loss": 3.8856,
+      "step": 1758720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1196235757811217e-05,
+      "loss": 3.8846,
+      "step": 1759232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1187849810300697e-05,
+      "loss": 3.8973,
+      "step": 1759744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1179463862790174e-05,
+      "loss": 3.9001,
+      "step": 1760256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1171094294083386e-05,
+      "loss": 3.8937,
+      "step": 1760768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.116270834657287e-05,
+      "loss": 3.9022,
+      "step": 1761280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.115435515666981e-05,
+      "loss": 3.8897,
+      "step": 1761792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.114596920915929e-05,
+      "loss": 3.8856,
+      "step": 1762304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1137583261648768e-05,
+      "loss": 3.8953,
+      "step": 1762816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1129197314138248e-05,
+      "loss": 3.8769,
+      "step": 1763328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.112081136662773e-05,
+      "loss": 3.8892,
+      "step": 1763840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.111242541911721e-05,
+      "loss": 3.8825,
+      "step": 1764352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.110403947160669e-05,
+      "loss": 3.8902,
+      "step": 1764864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.109565352409617e-05,
+      "loss": 3.8979,
+      "step": 1765376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1087267576585648e-05,
+      "loss": 3.9002,
+      "step": 1765888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1078881629075128e-05,
+      "loss": 3.8926,
+      "step": 1766400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1070495681564608e-05,
+      "loss": 3.8824,
+      "step": 1766912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1062109734054088e-05,
+      "loss": 3.8987,
+      "step": 1767424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1053756544151033e-05,
+      "loss": 3.881,
+      "step": 1767936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1045370596640513e-05,
+      "loss": 3.8889,
+      "step": 1768448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.103700102793372e-05,
+      "loss": 3.8801,
+      "step": 1768960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.10286150804232e-05,
+      "loss": 3.8918,
+      "step": 1769472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1020229132912685e-05,
+      "loss": 3.8839,
+      "step": 1769984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1011843185402165e-05,
+      "loss": 3.8874,
+      "step": 1770496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1003457237891645e-05,
+      "loss": 3.8893,
+      "step": 1771008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.099507129038112e-05,
+      "loss": 3.8964,
+      "step": 1771520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0986701721674334e-05,
+      "loss": 3.8913,
+      "step": 1772032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0978315774163814e-05,
+      "loss": 3.8875,
+      "step": 1772544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0969929826653294e-05,
+      "loss": 3.8923,
+      "step": 1773056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0961543879142774e-05,
+      "loss": 3.889,
+      "step": 1773568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0953157931632254e-05,
+      "loss": 3.8882,
+      "step": 1774080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0944771984121734e-05,
+      "loss": 3.8748,
+      "step": 1774592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0936386036611214e-05,
+      "loss": 3.8786,
+      "step": 1775104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0928000089100694e-05,
+      "loss": 3.8829,
+      "step": 1775616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0919614141590174e-05,
+      "loss": 3.8903,
+      "step": 1776128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.091122819407965e-05,
+      "loss": 3.89,
+      "step": 1776640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0902842246569133e-05,
+      "loss": 3.8862,
+      "step": 1777152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0894472677862346e-05,
+      "loss": 3.8858,
+      "step": 1777664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0886086730351826e-05,
+      "loss": 3.8982,
+      "step": 1778176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0877700782841302e-05,
+      "loss": 3.8831,
+      "step": 1778688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0869314835330782e-05,
+      "loss": 3.9029,
+      "step": 1779200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0860928887820262e-05,
+      "loss": 3.8783,
+      "step": 1779712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0852542940309742e-05,
+      "loss": 3.8659,
+      "step": 1780224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0844156992799226e-05,
+      "loss": 3.908,
+      "step": 1780736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0835771045288706e-05,
+      "loss": 3.885,
+      "step": 1781248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0827385097778186e-05,
+      "loss": 3.8848,
+      "step": 1781760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0819015529071395e-05,
+      "loss": 3.8851,
+      "step": 1782272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0810645960364604e-05,
+      "loss": 3.8852,
+      "step": 1782784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0802260012854087e-05,
+      "loss": 3.8773,
+      "step": 1783296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0793874065343567e-05,
+      "loss": 3.8829,
+      "step": 1783808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0785488117833047e-05,
+      "loss": 3.8869,
+      "step": 1784320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0777118549126256e-05,
+      "loss": 3.888,
+      "step": 1784832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0768732601615736e-05,
+      "loss": 3.8997,
+      "step": 1785344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0760346654105216e-05,
+      "loss": 3.8782,
+      "step": 1785856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0751960706594696e-05,
+      "loss": 3.8763,
+      "step": 1786368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.074357475908418e-05,
+      "loss": 3.8968,
+      "step": 1786880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.073518881157366e-05,
+      "loss": 3.8701,
+      "step": 1787392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0726802864063136e-05,
+      "loss": 3.8719,
+      "step": 1787904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0718416916552616e-05,
+      "loss": 3.8855,
+      "step": 1788416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0710047347845828e-05,
+      "loss": 3.8888,
+      "step": 1788928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0701661400335305e-05,
+      "loss": 3.8744,
+      "step": 1789440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0693275452824788e-05,
+      "loss": 3.8849,
+      "step": 1789952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0684905884118e-05,
+      "loss": 3.8753,
+      "step": 1790464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.067651993660748e-05,
+      "loss": 3.8912,
+      "step": 1790976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0668133989096957e-05,
+      "loss": 3.8922,
+      "step": 1791488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0659748041586437e-05,
+      "loss": 3.8866,
+      "step": 1792000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0651362094075917e-05,
+      "loss": 3.8912,
+      "step": 1792512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0643008904172862e-05,
+      "loss": 3.8933,
+      "step": 1793024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0634622956662342e-05,
+      "loss": 3.9057,
+      "step": 1793536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0626237009151822e-05,
+      "loss": 3.8722,
+      "step": 1794048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0617851061641302e-05,
+      "loss": 3.9,
+      "step": 1794560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.060946511413078e-05,
+      "loss": 3.8887,
+      "step": 1795072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.060107916662026e-05,
+      "loss": 3.8806,
+      "step": 1795584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0592693219109742e-05,
+      "loss": 3.8909,
+      "step": 1796096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0584307271599222e-05,
+      "loss": 3.8881,
+      "step": 1796608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.057593770289243e-05,
+      "loss": 3.8949,
+      "step": 1797120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.056755175538191e-05,
+      "loss": 3.9018,
+      "step": 1797632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.055916580787139e-05,
+      "loss": 3.8767,
+      "step": 1798144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.055077986036087e-05,
+      "loss": 3.8741,
+      "step": 1798656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.054239391285035e-05,
+      "loss": 3.8832,
+      "step": 1799168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0534024344143563e-05,
+      "loss": 3.8943,
+      "step": 1799680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0525638396633043e-05,
+      "loss": 3.8877,
+      "step": 1800192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0517252449122523e-05,
+      "loss": 3.8844,
+      "step": 1800704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0508866501612003e-05,
+      "loss": 3.8804,
+      "step": 1801216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0500480554101483e-05,
+      "loss": 3.8829,
+      "step": 1801728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0492094606590963e-05,
+      "loss": 3.8874,
+      "step": 1802240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0483725037884175e-05,
+      "loss": 3.8825,
+      "step": 1802752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0475339090373655e-05,
+      "loss": 3.8947,
+      "step": 1803264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0466953142863135e-05,
+      "loss": 3.8942,
+      "step": 1803776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0458567195352612e-05,
+      "loss": 3.8904,
+      "step": 1804288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0450181247842092e-05,
+      "loss": 3.8804,
+      "step": 1804800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0441811679135304e-05,
+      "loss": 3.8901,
+      "step": 1805312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0433425731624788e-05,
+      "loss": 3.8827,
+      "step": 1805824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0425039784114264e-05,
+      "loss": 3.889,
+      "step": 1806336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0416653836603744e-05,
+      "loss": 3.8921,
+      "step": 1806848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0408267889093224e-05,
+      "loss": 3.8833,
+      "step": 1807360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0399881941582704e-05,
+      "loss": 3.8784,
+      "step": 1807872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0391512372875913e-05,
+      "loss": 3.88,
+      "step": 1808384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0383126425365397e-05,
+      "loss": 3.8811,
+      "step": 1808896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0374740477854877e-05,
+      "loss": 3.8815,
+      "step": 1809408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0366354530344357e-05,
+      "loss": 3.8665,
+      "step": 1809920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0357968582833836e-05,
+      "loss": 3.8858,
+      "step": 1810432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0349582635323316e-05,
+      "loss": 3.886,
+      "step": 1810944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0341196687812793e-05,
+      "loss": 3.8849,
+      "step": 1811456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0332827119106006e-05,
+      "loss": 3.8777,
+      "step": 1811968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.032444117159549e-05,
+      "loss": 3.894,
+      "step": 1812480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0316055224084965e-05,
+      "loss": 3.8802,
+      "step": 1812992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0307669276574445e-05,
+      "loss": 3.8745,
+      "step": 1813504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0299299707867658e-05,
+      "loss": 3.895,
+      "step": 1814016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0290913760357138e-05,
+      "loss": 3.8892,
+      "step": 1814528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.028254419165035e-05,
+      "loss": 3.8847,
+      "step": 1815040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.027415824413983e-05,
+      "loss": 3.8907,
+      "step": 1815552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.026577229662931e-05,
+      "loss": 3.8728,
+      "step": 1816064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.025738634911879e-05,
+      "loss": 3.8936,
+      "step": 1816576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0249000401608267e-05,
+      "loss": 3.877,
+      "step": 1817088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0240614454097747e-05,
+      "loss": 3.891,
+      "step": 1817600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.023224488539096e-05,
+      "loss": 3.8791,
+      "step": 1818112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.022385893788044e-05,
+      "loss": 3.886,
+      "step": 1818624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.021547299036992e-05,
+      "loss": 3.8803,
+      "step": 1819136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.02070870428594e-05,
+      "loss": 3.8803,
+      "step": 1819648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.019871747415261e-05,
+      "loss": 3.8782,
+      "step": 1820160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0190331526642088e-05,
+      "loss": 3.8825,
+      "step": 1820672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0181961957935304e-05,
+      "loss": 3.8846,
+      "step": 1821184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0173576010424784e-05,
+      "loss": 3.8938,
+      "step": 1821696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0165190062914264e-05,
+      "loss": 3.8798,
+      "step": 1822208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.015680411540374e-05,
+      "loss": 3.8976,
+      "step": 1822720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.014841816789322e-05,
+      "loss": 3.8704,
+      "step": 1823232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.01400322203827e-05,
+      "loss": 3.8973,
+      "step": 1823744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.013164627287218e-05,
+      "loss": 3.8915,
+      "step": 1824256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.012326032536166e-05,
+      "loss": 3.8902,
+      "step": 1824768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0114874377851144e-05,
+      "loss": 3.8842,
+      "step": 1825280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0106504809144353e-05,
+      "loss": 3.8866,
+      "step": 1825792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0098118861633833e-05,
+      "loss": 3.8897,
+      "step": 1826304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0089732914123313e-05,
+      "loss": 3.8964,
+      "step": 1826816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0081363345416525e-05,
+      "loss": 3.8758,
+      "step": 1827328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0072977397906005e-05,
+      "loss": 3.8867,
+      "step": 1827840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0064591450395485e-05,
+      "loss": 3.8805,
+      "step": 1828352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0056205502884965e-05,
+      "loss": 3.8859,
+      "step": 1828864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0047819555374445e-05,
+      "loss": 3.8796,
+      "step": 1829376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.003943360786392e-05,
+      "loss": 3.8852,
+      "step": 1829888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.00310476603534e-05,
+      "loss": 3.8883,
+      "step": 1830400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.002266171284288e-05,
+      "loss": 3.8736,
+      "step": 1830912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0014275765332365e-05,
+      "loss": 3.8936,
+      "step": 1831424
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9861526489257812,
+      "eval_runtime": 306.5317,
+      "eval_samples_per_second": 1244.867,
+      "eval_steps_per_second": 38.903,
+      "step": 1831680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0005889817821845e-05,
+      "loss": 3.8848,
+      "step": 1831936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9997503870311325e-05,
+      "loss": 3.8833,
+      "step": 1832448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.99891179228008e-05,
+      "loss": 3.8902,
+      "step": 1832960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.998073197529028e-05,
+      "loss": 3.8875,
+      "step": 1833472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.997234602777976e-05,
+      "loss": 3.8937,
+      "step": 1833984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.996396008026924e-05,
+      "loss": 3.8849,
+      "step": 1834496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.995557413275872e-05,
+      "loss": 3.8828,
+      "step": 1835008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9947188185248204e-05,
+      "loss": 3.8834,
+      "step": 1835520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9938802237737684e-05,
+      "loss": 3.8874,
+      "step": 1836032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.993041629022716e-05,
+      "loss": 3.8949,
+      "step": 1836544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.992203034271664e-05,
+      "loss": 3.8913,
+      "step": 1837056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.991364439520612e-05,
+      "loss": 3.8944,
+      "step": 1837568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9905274826499333e-05,
+      "loss": 3.881,
+      "step": 1838080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9896888878988813e-05,
+      "loss": 3.8809,
+      "step": 1838592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9888502931478293e-05,
+      "loss": 3.8914,
+      "step": 1839104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9880116983967773e-05,
+      "loss": 3.8735,
+      "step": 1839616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9871731036457253e-05,
+      "loss": 3.8809,
+      "step": 1840128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9863345088946733e-05,
+      "loss": 3.877,
+      "step": 1840640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9854959141436213e-05,
+      "loss": 3.883,
+      "step": 1841152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.984657319392569e-05,
+      "loss": 3.8889,
+      "step": 1841664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9838187246415173e-05,
+      "loss": 3.8944,
+      "step": 1842176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9829801298904653e-05,
+      "loss": 3.887,
+      "step": 1842688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9821431730197865e-05,
+      "loss": 3.8791,
+      "step": 1843200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9813045782687342e-05,
+      "loss": 3.8892,
+      "step": 1843712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9804659835176822e-05,
+      "loss": 3.8808,
+      "step": 1844224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9796273887666302e-05,
+      "loss": 3.8825,
+      "step": 1844736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9787887940155782e-05,
+      "loss": 3.8734,
+      "step": 1845248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9779501992645265e-05,
+      "loss": 3.8859,
+      "step": 1845760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9771116045134745e-05,
+      "loss": 3.8829,
+      "step": 1846272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9762730097624222e-05,
+      "loss": 3.8773,
+      "step": 1846784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9754360528917434e-05,
+      "loss": 3.8821,
+      "step": 1847296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9745990960210643e-05,
+      "loss": 3.8905,
+      "step": 1847808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9737605012700127e-05,
+      "loss": 3.883,
+      "step": 1848320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9729219065189607e-05,
+      "loss": 3.8828,
+      "step": 1848832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9720833117679087e-05,
+      "loss": 3.8865,
+      "step": 1849344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9712447170168566e-05,
+      "loss": 3.8891,
+      "step": 1849856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9704061222658046e-05,
+      "loss": 3.8797,
+      "step": 1850368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9695691653951256e-05,
+      "loss": 3.8683,
+      "step": 1850880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9687305706440735e-05,
+      "loss": 3.8751,
+      "step": 1851392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9678919758930215e-05,
+      "loss": 3.878,
+      "step": 1851904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9670533811419695e-05,
+      "loss": 3.8818,
+      "step": 1852416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9662147863909175e-05,
+      "loss": 3.8858,
+      "step": 1852928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9653761916398655e-05,
+      "loss": 3.8793,
+      "step": 1853440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9645392347691868e-05,
+      "loss": 3.881,
+      "step": 1853952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9637006400181344e-05,
+      "loss": 3.8904,
+      "step": 1854464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9628620452670828e-05,
+      "loss": 3.8773,
+      "step": 1854976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9620234505160308e-05,
+      "loss": 3.8931,
+      "step": 1855488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9611848557649788e-05,
+      "loss": 3.8759,
+      "step": 1856000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9603462610139268e-05,
+      "loss": 3.8571,
+      "step": 1856512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9595076662628748e-05,
+      "loss": 3.9018,
+      "step": 1857024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9586707093921957e-05,
+      "loss": 3.8794,
+      "step": 1857536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9578321146411437e-05,
+      "loss": 3.883,
+      "step": 1858048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.956993519890092e-05,
+      "loss": 3.8766,
+      "step": 1858560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.95615492513904e-05,
+      "loss": 3.8775,
+      "step": 1859072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.955317968268361e-05,
+      "loss": 3.8734,
+      "step": 1859584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.954479373517309e-05,
+      "loss": 3.877,
+      "step": 1860096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.953640778766257e-05,
+      "loss": 3.8832,
+      "step": 1860608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.952802184015205e-05,
+      "loss": 3.883,
+      "step": 1861120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.951965227144526e-05,
+      "loss": 3.8949,
+      "step": 1861632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.951126632393474e-05,
+      "loss": 3.8721,
+      "step": 1862144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.950288037642422e-05,
+      "loss": 3.8707,
+      "step": 1862656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.94944944289137e-05,
+      "loss": 3.8916,
+      "step": 1863168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9486108481403178e-05,
+      "loss": 3.8624,
+      "step": 1863680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9477722533892658e-05,
+      "loss": 3.8655,
+      "step": 1864192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9469336586382138e-05,
+      "loss": 3.8803,
+      "step": 1864704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.946095063887162e-05,
+      "loss": 3.884,
+      "step": 1865216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.945258107016483e-05,
+      "loss": 3.8721,
+      "step": 1865728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9444211501458043e-05,
+      "loss": 3.8784,
+      "step": 1866240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9435825553947523e-05,
+      "loss": 3.8691,
+      "step": 1866752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9427455985240735e-05,
+      "loss": 3.8871,
+      "step": 1867264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9419070037730215e-05,
+      "loss": 3.882,
+      "step": 1867776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9410684090219695e-05,
+      "loss": 3.8827,
+      "step": 1868288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9402298142709175e-05,
+      "loss": 3.8857,
+      "step": 1868800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.939391219519865e-05,
+      "loss": 3.8882,
+      "step": 1869312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.938552624768813e-05,
+      "loss": 3.8985,
+      "step": 1869824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.937714030017761e-05,
+      "loss": 3.8697,
+      "step": 1870336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.936875435266709e-05,
+      "loss": 3.8914,
+      "step": 1870848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9360368405156575e-05,
+      "loss": 3.8786,
+      "step": 1871360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9351982457646055e-05,
+      "loss": 3.8761,
+      "step": 1871872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.934359651013553e-05,
+      "loss": 3.886,
+      "step": 1872384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9335226941428744e-05,
+      "loss": 3.8849,
+      "step": 1872896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9326840993918224e-05,
+      "loss": 3.8877,
+      "step": 1873408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9318455046407704e-05,
+      "loss": 3.8973,
+      "step": 1873920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9310069098897184e-05,
+      "loss": 3.8727,
+      "step": 1874432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9301699530190396e-05,
+      "loss": 3.8664,
+      "step": 1874944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9293313582679876e-05,
+      "loss": 3.878,
+      "step": 1875456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9284927635169356e-05,
+      "loss": 3.8881,
+      "step": 1875968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9276541687658833e-05,
+      "loss": 3.8838,
+      "step": 1876480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9268155740148312e-05,
+      "loss": 3.882,
+      "step": 1876992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9259769792637792e-05,
+      "loss": 3.8733,
+      "step": 1877504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9251383845127276e-05,
+      "loss": 3.8736,
+      "step": 1878016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9242997897616756e-05,
+      "loss": 3.8842,
+      "step": 1878528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9234628328909965e-05,
+      "loss": 3.8762,
+      "step": 1879040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9226242381399445e-05,
+      "loss": 3.8854,
+      "step": 1879552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9217856433888925e-05,
+      "loss": 3.8898,
+      "step": 1880064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9209470486378405e-05,
+      "loss": 3.8851,
+      "step": 1880576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9201084538867885e-05,
+      "loss": 3.8755,
+      "step": 1881088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9192698591357365e-05,
+      "loss": 3.88,
+      "step": 1881600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9184312643846845e-05,
+      "loss": 3.8761,
+      "step": 1882112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9175926696336325e-05,
+      "loss": 3.8805,
+      "step": 1882624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9167540748825804e-05,
+      "loss": 3.8868,
+      "step": 1883136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9159171180119014e-05,
+      "loss": 3.8823,
+      "step": 1883648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.915080161141223e-05,
+      "loss": 3.8772,
+      "step": 1884160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.914241566390171e-05,
+      "loss": 3.8672,
+      "step": 1884672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9134029716391186e-05,
+      "loss": 3.8805,
+      "step": 1885184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9125643768880666e-05,
+      "loss": 3.8732,
+      "step": 1885696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.911727420017388e-05,
+      "loss": 3.8596,
+      "step": 1886208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.910888825266336e-05,
+      "loss": 3.882,
+      "step": 1886720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.910050230515284e-05,
+      "loss": 3.879,
+      "step": 1887232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9092116357642318e-05,
+      "loss": 3.8794,
+      "step": 1887744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9083730410131798e-05,
+      "loss": 3.8695,
+      "step": 1888256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9075360841425007e-05,
+      "loss": 3.8853,
+      "step": 1888768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9066974893914487e-05,
+      "loss": 3.8773,
+      "step": 1889280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9058588946403967e-05,
+      "loss": 3.869,
+      "step": 1889792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9050202998893447e-05,
+      "loss": 3.892,
+      "step": 1890304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.904181705138293e-05,
+      "loss": 3.886,
+      "step": 1890816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.903343110387241e-05,
+      "loss": 3.8791,
+      "step": 1891328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.902506153516562e-05,
+      "loss": 3.8829,
+      "step": 1891840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.90166755876551e-05,
+      "loss": 3.866,
+      "step": 1892352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.900828964014458e-05,
+      "loss": 3.8853,
+      "step": 1892864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.899990369263406e-05,
+      "loss": 3.8762,
+      "step": 1893376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8991534123927272e-05,
+      "loss": 3.8821,
+      "step": 1893888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8983148176416752e-05,
+      "loss": 3.874,
+      "step": 1894400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8974762228906232e-05,
+      "loss": 3.8825,
+      "step": 1894912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8966376281395712e-05,
+      "loss": 3.8755,
+      "step": 1895424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.895799033388519e-05,
+      "loss": 3.8716,
+      "step": 1895936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.89496207651784e-05,
+      "loss": 3.8707,
+      "step": 1896448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8941234817667884e-05,
+      "loss": 3.8808,
+      "step": 1896960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8932848870157364e-05,
+      "loss": 3.8757,
+      "step": 1897472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.892446292264684e-05,
+      "loss": 3.8872,
+      "step": 1897984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.891607697513632e-05,
+      "loss": 3.8758,
+      "step": 1898496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.89076910276258e-05,
+      "loss": 3.8914,
+      "step": 1899008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.889930508011528e-05,
+      "loss": 3.8652,
+      "step": 1899520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.889091913260476e-05,
+      "loss": 3.8886,
+      "step": 1900032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.888253318509424e-05,
+      "loss": 3.8863,
+      "step": 1900544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8874163616387453e-05,
+      "loss": 3.8867,
+      "step": 1901056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8865794047680662e-05,
+      "loss": 3.8778,
+      "step": 1901568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8857408100170142e-05,
+      "loss": 3.8791,
+      "step": 1902080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8849022152659622e-05,
+      "loss": 3.8812,
+      "step": 1902592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8840636205149105e-05,
+      "loss": 3.8965,
+      "step": 1903104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8832266636442314e-05,
+      "loss": 3.8694,
+      "step": 1903616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8823880688931794e-05,
+      "loss": 3.8764,
+      "step": 1904128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8815494741421274e-05,
+      "loss": 3.8736,
+      "step": 1904640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8807108793910754e-05,
+      "loss": 3.8848,
+      "step": 1905152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8798722846400234e-05,
+      "loss": 3.8713,
+      "step": 1905664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8790336898889714e-05,
+      "loss": 3.8813,
+      "step": 1906176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8781950951379194e-05,
+      "loss": 3.8805,
+      "step": 1906688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8773565003868674e-05,
+      "loss": 3.875,
+      "step": 1907200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8765179056358154e-05,
+      "loss": 3.8824,
+      "step": 1907712
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.984157085418701,
+      "eval_runtime": 299.1369,
+      "eval_samples_per_second": 1275.64,
+      "eval_steps_per_second": 39.865,
+      "step": 1908000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8756793108847634e-05,
+      "loss": 3.8733,
+      "step": 1908224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8748423540140843e-05,
+      "loss": 3.8739,
+      "step": 1908736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8740037592630323e-05,
+      "loss": 3.8874,
+      "step": 1909248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8731651645119806e-05,
+      "loss": 3.8837,
+      "step": 1909760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8723265697609286e-05,
+      "loss": 3.8863,
+      "step": 1910272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8714879750098766e-05,
+      "loss": 3.8798,
+      "step": 1910784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8706493802588246e-05,
+      "loss": 3.875,
+      "step": 1911296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8698107855077726e-05,
+      "loss": 3.8792,
+      "step": 1911808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8689738286370935e-05,
+      "loss": 3.8789,
+      "step": 1912320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8681352338860415e-05,
+      "loss": 3.8904,
+      "step": 1912832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.86729663913499e-05,
+      "loss": 3.886,
+      "step": 1913344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.866458044383938e-05,
+      "loss": 3.8861,
+      "step": 1913856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8656210875132588e-05,
+      "loss": 3.883,
+      "step": 1914368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8647824927622068e-05,
+      "loss": 3.868,
+      "step": 1914880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8639438980111548e-05,
+      "loss": 3.8852,
+      "step": 1915392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8631053032601024e-05,
+      "loss": 3.8693,
+      "step": 1915904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8622667085090508e-05,
+      "loss": 3.8745,
+      "step": 1916416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8614281137579987e-05,
+      "loss": 3.8742,
+      "step": 1916928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8605895190069467e-05,
+      "loss": 3.8737,
+      "step": 1917440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8597509242558947e-05,
+      "loss": 3.8827,
+      "step": 1917952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8589139673852156e-05,
+      "loss": 3.891,
+      "step": 1918464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8580753726341636e-05,
+      "loss": 3.8866,
+      "step": 1918976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8572367778831116e-05,
+      "loss": 3.8722,
+      "step": 1919488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.85639818313206e-05,
+      "loss": 3.8798,
+      "step": 1920000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.855561226261381e-05,
+      "loss": 3.8727,
+      "step": 1920512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.854722631510329e-05,
+      "loss": 3.8784,
+      "step": 1921024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.853884036759277e-05,
+      "loss": 3.8678,
+      "step": 1921536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.853045442008225e-05,
+      "loss": 3.8818,
+      "step": 1922048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.852208485137546e-05,
+      "loss": 3.8719,
+      "step": 1922560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8513715282668674e-05,
+      "loss": 3.8732,
+      "step": 1923072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8505345713961883e-05,
+      "loss": 3.8741,
+      "step": 1923584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8496959766451363e-05,
+      "loss": 3.8872,
+      "step": 1924096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8488590197744575e-05,
+      "loss": 3.8773,
+      "step": 1924608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8480204250234055e-05,
+      "loss": 3.8793,
+      "step": 1925120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8471818302723535e-05,
+      "loss": 3.8788,
+      "step": 1925632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8463432355213015e-05,
+      "loss": 3.8792,
+      "step": 1926144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8455046407702495e-05,
+      "loss": 3.8795,
+      "step": 1926656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.844666046019197e-05,
+      "loss": 3.8631,
+      "step": 1927168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.843827451268145e-05,
+      "loss": 3.8694,
+      "step": 1927680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.842988856517093e-05,
+      "loss": 3.8701,
+      "step": 1928192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8421502617660415e-05,
+      "loss": 3.8706,
+      "step": 1928704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8413116670149895e-05,
+      "loss": 3.878,
+      "step": 1929216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8404730722639375e-05,
+      "loss": 3.8749,
+      "step": 1929728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8396344775128855e-05,
+      "loss": 3.8812,
+      "step": 1930240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.838795882761833e-05,
+      "loss": 3.8781,
+      "step": 1930752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.837957288010781e-05,
+      "loss": 3.876,
+      "step": 1931264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.837118693259729e-05,
+      "loss": 3.8864,
+      "step": 1931776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8362817363890504e-05,
+      "loss": 3.872,
+      "step": 1932288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8354431416379984e-05,
+      "loss": 3.8546,
+      "step": 1932800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8346061847673196e-05,
+      "loss": 3.8912,
+      "step": 1933312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8337675900162676e-05,
+      "loss": 3.8739,
+      "step": 1933824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8329289952652153e-05,
+      "loss": 3.8772,
+      "step": 1934336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8320904005141633e-05,
+      "loss": 3.8753,
+      "step": 1934848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8312518057631116e-05,
+      "loss": 3.8691,
+      "step": 1935360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8304132110120596e-05,
+      "loss": 3.8666,
+      "step": 1935872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8295746162610076e-05,
+      "loss": 3.871,
+      "step": 1936384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8287360215099556e-05,
+      "loss": 3.8808,
+      "step": 1936896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8278974267589036e-05,
+      "loss": 3.8748,
+      "step": 1937408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8270588320078512e-05,
+      "loss": 3.8861,
+      "step": 1937920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8262202372567992e-05,
+      "loss": 3.8699,
+      "step": 1938432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8253816425057472e-05,
+      "loss": 3.8597,
+      "step": 1938944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8245446856350685e-05,
+      "loss": 3.8868,
+      "step": 1939456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8237077287643897e-05,
+      "loss": 3.8598,
+      "step": 1939968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8228691340133377e-05,
+      "loss": 3.86,
+      "step": 1940480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8220305392622857e-05,
+      "loss": 3.8761,
+      "step": 1940992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8211919445112334e-05,
+      "loss": 3.8728,
+      "step": 1941504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.820354987640555e-05,
+      "loss": 3.8667,
+      "step": 1942016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.819516392889503e-05,
+      "loss": 3.8711,
+      "step": 1942528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.818677798138451e-05,
+      "loss": 3.8648,
+      "step": 1943040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8178392033873986e-05,
+      "loss": 3.8799,
+      "step": 1943552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8170006086363466e-05,
+      "loss": 3.8802,
+      "step": 1944064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8161620138852946e-05,
+      "loss": 3.8769,
+      "step": 1944576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8153234191342426e-05,
+      "loss": 3.8795,
+      "step": 1945088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.814484824383191e-05,
+      "loss": 3.8818,
+      "step": 1945600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.813647867512512e-05,
+      "loss": 3.8912,
+      "step": 1946112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.812810910641833e-05,
+      "loss": 3.8655,
+      "step": 1946624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8119723158907807e-05,
+      "loss": 3.8873,
+      "step": 1947136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8111337211397287e-05,
+      "loss": 3.8708,
+      "step": 1947648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.810295126388677e-05,
+      "loss": 3.8753,
+      "step": 1948160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.809456531637625e-05,
+      "loss": 3.8763,
+      "step": 1948672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.808619574766946e-05,
+      "loss": 3.88,
+      "step": 1949184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.807780980015894e-05,
+      "loss": 3.8822,
+      "step": 1949696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.806942385264842e-05,
+      "loss": 3.8903,
+      "step": 1950208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.80610379051379e-05,
+      "loss": 3.8631,
+      "step": 1950720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.805265195762738e-05,
+      "loss": 3.865,
+      "step": 1951232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8044266010116863e-05,
+      "loss": 3.8722,
+      "step": 1951744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.803588006260634e-05,
+      "loss": 3.8824,
+      "step": 1952256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.802749411509582e-05,
+      "loss": 3.8728,
+      "step": 1952768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8019124546389032e-05,
+      "loss": 3.8747,
+      "step": 1953280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8010738598878512e-05,
+      "loss": 3.8715,
+      "step": 1953792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8002352651367992e-05,
+      "loss": 3.8732,
+      "step": 1954304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7993966703857472e-05,
+      "loss": 3.872,
+      "step": 1954816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7985580756346952e-05,
+      "loss": 3.8695,
+      "step": 1955328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.797721118764016e-05,
+      "loss": 3.8805,
+      "step": 1955840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.796882524012964e-05,
+      "loss": 3.8817,
+      "step": 1956352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.796043929261912e-05,
+      "loss": 3.8819,
+      "step": 1956864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.79520533451086e-05,
+      "loss": 3.8716,
+      "step": 1957376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.794366739759808e-05,
+      "loss": 3.8728,
+      "step": 1957888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7935281450087564e-05,
+      "loss": 3.8733,
+      "step": 1958400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7926911881380773e-05,
+      "loss": 3.8752,
+      "step": 1958912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7918542312673986e-05,
+      "loss": 3.8802,
+      "step": 1959424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7910156365163462e-05,
+      "loss": 3.876,
+      "step": 1959936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7901770417652942e-05,
+      "loss": 3.8717,
+      "step": 1960448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7893384470142425e-05,
+      "loss": 3.8608,
+      "step": 1960960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7884998522631905e-05,
+      "loss": 3.8767,
+      "step": 1961472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7876612575121385e-05,
+      "loss": 3.8652,
+      "step": 1961984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7868226627610865e-05,
+      "loss": 3.8552,
+      "step": 1962496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7859840680100342e-05,
+      "loss": 3.8722,
+      "step": 1963008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7851454732589822e-05,
+      "loss": 3.8829,
+      "step": 1963520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7843068785079302e-05,
+      "loss": 3.8693,
+      "step": 1964032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7834682837568785e-05,
+      "loss": 3.8654,
+      "step": 1964544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7826296890058265e-05,
+      "loss": 3.8766,
+      "step": 1965056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7817943700155207e-05,
+      "loss": 3.8724,
+      "step": 1965568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7809557752644687e-05,
+      "loss": 3.862,
+      "step": 1966080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7801171805134167e-05,
+      "loss": 3.8878,
+      "step": 1966592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7792785857623647e-05,
+      "loss": 3.8806,
+      "step": 1967104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7784399910113127e-05,
+      "loss": 3.8703,
+      "step": 1967616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.777603034140634e-05,
+      "loss": 3.8786,
+      "step": 1968128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7767644393895816e-05,
+      "loss": 3.8649,
+      "step": 1968640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7759274825189028e-05,
+      "loss": 3.8778,
+      "step": 1969152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7750888877678508e-05,
+      "loss": 3.8728,
+      "step": 1969664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7742502930167988e-05,
+      "loss": 3.8746,
+      "step": 1970176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7734116982657468e-05,
+      "loss": 3.8684,
+      "step": 1970688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7725731035146948e-05,
+      "loss": 3.879,
+      "step": 1971200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7717345087636428e-05,
+      "loss": 3.8723,
+      "step": 1971712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7708959140125908e-05,
+      "loss": 3.8687,
+      "step": 1972224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7700573192615388e-05,
+      "loss": 3.8643,
+      "step": 1972736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.76922036239086e-05,
+      "loss": 3.8749,
+      "step": 1973248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7683834055201813e-05,
+      "loss": 3.8679,
+      "step": 1973760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.767544810769129e-05,
+      "loss": 3.8813,
+      "step": 1974272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.766706216018077e-05,
+      "loss": 3.8712,
+      "step": 1974784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.765867621267025e-05,
+      "loss": 3.8838,
+      "step": 1975296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.765029026515973e-05,
+      "loss": 3.8572,
+      "step": 1975808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.764190431764921e-05,
+      "loss": 3.8842,
+      "step": 1976320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.763351837013869e-05,
+      "loss": 3.879,
+      "step": 1976832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7625132422628172e-05,
+      "loss": 3.8795,
+      "step": 1977344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.761674647511765e-05,
+      "loss": 3.875,
+      "step": 1977856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.760837690641086e-05,
+      "loss": 3.8734,
+      "step": 1978368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.759999095890034e-05,
+      "loss": 3.8757,
+      "step": 1978880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.759160501138982e-05,
+      "loss": 3.8939,
+      "step": 1979392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.75832190638793e-05,
+      "loss": 3.8636,
+      "step": 1979904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.757483311636878e-05,
+      "loss": 3.8739,
+      "step": 1980416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.756644716885826e-05,
+      "loss": 3.8715,
+      "step": 1980928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.755806122134774e-05,
+      "loss": 3.8752,
+      "step": 1981440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.754967527383722e-05,
+      "loss": 3.8649,
+      "step": 1981952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.754130570513043e-05,
+      "loss": 3.8784,
+      "step": 1982464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.753291975761991e-05,
+      "loss": 3.8717,
+      "step": 1982976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7524533810109394e-05,
+      "loss": 3.8729,
+      "step": 1983488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7516147862598874e-05,
+      "loss": 3.8737,
+      "step": 1984000
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9818334579467773,
+      "eval_runtime": 299.249,
+      "eval_samples_per_second": 1275.162,
+      "eval_steps_per_second": 39.85,
+      "step": 1984320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7507761915088354e-05,
+      "loss": 3.8668,
+      "step": 1984512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.749937596757783e-05,
+      "loss": 3.8685,
+      "step": 1985024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.749099002006731e-05,
+      "loss": 3.8798,
+      "step": 1985536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.748260407255679e-05,
+      "loss": 3.8784,
+      "step": 1986048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.747421812504627e-05,
+      "loss": 3.8792,
+      "step": 1986560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.746583217753575e-05,
+      "loss": 3.8735,
+      "step": 1987072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7457446230025233e-05,
+      "loss": 3.873,
+      "step": 1987584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.744906028251471e-05,
+      "loss": 3.8714,
+      "step": 1988096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.744067433500419e-05,
+      "loss": 3.8734,
+      "step": 1988608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.743228838749367e-05,
+      "loss": 3.8843,
+      "step": 1989120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.742390243998315e-05,
+      "loss": 3.8797,
+      "step": 1989632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.741551649247263e-05,
+      "loss": 3.8803,
+      "step": 1990144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7407146923765842e-05,
+      "loss": 3.8836,
+      "step": 1990656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7398760976255322e-05,
+      "loss": 3.8587,
+      "step": 1991168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7390375028744802e-05,
+      "loss": 3.8832,
+      "step": 1991680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7381989081234282e-05,
+      "loss": 3.8615,
+      "step": 1992192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.737361951252749e-05,
+      "loss": 3.8694,
+      "step": 1992704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.736523356501697e-05,
+      "loss": 3.8674,
+      "step": 1993216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.735684761750645e-05,
+      "loss": 3.8685,
+      "step": 1993728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7348461669995934e-05,
+      "loss": 3.8784,
+      "step": 1994240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7340075722485414e-05,
+      "loss": 3.8845,
+      "step": 1994752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.733168977497489e-05,
+      "loss": 3.8804,
+      "step": 1995264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.732330382746437e-05,
+      "loss": 3.8681,
+      "step": 1995776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.731491787995385e-05,
+      "loss": 3.868,
+      "step": 1996288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.730653193244333e-05,
+      "loss": 3.8741,
+      "step": 1996800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.729814598493281e-05,
+      "loss": 3.8685,
+      "step": 1997312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7289760037422294e-05,
+      "loss": 3.865,
+      "step": 1997824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7281374089911774e-05,
+      "loss": 3.8789,
+      "step": 1998336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7273004521204983e-05,
+      "loss": 3.8651,
+      "step": 1998848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7264618573694463e-05,
+      "loss": 3.8728,
+      "step": 1999360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7256232626183943e-05,
+      "loss": 3.863,
+      "step": 1999872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7247846678673423e-05,
+      "loss": 3.8805,
+      "step": 2000384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7239460731162903e-05,
+      "loss": 3.8754,
+      "step": 2000896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7231091162456115e-05,
+      "loss": 3.8702,
+      "step": 2001408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7222705214945595e-05,
+      "loss": 3.8779,
+      "step": 2001920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7214335646238804e-05,
+      "loss": 3.8684,
+      "step": 2002432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7205949698728284e-05,
+      "loss": 3.8772,
+      "step": 2002944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7197563751217764e-05,
+      "loss": 3.8632,
+      "step": 2003456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7189177803707244e-05,
+      "loss": 3.8585,
+      "step": 2003968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7180808235000457e-05,
+      "loss": 3.8632,
+      "step": 2004480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7172422287489937e-05,
+      "loss": 3.8712,
+      "step": 2004992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7164036339979417e-05,
+      "loss": 3.8698,
+      "step": 2005504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7155650392468897e-05,
+      "loss": 3.8657,
+      "step": 2006016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7147264444958373e-05,
+      "loss": 3.8749,
+      "step": 2006528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7138878497447857e-05,
+      "loss": 3.8729,
+      "step": 2007040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7130492549937337e-05,
+      "loss": 3.8726,
+      "step": 2007552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7122139360034278e-05,
+      "loss": 3.8844,
+      "step": 2008064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7113753412523758e-05,
+      "loss": 3.8661,
+      "step": 2008576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7105367465013238e-05,
+      "loss": 3.8455,
+      "step": 2009088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7096981517502718e-05,
+      "loss": 3.886,
+      "step": 2009600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7088595569992198e-05,
+      "loss": 3.8689,
+      "step": 2010112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7080209622481678e-05,
+      "loss": 3.872,
+      "step": 2010624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7071823674971158e-05,
+      "loss": 3.8666,
+      "step": 2011136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7063437727460638e-05,
+      "loss": 3.8665,
+      "step": 2011648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7055051779950118e-05,
+      "loss": 3.8609,
+      "step": 2012160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7046665832439598e-05,
+      "loss": 3.8618,
+      "step": 2012672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7038279884929078e-05,
+      "loss": 3.8805,
+      "step": 2013184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.702991031622229e-05,
+      "loss": 3.8641,
+      "step": 2013696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.702152436871177e-05,
+      "loss": 3.884,
+      "step": 2014208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.701313842120125e-05,
+      "loss": 3.8661,
+      "step": 2014720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7004752473690727e-05,
+      "loss": 3.8519,
+      "step": 2015232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6996366526180207e-05,
+      "loss": 3.8805,
+      "step": 2015744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.698799695747342e-05,
+      "loss": 3.8626,
+      "step": 2016256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.697962738876663e-05,
+      "loss": 3.8515,
+      "step": 2016768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.697124144125611e-05,
+      "loss": 3.8704,
+      "step": 2017280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.696285549374559e-05,
+      "loss": 3.8687,
+      "step": 2017792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.695446954623507e-05,
+      "loss": 3.8613,
+      "step": 2018304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.694608359872455e-05,
+      "loss": 3.8648,
+      "step": 2018816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6937697651214028e-05,
+      "loss": 3.8615,
+      "step": 2019328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.692931170370351e-05,
+      "loss": 3.873,
+      "step": 2019840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.692092575619299e-05,
+      "loss": 3.8783,
+      "step": 2020352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.69125561874862e-05,
+      "loss": 3.8671,
+      "step": 2020864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6904186618779413e-05,
+      "loss": 3.8751,
+      "step": 2021376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6895800671268893e-05,
+      "loss": 3.8743,
+      "step": 2021888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6887414723758373e-05,
+      "loss": 3.8857,
+      "step": 2022400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6879045155051585e-05,
+      "loss": 3.863,
+      "step": 2022912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6870659207541065e-05,
+      "loss": 3.8842,
+      "step": 2023424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6862273260030545e-05,
+      "loss": 3.8643,
+      "step": 2023936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6853887312520022e-05,
+      "loss": 3.867,
+      "step": 2024448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6845517743813234e-05,
+      "loss": 3.8748,
+      "step": 2024960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6837131796302718e-05,
+      "loss": 3.869,
+      "step": 2025472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6828745848792198e-05,
+      "loss": 3.8814,
+      "step": 2025984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6820359901281674e-05,
+      "loss": 3.8851,
+      "step": 2026496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6811973953771154e-05,
+      "loss": 3.8597,
+      "step": 2027008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6803588006260634e-05,
+      "loss": 3.8589,
+      "step": 2027520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6795202058750114e-05,
+      "loss": 3.8643,
+      "step": 2028032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6786816111239594e-05,
+      "loss": 3.8805,
+      "step": 2028544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6778430163729074e-05,
+      "loss": 3.8639,
+      "step": 2029056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6770044216218557e-05,
+      "loss": 3.8707,
+      "step": 2029568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6761658268708034e-05,
+      "loss": 3.8712,
+      "step": 2030080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6753272321197514e-05,
+      "loss": 3.8599,
+      "step": 2030592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6744886373686994e-05,
+      "loss": 3.8736,
+      "step": 2031104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6736500426176474e-05,
+      "loss": 3.8621,
+      "step": 2031616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6728114478665954e-05,
+      "loss": 3.8763,
+      "step": 2032128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6719728531155434e-05,
+      "loss": 3.8765,
+      "step": 2032640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6711358962448646e-05,
+      "loss": 3.8792,
+      "step": 2033152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6702973014938126e-05,
+      "loss": 3.8635,
+      "step": 2033664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6694587067427606e-05,
+      "loss": 3.8699,
+      "step": 2034176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6686201119917086e-05,
+      "loss": 3.8651,
+      "step": 2034688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6677847930014028e-05,
+      "loss": 3.8717,
+      "step": 2035200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6669461982503507e-05,
+      "loss": 3.8757,
+      "step": 2035712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6661076034992987e-05,
+      "loss": 3.866,
+      "step": 2036224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6652690087482467e-05,
+      "loss": 3.8698,
+      "step": 2036736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6644304139971947e-05,
+      "loss": 3.8582,
+      "step": 2037248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6635918192461427e-05,
+      "loss": 3.8686,
+      "step": 2037760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6627532244950907e-05,
+      "loss": 3.8641,
+      "step": 2038272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6619146297440387e-05,
+      "loss": 3.8486,
+      "step": 2038784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6610760349929867e-05,
+      "loss": 3.8678,
+      "step": 2039296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6602374402419347e-05,
+      "loss": 3.8722,
+      "step": 2039808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6593988454908827e-05,
+      "loss": 3.8653,
+      "step": 2040320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6585602507398307e-05,
+      "loss": 3.8605,
+      "step": 2040832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6577232938691516e-05,
+      "loss": 3.8697,
+      "step": 2041344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.656886336998473e-05,
+      "loss": 3.8687,
+      "step": 2041856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6560477422474212e-05,
+      "loss": 3.8567,
+      "step": 2042368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.655209147496369e-05,
+      "loss": 3.8869,
+      "step": 2042880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.654370552745317e-05,
+      "loss": 3.8701,
+      "step": 2043392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.653531957994265e-05,
+      "loss": 3.8658,
+      "step": 2043904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6526950011235858e-05,
+      "loss": 3.8744,
+      "step": 2044416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6518564063725337e-05,
+      "loss": 3.8606,
+      "step": 2044928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.651017811621482e-05,
+      "loss": 3.8711,
+      "step": 2045440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.65017921687043e-05,
+      "loss": 3.8694,
+      "step": 2045952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.649340622119378e-05,
+      "loss": 3.8645,
+      "step": 2046464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.648503665248699e-05,
+      "loss": 3.8689,
+      "step": 2046976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.647665070497647e-05,
+      "loss": 3.8704,
+      "step": 2047488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.646826475746595e-05,
+      "loss": 3.8697,
+      "step": 2048000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.645987880995543e-05,
+      "loss": 3.8605,
+      "step": 2048512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6451492862444913e-05,
+      "loss": 3.8646,
+      "step": 2049024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6443106914934393e-05,
+      "loss": 3.8636,
+      "step": 2049536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6434737346227602e-05,
+      "loss": 3.8629,
+      "step": 2050048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6426351398717082e-05,
+      "loss": 3.8809,
+      "step": 2050560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6417965451206562e-05,
+      "loss": 3.8684,
+      "step": 2051072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6409579503696042e-05,
+      "loss": 3.8794,
+      "step": 2051584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6401193556185522e-05,
+      "loss": 3.8493,
+      "step": 2052096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6392823987478734e-05,
+      "loss": 3.8844,
+      "step": 2052608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6384438039968214e-05,
+      "loss": 3.872,
+      "step": 2053120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.637605209245769e-05,
+      "loss": 3.8736,
+      "step": 2053632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.636766614494717e-05,
+      "loss": 3.8711,
+      "step": 2054144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.635928019743665e-05,
+      "loss": 3.8656,
+      "step": 2054656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6350910628729867e-05,
+      "loss": 3.8702,
+      "step": 2055168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6342524681219343e-05,
+      "loss": 3.885,
+      "step": 2055680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6334138733708823e-05,
+      "loss": 3.8584,
+      "step": 2056192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6325752786198303e-05,
+      "loss": 3.8707,
+      "step": 2056704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6317366838687783e-05,
+      "loss": 3.8644,
+      "step": 2057216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6308980891177263e-05,
+      "loss": 3.8714,
+      "step": 2057728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6300594943666743e-05,
+      "loss": 3.8625,
+      "step": 2058240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6292208996156223e-05,
+      "loss": 3.8681,
+      "step": 2058752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6283839427449436e-05,
+      "loss": 3.8686,
+      "step": 2059264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6275453479938915e-05,
+      "loss": 3.8723,
+      "step": 2059776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6267067532428395e-05,
+      "loss": 3.8655,
+      "step": 2060288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.980043649673462,
+      "eval_runtime": 297.2061,
+      "eval_samples_per_second": 1283.927,
+      "eval_steps_per_second": 40.124,
+      "step": 2060640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6258681584917872e-05,
+      "loss": 3.8601,
+      "step": 2060800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6250312016211084e-05,
+      "loss": 3.8663,
+      "step": 2061312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6241926068700568e-05,
+      "loss": 3.8722,
+      "step": 2061824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6233540121190044e-05,
+      "loss": 3.8735,
+      "step": 2062336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6225154173679524e-05,
+      "loss": 3.8771,
+      "step": 2062848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6216784604972737e-05,
+      "loss": 3.8706,
+      "step": 2063360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6208398657462217e-05,
+      "loss": 3.8665,
+      "step": 2063872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6200012709951697e-05,
+      "loss": 3.8729,
+      "step": 2064384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6191626762441177e-05,
+      "loss": 3.8648,
+      "step": 2064896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.618327357253812e-05,
+      "loss": 3.8779,
+      "step": 2065408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6174887625027598e-05,
+      "loss": 3.8772,
+      "step": 2065920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6166501677517078e-05,
+      "loss": 3.8788,
+      "step": 2066432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.615813210881029e-05,
+      "loss": 3.8752,
+      "step": 2066944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.614974616129977e-05,
+      "loss": 3.8559,
+      "step": 2067456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.614136021378925e-05,
+      "loss": 3.8788,
+      "step": 2067968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.613297426627873e-05,
+      "loss": 3.8547,
+      "step": 2068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.612458831876821e-05,
+      "loss": 3.8675,
+      "step": 2068992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.611620237125769e-05,
+      "loss": 3.8661,
+      "step": 2069504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6107816423747167e-05,
+      "loss": 3.8607,
+      "step": 2070016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.609943047623665e-05,
+      "loss": 3.8692,
+      "step": 2070528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.609104452872613e-05,
+      "loss": 3.8814,
+      "step": 2071040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.608265858121561e-05,
+      "loss": 3.8716,
+      "step": 2071552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.607427263370509e-05,
+      "loss": 3.8691,
+      "step": 2072064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.606588668619457e-05,
+      "loss": 3.8668,
+      "step": 2072576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.605750073868405e-05,
+      "loss": 3.8644,
+      "step": 2073088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6049114791173527e-05,
+      "loss": 3.8632,
+      "step": 2073600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6040728843663007e-05,
+      "loss": 3.8571,
+      "step": 2074112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.603234289615249e-05,
+      "loss": 3.873,
+      "step": 2074624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.60239733274457e-05,
+      "loss": 3.8627,
+      "step": 2075136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.601558737993518e-05,
+      "loss": 3.8665,
+      "step": 2075648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.600721781122839e-05,
+      "loss": 3.8637,
+      "step": 2076160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.599883186371787e-05,
+      "loss": 3.8747,
+      "step": 2076672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.599044591620735e-05,
+      "loss": 3.8653,
+      "step": 2077184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.598205996869683e-05,
+      "loss": 3.8702,
+      "step": 2077696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.597367402118631e-05,
+      "loss": 3.8721,
+      "step": 2078208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.596528807367579e-05,
+      "loss": 3.8661,
+      "step": 2078720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.595690212616527e-05,
+      "loss": 3.8739,
+      "step": 2079232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.594853255745848e-05,
+      "loss": 3.8543,
+      "step": 2079744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.594014660994796e-05,
+      "loss": 3.8564,
+      "step": 2080256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5931760662437444e-05,
+      "loss": 3.8593,
+      "step": 2080768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5923374714926924e-05,
+      "loss": 3.867,
+      "step": 2081280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5914988767416404e-05,
+      "loss": 3.8661,
+      "step": 2081792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5906619198709613e-05,
+      "loss": 3.8643,
+      "step": 2082304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5898233251199093e-05,
+      "loss": 3.8682,
+      "step": 2082816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5889847303688573e-05,
+      "loss": 3.8691,
+      "step": 2083328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5881461356178053e-05,
+      "loss": 3.8665,
+      "step": 2083840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5873075408667533e-05,
+      "loss": 3.8759,
+      "step": 2084352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5864689461157012e-05,
+      "loss": 3.8657,
+      "step": 2084864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5856303513646492e-05,
+      "loss": 3.8418,
+      "step": 2085376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5847933944939705e-05,
+      "loss": 3.8806,
+      "step": 2085888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.583954799742918e-05,
+      "loss": 3.8655,
+      "step": 2086400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.583116204991866e-05,
+      "loss": 3.8658,
+      "step": 2086912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5822776102408145e-05,
+      "loss": 3.8617,
+      "step": 2087424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5814390154897625e-05,
+      "loss": 3.8615,
+      "step": 2087936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5806004207387105e-05,
+      "loss": 3.8579,
+      "step": 2088448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5797618259876585e-05,
+      "loss": 3.8569,
+      "step": 2088960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5789248691169794e-05,
+      "loss": 3.8789,
+      "step": 2089472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5780862743659274e-05,
+      "loss": 3.8585,
+      "step": 2089984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5772476796148754e-05,
+      "loss": 3.8806,
+      "step": 2090496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5764090848638237e-05,
+      "loss": 3.8619,
+      "step": 2091008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5755704901127714e-05,
+      "loss": 3.8466,
+      "step": 2091520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5747318953617194e-05,
+      "loss": 3.8759,
+      "step": 2092032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5738933006106673e-05,
+      "loss": 3.8562,
+      "step": 2092544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5730547058596153e-05,
+      "loss": 3.8506,
+      "step": 2093056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5722177489889363e-05,
+      "loss": 3.8636,
+      "step": 2093568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.571380792118258e-05,
+      "loss": 3.8623,
+      "step": 2094080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.570542197367206e-05,
+      "loss": 3.857,
+      "step": 2094592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5697036026161535e-05,
+      "loss": 3.8654,
+      "step": 2095104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5688650078651015e-05,
+      "loss": 3.8587,
+      "step": 2095616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5680264131140495e-05,
+      "loss": 3.8652,
+      "step": 2096128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5671894562433707e-05,
+      "loss": 3.8707,
+      "step": 2096640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5663508614923187e-05,
+      "loss": 3.8667,
+      "step": 2097152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5655122667412667e-05,
+      "loss": 3.8725,
+      "step": 2097664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5646736719902147e-05,
+      "loss": 3.869,
+      "step": 2098176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5638350772391627e-05,
+      "loss": 3.8806,
+      "step": 2098688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5629964824881107e-05,
+      "loss": 3.8603,
+      "step": 2099200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5621595256174316e-05,
+      "loss": 3.8789,
+      "step": 2099712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5613225687467532e-05,
+      "loss": 3.858,
+      "step": 2100224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.560483973995701e-05,
+      "loss": 3.8623,
+      "step": 2100736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.559645379244649e-05,
+      "loss": 3.8677,
+      "step": 2101248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.558806784493597e-05,
+      "loss": 3.8675,
+      "step": 2101760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.557968189742545e-05,
+      "loss": 3.8732,
+      "step": 2102272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.557129594991493e-05,
+      "loss": 3.8832,
+      "step": 2102784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.556291000240441e-05,
+      "loss": 3.8605,
+      "step": 2103296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5554524054893892e-05,
+      "loss": 3.8527,
+      "step": 2103808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.554613810738337e-05,
+      "loss": 3.8564,
+      "step": 2104320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5537752159872848e-05,
+      "loss": 3.878,
+      "step": 2104832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.552938259116606e-05,
+      "loss": 3.8589,
+      "step": 2105344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5520996643655537e-05,
+      "loss": 3.8703,
+      "step": 2105856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.551261069614502e-05,
+      "loss": 3.8681,
+      "step": 2106368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.55042247486345e-05,
+      "loss": 3.8549,
+      "step": 2106880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.549583880112398e-05,
+      "loss": 3.8661,
+      "step": 2107392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.548746923241719e-05,
+      "loss": 3.861,
+      "step": 2107904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.547908328490667e-05,
+      "loss": 3.8692,
+      "step": 2108416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.547069733739615e-05,
+      "loss": 3.8742,
+      "step": 2108928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.546231138988563e-05,
+      "loss": 3.8737,
+      "step": 2109440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5453941821178842e-05,
+      "loss": 3.8605,
+      "step": 2109952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5445555873668322e-05,
+      "loss": 3.8646,
+      "step": 2110464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5437169926157802e-05,
+      "loss": 3.8608,
+      "step": 2110976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5428783978647282e-05,
+      "loss": 3.8662,
+      "step": 2111488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5420398031136762e-05,
+      "loss": 3.8717,
+      "step": 2112000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5412012083626242e-05,
+      "loss": 3.8617,
+      "step": 2112512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5403626136115722e-05,
+      "loss": 3.8664,
+      "step": 2113024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5395240188605202e-05,
+      "loss": 3.8523,
+      "step": 2113536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5386854241094682e-05,
+      "loss": 3.8652,
+      "step": 2114048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.537846829358416e-05,
+      "loss": 3.8599,
+      "step": 2114560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5370115103681103e-05,
+      "loss": 3.8451,
+      "step": 2115072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5361729156170583e-05,
+      "loss": 3.8637,
+      "step": 2115584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5353343208660063e-05,
+      "loss": 3.8681,
+      "step": 2116096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5344957261149547e-05,
+      "loss": 3.8595,
+      "step": 2116608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5336571313639023e-05,
+      "loss": 3.8566,
+      "step": 2117120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5328185366128503e-05,
+      "loss": 3.8642,
+      "step": 2117632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5319799418617983e-05,
+      "loss": 3.8633,
+      "step": 2118144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5311413471107463e-05,
+      "loss": 3.8536,
+      "step": 2118656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5303043902400675e-05,
+      "loss": 3.8761,
+      "step": 2119168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5294674333693888e-05,
+      "loss": 3.8653,
+      "step": 2119680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5286288386183368e-05,
+      "loss": 3.8631,
+      "step": 2120192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5277918817476577e-05,
+      "loss": 3.8702,
+      "step": 2120704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5269532869966057e-05,
+      "loss": 3.8574,
+      "step": 2121216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.526116330125927e-05,
+      "loss": 3.8661,
+      "step": 2121728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5252777353748748e-05,
+      "loss": 3.8684,
+      "step": 2122240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.524439140623823e-05,
+      "loss": 3.8618,
+      "step": 2122752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.523600545872771e-05,
+      "loss": 3.8652,
+      "step": 2123264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.522761951121719e-05,
+      "loss": 3.8648,
+      "step": 2123776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5219233563706668e-05,
+      "loss": 3.8658,
+      "step": 2124288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5210847616196147e-05,
+      "loss": 3.8558,
+      "step": 2124800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5202461668685627e-05,
+      "loss": 3.8565,
+      "step": 2125312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5194075721175107e-05,
+      "loss": 3.862,
+      "step": 2125824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5185706152468318e-05,
+      "loss": 3.8565,
+      "step": 2126336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5177320204957798e-05,
+      "loss": 3.8756,
+      "step": 2126848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5168934257447278e-05,
+      "loss": 3.8608,
+      "step": 2127360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.516054830993676e-05,
+      "loss": 3.8766,
+      "step": 2127872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.515216236242624e-05,
+      "loss": 3.8435,
+      "step": 2128384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5143792793719449e-05,
+      "loss": 3.88,
+      "step": 2128896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.513540684620893e-05,
+      "loss": 3.8646,
+      "step": 2129408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.512702089869841e-05,
+      "loss": 3.8758,
+      "step": 2129920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.511863495118789e-05,
+      "loss": 3.8672,
+      "step": 2130432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.511024900367737e-05,
+      "loss": 3.8588,
+      "step": 2130944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5101863056166849e-05,
+      "loss": 3.8688,
+      "step": 2131456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5093477108656329e-05,
+      "loss": 3.8864,
+      "step": 2131968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5085107539949541e-05,
+      "loss": 3.8498,
+      "step": 2132480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5076721592439023e-05,
+      "loss": 3.8678,
+      "step": 2132992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.50683356449285e-05,
+      "loss": 3.8618,
+      "step": 2133504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.505994969741798e-05,
+      "loss": 3.8655,
+      "step": 2134016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.505156374990746e-05,
+      "loss": 3.8569,
+      "step": 2134528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.504317780239694e-05,
+      "loss": 3.8645,
+      "step": 2135040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.503479185488642e-05,
+      "loss": 3.8602,
+      "step": 2135552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.50264059073759e-05,
+      "loss": 3.8679,
+      "step": 2136064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5018036338669111e-05,
+      "loss": 3.8649,
+      "step": 2136576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9789927005767822,
+      "eval_runtime": 298.1731,
+      "eval_samples_per_second": 1279.763,
+      "eval_steps_per_second": 39.994,
+      "step": 2136960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5009650391158591e-05,
+      "loss": 3.8653,
+      "step": 2137088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5001264443648071e-05,
+      "loss": 3.859,
+      "step": 2137600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4992878496137553e-05,
+      "loss": 3.8701,
+      "step": 2138112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.498449254862703e-05,
+      "loss": 3.8709,
+      "step": 2138624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.497610660111651e-05,
+      "loss": 3.8686,
+      "step": 2139136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4967720653605991e-05,
+      "loss": 3.8657,
+      "step": 2139648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4959334706095471e-05,
+      "loss": 3.8681,
+      "step": 2140160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4950948758584951e-05,
+      "loss": 3.866,
+      "step": 2140672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4942562811074431e-05,
+      "loss": 3.86,
+      "step": 2141184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4934176863563911e-05,
+      "loss": 3.8692,
+      "step": 2141696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.492579091605339e-05,
+      "loss": 3.8768,
+      "step": 2142208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.491740496854287e-05,
+      "loss": 3.8726,
+      "step": 2142720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4909035399836083e-05,
+      "loss": 3.8747,
+      "step": 2143232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.490064945232556e-05,
+      "loss": 3.8522,
+      "step": 2143744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.489226350481504e-05,
+      "loss": 3.874,
+      "step": 2144256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4883877557304522e-05,
+      "loss": 3.8517,
+      "step": 2144768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4875491609794002e-05,
+      "loss": 3.8656,
+      "step": 2145280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4867105662283482e-05,
+      "loss": 3.8586,
+      "step": 2145792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4858719714772961e-05,
+      "loss": 3.8557,
+      "step": 2146304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4850333767262441e-05,
+      "loss": 3.8649,
+      "step": 2146816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.484194781975192e-05,
+      "loss": 3.8774,
+      "step": 2147328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.48335618722414e-05,
+      "loss": 3.8694,
+      "step": 2147840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4825192303534612e-05,
+      "loss": 3.8655,
+      "step": 2148352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4816806356024094e-05,
+      "loss": 3.864,
+      "step": 2148864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.480842040851357e-05,
+      "loss": 3.8644,
+      "step": 2149376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.480003446100305e-05,
+      "loss": 3.8606,
+      "step": 2149888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4791648513492532e-05,
+      "loss": 3.8507,
+      "step": 2150400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4783262565982012e-05,
+      "loss": 3.8676,
+      "step": 2150912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4774892997275223e-05,
+      "loss": 3.8575,
+      "step": 2151424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4766507049764703e-05,
+      "loss": 3.8619,
+      "step": 2151936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4758121102254183e-05,
+      "loss": 3.8616,
+      "step": 2152448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4749751533547393e-05,
+      "loss": 3.8686,
+      "step": 2152960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4741365586036873e-05,
+      "loss": 3.8582,
+      "step": 2153472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4732979638526353e-05,
+      "loss": 3.8671,
+      "step": 2153984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4724593691015833e-05,
+      "loss": 3.8676,
+      "step": 2154496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4716207743505315e-05,
+      "loss": 3.8621,
+      "step": 2155008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4707821795994795e-05,
+      "loss": 3.8714,
+      "step": 2155520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4699435848484275e-05,
+      "loss": 3.8472,
+      "step": 2156032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4691049900973753e-05,
+      "loss": 3.8563,
+      "step": 2156544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4682663953463233e-05,
+      "loss": 3.8565,
+      "step": 2157056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4674294384756446e-05,
+      "loss": 3.8578,
+      "step": 2157568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4665908437245924e-05,
+      "loss": 3.8616,
+      "step": 2158080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4657522489735404e-05,
+      "loss": 3.8619,
+      "step": 2158592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4649136542224884e-05,
+      "loss": 3.8644,
+      "step": 2159104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4640766973518096e-05,
+      "loss": 3.8655,
+      "step": 2159616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4632381026007574e-05,
+      "loss": 3.861,
+      "step": 2160128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4623995078497054e-05,
+      "loss": 3.8707,
+      "step": 2160640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4615609130986534e-05,
+      "loss": 3.858,
+      "step": 2161152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4607223183476016e-05,
+      "loss": 3.8402,
+      "step": 2161664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4598837235965496e-05,
+      "loss": 3.8743,
+      "step": 2162176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4590467667258707e-05,
+      "loss": 3.8626,
+      "step": 2162688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4582081719748187e-05,
+      "loss": 3.8604,
+      "step": 2163200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4573695772237667e-05,
+      "loss": 3.8587,
+      "step": 2163712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4565309824727147e-05,
+      "loss": 3.858,
+      "step": 2164224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4556923877216627e-05,
+      "loss": 3.855,
+      "step": 2164736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4548537929706105e-05,
+      "loss": 3.8523,
+      "step": 2165248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4540151982195585e-05,
+      "loss": 3.8745,
+      "step": 2165760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4531766034685065e-05,
+      "loss": 3.8556,
+      "step": 2166272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4523396465978279e-05,
+      "loss": 3.8771,
+      "step": 2166784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4515010518467756e-05,
+      "loss": 3.861,
+      "step": 2167296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4506624570957235e-05,
+      "loss": 3.8383,
+      "step": 2167808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4498238623446717e-05,
+      "loss": 3.8671,
+      "step": 2168320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4489852675936197e-05,
+      "loss": 3.8543,
+      "step": 2168832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4481466728425677e-05,
+      "loss": 3.8473,
+      "step": 2169344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4473097159718888e-05,
+      "loss": 3.8549,
+      "step": 2169856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4464711212208368e-05,
+      "loss": 3.8606,
+      "step": 2170368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4456325264697848e-05,
+      "loss": 3.8542,
+      "step": 2170880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4447939317187328e-05,
+      "loss": 3.8617,
+      "step": 2171392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.443955336967681e-05,
+      "loss": 3.849,
+      "step": 2171904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4431167422166286e-05,
+      "loss": 3.8628,
+      "step": 2172416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4422781474655766e-05,
+      "loss": 3.867,
+      "step": 2172928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4414395527145247e-05,
+      "loss": 3.8625,
+      "step": 2173440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.440602595843846e-05,
+      "loss": 3.8659,
+      "step": 2173952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4397640010927938e-05,
+      "loss": 3.8655,
+      "step": 2174464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4389254063417418e-05,
+      "loss": 3.8772,
+      "step": 2174976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.438088449471063e-05,
+      "loss": 3.858,
+      "step": 2175488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4372514926003841e-05,
+      "loss": 3.8705,
+      "step": 2176000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4364128978493321e-05,
+      "loss": 3.8581,
+      "step": 2176512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4355743030982801e-05,
+      "loss": 3.8585,
+      "step": 2177024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4347357083472281e-05,
+      "loss": 3.8662,
+      "step": 2177536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.433897113596176e-05,
+      "loss": 3.8556,
+      "step": 2178048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.433058518845124e-05,
+      "loss": 3.8678,
+      "step": 2178560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.432219924094072e-05,
+      "loss": 3.8786,
+      "step": 2179072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4313813293430201e-05,
+      "loss": 3.8615,
+      "step": 2179584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.430544372472341e-05,
+      "loss": 3.8426,
+      "step": 2180096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4297057777212892e-05,
+      "loss": 3.8542,
+      "step": 2180608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4288671829702372e-05,
+      "loss": 3.8693,
+      "step": 2181120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4280285882191852e-05,
+      "loss": 3.8554,
+      "step": 2181632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4271899934681332e-05,
+      "loss": 3.864,
+      "step": 2182144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4263513987170812e-05,
+      "loss": 3.8642,
+      "step": 2182656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4255144418464023e-05,
+      "loss": 3.8507,
+      "step": 2183168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4246758470953502e-05,
+      "loss": 3.8633,
+      "step": 2183680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4238372523442982e-05,
+      "loss": 3.8556,
+      "step": 2184192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4229986575932464e-05,
+      "loss": 3.8692,
+      "step": 2184704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.422160062842194e-05,
+      "loss": 3.8679,
+      "step": 2185216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.421321468091142e-05,
+      "loss": 3.8695,
+      "step": 2185728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4204828733400902e-05,
+      "loss": 3.855,
+      "step": 2186240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4196442785890382e-05,
+      "loss": 3.8589,
+      "step": 2186752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4188073217183593e-05,
+      "loss": 3.8621,
+      "step": 2187264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4179703648476805e-05,
+      "loss": 3.8597,
+      "step": 2187776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4171317700966285e-05,
+      "loss": 3.8674,
+      "step": 2188288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4162931753455764e-05,
+      "loss": 3.8561,
+      "step": 2188800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4154545805945244e-05,
+      "loss": 3.8608,
+      "step": 2189312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4146176237238456e-05,
+      "loss": 3.8526,
+      "step": 2189824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4137790289727936e-05,
+      "loss": 3.86,
+      "step": 2190336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4129404342217414e-05,
+      "loss": 3.8543,
+      "step": 2190848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4121018394706894e-05,
+      "loss": 3.8435,
+      "step": 2191360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4112632447196374e-05,
+      "loss": 3.8559,
+      "step": 2191872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4104246499685856e-05,
+      "loss": 3.8669,
+      "step": 2192384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4095876930979065e-05,
+      "loss": 3.8537,
+      "step": 2192896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4087490983468547e-05,
+      "loss": 3.8574,
+      "step": 2193408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4079105035958027e-05,
+      "loss": 3.8574,
+      "step": 2193920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4070719088447507e-05,
+      "loss": 3.8629,
+      "step": 2194432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4062333140936987e-05,
+      "loss": 3.8521,
+      "step": 2194944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4053947193426466e-05,
+      "loss": 3.8677,
+      "step": 2195456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4045561245915945e-05,
+      "loss": 3.8594,
+      "step": 2195968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4037175298405425e-05,
+      "loss": 3.8655,
+      "step": 2196480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4028805729698639e-05,
+      "loss": 3.8641,
+      "step": 2196992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4020419782188119e-05,
+      "loss": 3.8532,
+      "step": 2197504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4012033834677595e-05,
+      "loss": 3.8623,
+      "step": 2198016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4003647887167075e-05,
+      "loss": 3.8633,
+      "step": 2198528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3995261939656557e-05,
+      "loss": 3.8582,
+      "step": 2199040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.398689237094977e-05,
+      "loss": 3.8603,
+      "step": 2199552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.397852280224298e-05,
+      "loss": 3.8559,
+      "step": 2200064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.397013685473246e-05,
+      "loss": 3.8632,
+      "step": 2200576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.396175090722194e-05,
+      "loss": 3.8544,
+      "step": 2201088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3953364959711418e-05,
+      "loss": 3.8535,
+      "step": 2201600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3944979012200898e-05,
+      "loss": 3.8534,
+      "step": 2202112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3936593064690378e-05,
+      "loss": 3.854,
+      "step": 2202624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3928207117179858e-05,
+      "loss": 3.8735,
+      "step": 2203136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.391982116966934e-05,
+      "loss": 3.8581,
+      "step": 2203648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.391143522215882e-05,
+      "loss": 3.8734,
+      "step": 2204160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.39030492746483e-05,
+      "loss": 3.8442,
+      "step": 2204672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.389467970594151e-05,
+      "loss": 3.8689,
+      "step": 2205184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.388629375843099e-05,
+      "loss": 3.8655,
+      "step": 2205696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.387790781092047e-05,
+      "loss": 3.8675,
+      "step": 2206208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3869521863409949e-05,
+      "loss": 3.8643,
+      "step": 2206720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3861135915899429e-05,
+      "loss": 3.855,
+      "step": 2207232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3852749968388909e-05,
+      "loss": 3.8624,
+      "step": 2207744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3844364020878389e-05,
+      "loss": 3.8781,
+      "step": 2208256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3836010830975332e-05,
+      "loss": 3.8484,
+      "step": 2208768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3827624883464812e-05,
+      "loss": 3.8603,
+      "step": 2209280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3819238935954294e-05,
+      "loss": 3.8592,
+      "step": 2209792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3810852988443774e-05,
+      "loss": 3.861,
+      "step": 2210304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.380246704093325e-05,
+      "loss": 3.8575,
+      "step": 2210816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3794081093422732e-05,
+      "loss": 3.8613,
+      "step": 2211328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3785695145912212e-05,
+      "loss": 3.8543,
+      "step": 2211840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3777309198401692e-05,
+      "loss": 3.8636,
+      "step": 2212352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3768923250891172e-05,
+      "loss": 3.8601,
+      "step": 2212864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9774677753448486,
+      "eval_runtime": 317.6092,
+      "eval_samples_per_second": 1201.448,
+      "eval_steps_per_second": 37.546,
+      "step": 2213280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3760537303380652e-05,
+      "loss": 3.8559,
+      "step": 2213376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3752167734673862e-05,
+      "loss": 3.8531,
+      "step": 2213888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3743781787163342e-05,
+      "loss": 3.8671,
+      "step": 2214400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3735395839652822e-05,
+      "loss": 3.8693,
+      "step": 2214912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3727009892142304e-05,
+      "loss": 3.8631,
+      "step": 2215424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.371862394463178e-05,
+      "loss": 3.8619,
+      "step": 2215936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.371023799712126e-05,
+      "loss": 3.862,
+      "step": 2216448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3701852049610742e-05,
+      "loss": 3.8663,
+      "step": 2216960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3693466102100222e-05,
+      "loss": 3.8518,
+      "step": 2217472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3685112912197165e-05,
+      "loss": 3.871,
+      "step": 2217984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3676726964686645e-05,
+      "loss": 3.8713,
+      "step": 2218496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3668341017176125e-05,
+      "loss": 3.8661,
+      "step": 2219008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3659971448469336e-05,
+      "loss": 3.8766,
+      "step": 2219520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3651585500958816e-05,
+      "loss": 3.8434,
+      "step": 2220032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3643199553448296e-05,
+      "loss": 3.8653,
+      "step": 2220544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3634813605937776e-05,
+      "loss": 3.8504,
+      "step": 2221056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3626444037230987e-05,
+      "loss": 3.8616,
+      "step": 2221568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3618058089720467e-05,
+      "loss": 3.8569,
+      "step": 2222080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3609672142209948e-05,
+      "loss": 3.8524,
+      "step": 2222592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3601286194699428e-05,
+      "loss": 3.8548,
+      "step": 2223104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3592900247188905e-05,
+      "loss": 3.8783,
+      "step": 2223616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3584514299678387e-05,
+      "loss": 3.8656,
+      "step": 2224128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3576128352167867e-05,
+      "loss": 3.8605,
+      "step": 2224640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3567742404657346e-05,
+      "loss": 3.8585,
+      "step": 2225152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3559356457146826e-05,
+      "loss": 3.8644,
+      "step": 2225664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3550970509636306e-05,
+      "loss": 3.8515,
+      "step": 2226176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3542584562125785e-05,
+      "loss": 3.8478,
+      "step": 2226688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3534214993418997e-05,
+      "loss": 3.8637,
+      "step": 2227200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3525829045908479e-05,
+      "loss": 3.8569,
+      "step": 2227712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3517443098397959e-05,
+      "loss": 3.8533,
+      "step": 2228224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3509057150887435e-05,
+      "loss": 3.8596,
+      "step": 2228736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3500671203376917e-05,
+      "loss": 3.8619,
+      "step": 2229248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3492285255866397e-05,
+      "loss": 3.8601,
+      "step": 2229760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3483899308355877e-05,
+      "loss": 3.861,
+      "step": 2230272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3475529739649088e-05,
+      "loss": 3.861,
+      "step": 2230784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3467143792138568e-05,
+      "loss": 3.8656,
+      "step": 2231296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3458757844628048e-05,
+      "loss": 3.8668,
+      "step": 2231808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3450371897117528e-05,
+      "loss": 3.838,
+      "step": 2232320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3441985949607007e-05,
+      "loss": 3.8516,
+      "step": 2232832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3433600002096489e-05,
+      "loss": 3.8528,
+      "step": 2233344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3425214054585966e-05,
+      "loss": 3.8542,
+      "step": 2233856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3416828107075446e-05,
+      "loss": 3.859,
+      "step": 2234368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3408474917172389e-05,
+      "loss": 3.8599,
+      "step": 2234880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3400088969661869e-05,
+      "loss": 3.8578,
+      "step": 2235392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3391719400955083e-05,
+      "loss": 3.8622,
+      "step": 2235904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.338333345344456e-05,
+      "loss": 3.8604,
+      "step": 2236416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3374947505934041e-05,
+      "loss": 3.8634,
+      "step": 2236928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3366561558423521e-05,
+      "loss": 3.8586,
+      "step": 2237440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3358175610913001e-05,
+      "loss": 3.8369,
+      "step": 2237952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3349789663402481e-05,
+      "loss": 3.8681,
+      "step": 2238464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3341403715891961e-05,
+      "loss": 3.8617,
+      "step": 2238976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.333301776838144e-05,
+      "loss": 3.8593,
+      "step": 2239488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3324648199674652e-05,
+      "loss": 3.8521,
+      "step": 2240000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3316262252164134e-05,
+      "loss": 3.8538,
+      "step": 2240512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3307876304653614e-05,
+      "loss": 3.8502,
+      "step": 2241024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.329949035714309e-05,
+      "loss": 3.8487,
+      "step": 2241536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3291104409632572e-05,
+      "loss": 3.8682,
+      "step": 2242048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3282718462122052e-05,
+      "loss": 3.8479,
+      "step": 2242560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3274332514611532e-05,
+      "loss": 3.8713,
+      "step": 2243072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3265962945904742e-05,
+      "loss": 3.8638,
+      "step": 2243584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3257576998394222e-05,
+      "loss": 3.8372,
+      "step": 2244096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3249191050883702e-05,
+      "loss": 3.8636,
+      "step": 2244608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3240821482176913e-05,
+      "loss": 3.8481,
+      "step": 2245120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3232435534666393e-05,
+      "loss": 3.8442,
+      "step": 2245632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3224049587155873e-05,
+      "loss": 3.8488,
+      "step": 2246144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3215663639645353e-05,
+      "loss": 3.8562,
+      "step": 2246656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3207277692134835e-05,
+      "loss": 3.8547,
+      "step": 2247168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3198908123428044e-05,
+      "loss": 3.8582,
+      "step": 2247680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3190522175917525e-05,
+      "loss": 3.8437,
+      "step": 2248192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3182136228407005e-05,
+      "loss": 3.8595,
+      "step": 2248704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3173750280896485e-05,
+      "loss": 3.8661,
+      "step": 2249216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3165364333385965e-05,
+      "loss": 3.8604,
+      "step": 2249728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3156994764679176e-05,
+      "loss": 3.8617,
+      "step": 2250240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3148608817168656e-05,
+      "loss": 3.8598,
+      "step": 2250752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3140222869658136e-05,
+      "loss": 3.8724,
+      "step": 2251264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3131836922147616e-05,
+      "loss": 3.852,
+      "step": 2251776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3123467353440827e-05,
+      "loss": 3.8642,
+      "step": 2252288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3115081405930307e-05,
+      "loss": 3.8575,
+      "step": 2252800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3106695458419788e-05,
+      "loss": 3.8554,
+      "step": 2253312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3098309510909268e-05,
+      "loss": 3.8598,
+      "step": 2253824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3089923563398745e-05,
+      "loss": 3.8512,
+      "step": 2254336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3081537615888226e-05,
+      "loss": 3.865,
+      "step": 2254848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3073151668377706e-05,
+      "loss": 3.8751,
+      "step": 2255360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3064765720867186e-05,
+      "loss": 3.8608,
+      "step": 2255872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3056379773356666e-05,
+      "loss": 3.8426,
+      "step": 2256384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3047993825846146e-05,
+      "loss": 3.8488,
+      "step": 2256896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3039607878335625e-05,
+      "loss": 3.8617,
+      "step": 2257408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3031221930825105e-05,
+      "loss": 3.8562,
+      "step": 2257920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3022852362118319e-05,
+      "loss": 3.8569,
+      "step": 2258432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3014482793411528e-05,
+      "loss": 3.8573,
+      "step": 2258944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3006096845901008e-05,
+      "loss": 3.8505,
+      "step": 2259456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.299771089839049e-05,
+      "loss": 3.8556,
+      "step": 2259968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2989341329683699e-05,
+      "loss": 3.8531,
+      "step": 2260480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.298095538217318e-05,
+      "loss": 3.8626,
+      "step": 2260992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.297256943466266e-05,
+      "loss": 3.8628,
+      "step": 2261504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.296418348715214e-05,
+      "loss": 3.8685,
+      "step": 2262016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.295579753964162e-05,
+      "loss": 3.8491,
+      "step": 2262528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2947411592131098e-05,
+      "loss": 3.855,
+      "step": 2263040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.293904202342431e-05,
+      "loss": 3.8615,
+      "step": 2263552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.293065607591379e-05,
+      "loss": 3.8532,
+      "step": 2264064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2922270128403272e-05,
+      "loss": 3.8664,
+      "step": 2264576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2913884180892749e-05,
+      "loss": 3.8522,
+      "step": 2265088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2905498233382229e-05,
+      "loss": 3.8531,
+      "step": 2265600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.289711228587171e-05,
+      "loss": 3.8472,
+      "step": 2266112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.288872633836119e-05,
+      "loss": 3.8604,
+      "step": 2266624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.288034039085067e-05,
+      "loss": 3.8516,
+      "step": 2267136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2871970822143881e-05,
+      "loss": 3.838,
+      "step": 2267648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2863584874633361e-05,
+      "loss": 3.8544,
+      "step": 2268160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2855198927122841e-05,
+      "loss": 3.8629,
+      "step": 2268672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2846812979612321e-05,
+      "loss": 3.852,
+      "step": 2269184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2838427032101801e-05,
+      "loss": 3.8504,
+      "step": 2269696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.283004108459128e-05,
+      "loss": 3.8517,
+      "step": 2270208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2821671515884492e-05,
+      "loss": 3.8595,
+      "step": 2270720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2813285568373973e-05,
+      "loss": 3.8479,
+      "step": 2271232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2804899620863453e-05,
+      "loss": 3.8616,
+      "step": 2271744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.279651367335293e-05,
+      "loss": 3.8602,
+      "step": 2272256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2788127725842412e-05,
+      "loss": 3.8588,
+      "step": 2272768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2779758157135624e-05,
+      "loss": 3.8587,
+      "step": 2273280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2771388588428835e-05,
+      "loss": 3.8514,
+      "step": 2273792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2763002640918315e-05,
+      "loss": 3.8567,
+      "step": 2274304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2754616693407795e-05,
+      "loss": 3.8598,
+      "step": 2274816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2746230745897275e-05,
+      "loss": 3.8538,
+      "step": 2275328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2737844798386753e-05,
+      "loss": 3.8602,
+      "step": 2275840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2729458850876233e-05,
+      "loss": 3.8508,
+      "step": 2276352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2721072903365713e-05,
+      "loss": 3.858,
+      "step": 2276864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2712686955855193e-05,
+      "loss": 3.8472,
+      "step": 2277376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2704301008344675e-05,
+      "loss": 3.8528,
+      "step": 2277888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2695915060834155e-05,
+      "loss": 3.8463,
+      "step": 2278400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2687545492127365e-05,
+      "loss": 3.8516,
+      "step": 2278912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2679159544616845e-05,
+      "loss": 3.8693,
+      "step": 2279424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2670773597106325e-05,
+      "loss": 3.8535,
+      "step": 2279936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2662387649595805e-05,
+      "loss": 3.8692,
+      "step": 2280448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2654001702085283e-05,
+      "loss": 3.8413,
+      "step": 2280960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2645615754574763e-05,
+      "loss": 3.8646,
+      "step": 2281472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2637229807064243e-05,
+      "loss": 3.8659,
+      "step": 2281984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2628843859553723e-05,
+      "loss": 3.8613,
+      "step": 2282496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2620474290846934e-05,
+      "loss": 3.861,
+      "step": 2283008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2612104722140147e-05,
+      "loss": 3.8494,
+      "step": 2283520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2603718774629628e-05,
+      "loss": 3.8613,
+      "step": 2284032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2595332827119108e-05,
+      "loss": 3.8732,
+      "step": 2284544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2586946879608585e-05,
+      "loss": 3.8461,
+      "step": 2285056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2578560932098066e-05,
+      "loss": 3.8511,
+      "step": 2285568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2570174984587546e-05,
+      "loss": 3.862,
+      "step": 2286080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2561789037077026e-05,
+      "loss": 3.8552,
+      "step": 2286592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2553403089566506e-05,
+      "loss": 3.8524,
+      "step": 2287104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.254504989966345e-05,
+      "loss": 3.8583,
+      "step": 2287616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.253666395215293e-05,
+      "loss": 3.8493,
+      "step": 2288128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2528278004642408e-05,
+      "loss": 3.8654,
+      "step": 2288640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2519892057131888e-05,
+      "loss": 3.852,
+      "step": 2289152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.976588249206543,
+      "eval_runtime": 317.4264,
+      "eval_samples_per_second": 1202.14,
+      "eval_steps_per_second": 37.568,
+      "step": 2289600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2511506109621368e-05,
+      "loss": 3.8566,
+      "step": 2289664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2503120162110848e-05,
+      "loss": 3.8513,
+      "step": 2290176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.249473421460033e-05,
+      "loss": 3.8619,
+      "step": 2290688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2486348267089808e-05,
+      "loss": 3.8668,
+      "step": 2291200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2477962319579288e-05,
+      "loss": 3.8589,
+      "step": 2291712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2469576372068767e-05,
+      "loss": 3.8565,
+      "step": 2292224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2461190424558247e-05,
+      "loss": 3.8572,
+      "step": 2292736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2452804477047727e-05,
+      "loss": 3.8579,
+      "step": 2293248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2444418529537207e-05,
+      "loss": 3.8479,
+      "step": 2293760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2436032582026689e-05,
+      "loss": 3.8694,
+      "step": 2294272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2427646634516167e-05,
+      "loss": 3.8643,
+      "step": 2294784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2419260687005647e-05,
+      "loss": 3.8633,
+      "step": 2295296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.241089111829886e-05,
+      "loss": 3.8724,
+      "step": 2295808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2402505170788338e-05,
+      "loss": 3.8368,
+      "step": 2296320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2394119223277818e-05,
+      "loss": 3.8639,
+      "step": 2296832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2385733275767298e-05,
+      "loss": 3.8465,
+      "step": 2297344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.237734732825678e-05,
+      "loss": 3.8572,
+      "step": 2297856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2368977759549989e-05,
+      "loss": 3.849,
+      "step": 2298368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.236059181203947e-05,
+      "loss": 3.8531,
+      "step": 2298880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.235220586452895e-05,
+      "loss": 3.8477,
+      "step": 2299392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2343819917018428e-05,
+      "loss": 3.8761,
+      "step": 2299904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2335433969507908e-05,
+      "loss": 3.8619,
+      "step": 2300416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.232704802199739e-05,
+      "loss": 3.857,
+      "step": 2300928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.231866207448687e-05,
+      "loss": 3.8538,
+      "step": 2301440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2310276126976348e-05,
+      "loss": 3.8613,
+      "step": 2301952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.230190655826956e-05,
+      "loss": 3.8484,
+      "step": 2302464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.229352061075904e-05,
+      "loss": 3.8464,
+      "step": 2302976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2285134663248519e-05,
+      "loss": 3.8583,
+      "step": 2303488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2276748715738e-05,
+      "loss": 3.8524,
+      "step": 2304000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.226836276822748e-05,
+      "loss": 3.8493,
+      "step": 2304512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.225997682071696e-05,
+      "loss": 3.8547,
+      "step": 2305024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2251590873206439e-05,
+      "loss": 3.8552,
+      "step": 2305536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.224320492569592e-05,
+      "loss": 3.8592,
+      "step": 2306048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2234835356989131e-05,
+      "loss": 3.8554,
+      "step": 2306560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.222644940947861e-05,
+      "loss": 3.8532,
+      "step": 2307072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2218063461968091e-05,
+      "loss": 3.8662,
+      "step": 2307584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2209693893261302e-05,
+      "loss": 3.8611,
+      "step": 2308096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2201307945750782e-05,
+      "loss": 3.8363,
+      "step": 2308608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2192921998240262e-05,
+      "loss": 3.8525,
+      "step": 2309120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2184536050729742e-05,
+      "loss": 3.8476,
+      "step": 2309632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2176150103219222e-05,
+      "loss": 3.8485,
+      "step": 2310144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2167780534512434e-05,
+      "loss": 3.8575,
+      "step": 2310656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2159394587001913e-05,
+      "loss": 3.8587,
+      "step": 2311168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2151008639491393e-05,
+      "loss": 3.8492,
+      "step": 2311680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2142622691980874e-05,
+      "loss": 3.8567,
+      "step": 2312192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2134253123274083e-05,
+      "loss": 3.8606,
+      "step": 2312704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2125867175763563e-05,
+      "loss": 3.8583,
+      "step": 2313216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2117481228253045e-05,
+      "loss": 3.8588,
+      "step": 2313728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2109095280742523e-05,
+      "loss": 3.8347,
+      "step": 2314240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2100709333232003e-05,
+      "loss": 3.8581,
+      "step": 2314752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2092339764525216e-05,
+      "loss": 3.8594,
+      "step": 2315264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2083953817014696e-05,
+      "loss": 3.8541,
+      "step": 2315776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2075567869504174e-05,
+      "loss": 3.8467,
+      "step": 2316288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2067181921993655e-05,
+      "loss": 3.8531,
+      "step": 2316800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2058812353286866e-05,
+      "loss": 3.8396,
+      "step": 2317312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2050426405776346e-05,
+      "loss": 3.8489,
+      "step": 2317824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2042040458265826e-05,
+      "loss": 3.8658,
+      "step": 2318336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2033654510755306e-05,
+      "loss": 3.8479,
+      "step": 2318848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2025268563244786e-05,
+      "loss": 3.8646,
+      "step": 2319360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2016882615734264e-05,
+      "loss": 3.8596,
+      "step": 2319872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2008496668223746e-05,
+      "loss": 3.8338,
+      "step": 2320384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2000127099516957e-05,
+      "loss": 3.8603,
+      "step": 2320896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1991741152006437e-05,
+      "loss": 3.8423,
+      "step": 2321408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1983355204495917e-05,
+      "loss": 3.8411,
+      "step": 2321920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1974969256985397e-05,
+      "loss": 3.8446,
+      "step": 2322432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1966599688278609e-05,
+      "loss": 3.8549,
+      "step": 2322944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1958213740768087e-05,
+      "loss": 3.8487,
+      "step": 2323456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1949827793257567e-05,
+      "loss": 3.8489,
+      "step": 2323968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1941441845747047e-05,
+      "loss": 3.8406,
+      "step": 2324480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1933055898236529e-05,
+      "loss": 3.8544,
+      "step": 2324992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1924669950726007e-05,
+      "loss": 3.8611,
+      "step": 2325504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1916300382019218e-05,
+      "loss": 3.8586,
+      "step": 2326016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.19079144345087e-05,
+      "loss": 3.8557,
+      "step": 2326528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1899528486998178e-05,
+      "loss": 3.8582,
+      "step": 2327040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1891142539487658e-05,
+      "loss": 3.8637,
+      "step": 2327552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1882756591977138e-05,
+      "loss": 3.8564,
+      "step": 2328064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.187438702327035e-05,
+      "loss": 3.8542,
+      "step": 2328576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1866001075759829e-05,
+      "loss": 3.8552,
+      "step": 2329088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.185761512824931e-05,
+      "loss": 3.8529,
+      "step": 2329600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.184922918073879e-05,
+      "loss": 3.8561,
+      "step": 2330112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1840843233228268e-05,
+      "loss": 3.8499,
+      "step": 2330624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1832457285717748e-05,
+      "loss": 3.8586,
+      "step": 2331136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.182407133820723e-05,
+      "loss": 3.8734,
+      "step": 2331648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.181568539069671e-05,
+      "loss": 3.8561,
+      "step": 2332160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.180731582198992e-05,
+      "loss": 3.8351,
+      "step": 2332672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1798946253283132e-05,
+      "loss": 3.849,
+      "step": 2333184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1790560305772612e-05,
+      "loss": 3.8574,
+      "step": 2333696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1782174358262091e-05,
+      "loss": 3.8532,
+      "step": 2334208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1773788410751571e-05,
+      "loss": 3.8545,
+      "step": 2334720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1765418842044782e-05,
+      "loss": 3.8575,
+      "step": 2335232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1757032894534264e-05,
+      "loss": 3.8444,
+      "step": 2335744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1748646947023742e-05,
+      "loss": 3.8503,
+      "step": 2336256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1740260999513222e-05,
+      "loss": 3.8506,
+      "step": 2336768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1731875052002702e-05,
+      "loss": 3.8604,
+      "step": 2337280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1723505483295915e-05,
+      "loss": 3.8554,
+      "step": 2337792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1715135914589125e-05,
+      "loss": 3.866,
+      "step": 2338304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1706749967078605e-05,
+      "loss": 3.8416,
+      "step": 2338816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1698364019568085e-05,
+      "loss": 3.8521,
+      "step": 2339328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1689978072057565e-05,
+      "loss": 3.8566,
+      "step": 2339840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1681592124547045e-05,
+      "loss": 3.8526,
+      "step": 2340352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1673206177036525e-05,
+      "loss": 3.8656,
+      "step": 2340864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1664820229526005e-05,
+      "loss": 3.8451,
+      "step": 2341376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1656434282015483e-05,
+      "loss": 3.8494,
+      "step": 2341888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1648048334504965e-05,
+      "loss": 3.8436,
+      "step": 2342400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1639662386994445e-05,
+      "loss": 3.8572,
+      "step": 2342912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1631276439483923e-05,
+      "loss": 3.844,
+      "step": 2343424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1622906870777136e-05,
+      "loss": 3.8366,
+      "step": 2343936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1614520923266616e-05,
+      "loss": 3.8492,
+      "step": 2344448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1606151354559826e-05,
+      "loss": 3.8533,
+      "step": 2344960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1597765407049306e-05,
+      "loss": 3.8516,
+      "step": 2345472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1589379459538786e-05,
+      "loss": 3.8543,
+      "step": 2345984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1580993512028266e-05,
+      "loss": 3.8439,
+      "step": 2346496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1572607564517746e-05,
+      "loss": 3.8574,
+      "step": 2347008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1564221617007226e-05,
+      "loss": 3.8464,
+      "step": 2347520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1555835669496706e-05,
+      "loss": 3.8531,
+      "step": 2348032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1547449721986186e-05,
+      "loss": 3.8584,
+      "step": 2348544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1539063774475666e-05,
+      "loss": 3.8598,
+      "step": 2349056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1530694205768877e-05,
+      "loss": 3.8538,
+      "step": 2349568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1522308258258357e-05,
+      "loss": 3.847,
+      "step": 2350080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1513922310747837e-05,
+      "loss": 3.8537,
+      "step": 2350592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1505536363237317e-05,
+      "loss": 3.8554,
+      "step": 2351104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1497150415726797e-05,
+      "loss": 3.846,
+      "step": 2351616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1488764468216277e-05,
+      "loss": 3.8596,
+      "step": 2352128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1480378520705757e-05,
+      "loss": 3.849,
+      "step": 2352640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1471992573195237e-05,
+      "loss": 3.8516,
+      "step": 2353152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1463606625684716e-05,
+      "loss": 3.845,
+      "step": 2353664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1455237056977927e-05,
+      "loss": 3.8496,
+      "step": 2354176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1446851109467407e-05,
+      "loss": 3.8451,
+      "step": 2354688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.143848154076062e-05,
+      "loss": 3.8457,
+      "step": 2355200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.14300955932501e-05,
+      "loss": 3.8636,
+      "step": 2355712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1421709645739578e-05,
+      "loss": 3.8525,
+      "step": 2356224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1413323698229058e-05,
+      "loss": 3.8618,
+      "step": 2356736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.140495412952227e-05,
+      "loss": 3.8422,
+      "step": 2357248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.139656818201175e-05,
+      "loss": 3.8575,
+      "step": 2357760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.138818223450123e-05,
+      "loss": 3.8628,
+      "step": 2358272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.137979628699071e-05,
+      "loss": 3.8605,
+      "step": 2358784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1371426718283921e-05,
+      "loss": 3.8521,
+      "step": 2359296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1363040770773401e-05,
+      "loss": 3.8489,
+      "step": 2359808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1354654823262881e-05,
+      "loss": 3.8592,
+      "step": 2360320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1346268875752361e-05,
+      "loss": 3.8704,
+      "step": 2360832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.133788292824184e-05,
+      "loss": 3.8474,
+      "step": 2361344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.132949698073132e-05,
+      "loss": 3.8453,
+      "step": 2361856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.13211110332208e-05,
+      "loss": 3.8559,
+      "step": 2362368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.131272508571028e-05,
+      "loss": 3.8488,
+      "step": 2362880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.130433913819976e-05,
+      "loss": 3.8486,
+      "step": 2363392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1295969569492971e-05,
+      "loss": 3.8543,
+      "step": 2363904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1287583621982451e-05,
+      "loss": 3.845,
+      "step": 2364416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1279197674471931e-05,
+      "loss": 3.8636,
+      "step": 2364928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1270811726961411e-05,
+      "loss": 3.845,
+      "step": 2365440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9753386974334717,
+      "eval_runtime": 294.1115,
+      "eval_samples_per_second": 1297.437,
+      "eval_steps_per_second": 40.546,
+      "step": 2365920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1262442158254622e-05,
+      "loss": 3.8473,
+      "step": 2365952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1254072589547835e-05,
+      "loss": 3.8519,
+      "step": 2366464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1245686642037315e-05,
+      "loss": 3.8575,
+      "step": 2366976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1237300694526793e-05,
+      "loss": 3.8619,
+      "step": 2367488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1228914747016274e-05,
+      "loss": 3.856,
+      "step": 2368000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1220528799505754e-05,
+      "loss": 3.856,
+      "step": 2368512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1212142851995233e-05,
+      "loss": 3.8533,
+      "step": 2369024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1203773283288445e-05,
+      "loss": 3.854,
+      "step": 2369536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1195387335777925e-05,
+      "loss": 3.8498,
+      "step": 2370048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1187001388267405e-05,
+      "loss": 3.8632,
+      "step": 2370560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1178615440756885e-05,
+      "loss": 3.861,
+      "step": 2371072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1170245872050096e-05,
+      "loss": 3.86,
+      "step": 2371584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1161876303343308e-05,
+      "loss": 3.8691,
+      "step": 2372096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1153490355832788e-05,
+      "loss": 3.8375,
+      "step": 2372608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1145104408322267e-05,
+      "loss": 3.8597,
+      "step": 2373120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1136718460811746e-05,
+      "loss": 3.8427,
+      "step": 2373632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1128332513301228e-05,
+      "loss": 3.8557,
+      "step": 2374144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1119946565790706e-05,
+      "loss": 3.8451,
+      "step": 2374656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1111560618280186e-05,
+      "loss": 3.8512,
+      "step": 2375168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1103174670769666e-05,
+      "loss": 3.8477,
+      "step": 2375680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1094788723259146e-05,
+      "loss": 3.8692,
+      "step": 2376192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1086402775748626e-05,
+      "loss": 3.8579,
+      "step": 2376704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1078016828238106e-05,
+      "loss": 3.8514,
+      "step": 2377216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1069630880727588e-05,
+      "loss": 3.8503,
+      "step": 2377728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1061261312020797e-05,
+      "loss": 3.8594,
+      "step": 2378240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1052875364510277e-05,
+      "loss": 3.8445,
+      "step": 2378752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1044489416999759e-05,
+      "loss": 3.8423,
+      "step": 2379264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1036103469489237e-05,
+      "loss": 3.856,
+      "step": 2379776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1027717521978717e-05,
+      "loss": 3.8446,
+      "step": 2380288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.101934795327193e-05,
+      "loss": 3.846,
+      "step": 2380800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.101096200576141e-05,
+      "loss": 3.853,
+      "step": 2381312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1002576058250887e-05,
+      "loss": 3.8505,
+      "step": 2381824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0994190110740369e-05,
+      "loss": 3.8575,
+      "step": 2382336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0985804163229849e-05,
+      "loss": 3.8485,
+      "step": 2382848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0977418215719327e-05,
+      "loss": 3.8534,
+      "step": 2383360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0969032268208807e-05,
+      "loss": 3.8619,
+      "step": 2383872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0960646320698289e-05,
+      "loss": 3.8586,
+      "step": 2384384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.09522767519915e-05,
+      "loss": 3.83,
+      "step": 2384896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0943890804480978e-05,
+      "loss": 3.8505,
+      "step": 2385408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.093550485697046e-05,
+      "loss": 3.8383,
+      "step": 2385920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.092713528826367e-05,
+      "loss": 3.8465,
+      "step": 2386432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.091874934075315e-05,
+      "loss": 3.8515,
+      "step": 2386944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.091036339324263e-05,
+      "loss": 3.8573,
+      "step": 2387456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.090197744573211e-05,
+      "loss": 3.849,
+      "step": 2387968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.089359149822159e-05,
+      "loss": 3.8507,
+      "step": 2388480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.088520555071107e-05,
+      "loss": 3.8551,
+      "step": 2388992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.087681960320055e-05,
+      "loss": 3.8548,
+      "step": 2389504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.086843365569003e-05,
+      "loss": 3.8518,
+      "step": 2390016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0860064086983241e-05,
+      "loss": 3.8333,
+      "step": 2390528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.085167813947272e-05,
+      "loss": 3.8537,
+      "step": 2391040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0843308570765932e-05,
+      "loss": 3.8586,
+      "step": 2391552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0834922623255413e-05,
+      "loss": 3.8516,
+      "step": 2392064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0826536675744892e-05,
+      "loss": 3.8418,
+      "step": 2392576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0818150728234371e-05,
+      "loss": 3.8472,
+      "step": 2393088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0809764780723851e-05,
+      "loss": 3.8443,
+      "step": 2393600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0801378833213331e-05,
+      "loss": 3.8371,
+      "step": 2394112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0793009264506542e-05,
+      "loss": 3.863,
+      "step": 2394624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0784623316996024e-05,
+      "loss": 3.8442,
+      "step": 2395136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0776237369485504e-05,
+      "loss": 3.8587,
+      "step": 2395648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0767851421974982e-05,
+      "loss": 3.8583,
+      "step": 2396160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0759465474464462e-05,
+      "loss": 3.8313,
+      "step": 2396672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0751079526953944e-05,
+      "loss": 3.852,
+      "step": 2397184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0742709958247154e-05,
+      "loss": 3.8452,
+      "step": 2397696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0734340389540365e-05,
+      "loss": 3.8412,
+      "step": 2398208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0725954442029845e-05,
+      "loss": 3.8364,
+      "step": 2398720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0717568494519325e-05,
+      "loss": 3.8553,
+      "step": 2399232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0709182547008805e-05,
+      "loss": 3.8505,
+      "step": 2399744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0700796599498285e-05,
+      "loss": 3.8397,
+      "step": 2400256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0692410651987765e-05,
+      "loss": 3.8396,
+      "step": 2400768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0684024704477245e-05,
+      "loss": 3.8529,
+      "step": 2401280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0675655135770456e-05,
+      "loss": 3.8565,
+      "step": 2401792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0667269188259936e-05,
+      "loss": 3.8543,
+      "step": 2402304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0658883240749416e-05,
+      "loss": 3.8534,
+      "step": 2402816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0650497293238896e-05,
+      "loss": 3.8534,
+      "step": 2403328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0642127724532106e-05,
+      "loss": 3.8592,
+      "step": 2403840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0633741777021586e-05,
+      "loss": 3.8565,
+      "step": 2404352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0625355829511068e-05,
+      "loss": 3.8527,
+      "step": 2404864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0616969882000546e-05,
+      "loss": 3.8523,
+      "step": 2405376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0608583934490026e-05,
+      "loss": 3.8504,
+      "step": 2405888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0600197986979506e-05,
+      "loss": 3.8558,
+      "step": 2406400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0591812039468986e-05,
+      "loss": 3.8447,
+      "step": 2406912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0583426091958466e-05,
+      "loss": 3.8562,
+      "step": 2407424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0575040144447946e-05,
+      "loss": 3.863,
+      "step": 2407936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0566670575741159e-05,
+      "loss": 3.8554,
+      "step": 2408448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0558284628230637e-05,
+      "loss": 3.8325,
+      "step": 2408960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0549898680720117e-05,
+      "loss": 3.8439,
+      "step": 2409472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.054152911201333e-05,
+      "loss": 3.853,
+      "step": 2409984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.053314316450281e-05,
+      "loss": 3.8506,
+      "step": 2410496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.052475721699229e-05,
+      "loss": 3.8485,
+      "step": 2411008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0516371269481769e-05,
+      "loss": 3.8571,
+      "step": 2411520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.050800170077498e-05,
+      "loss": 3.8467,
+      "step": 2412032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.049961575326446e-05,
+      "loss": 3.841,
+      "step": 2412544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.049122980575394e-05,
+      "loss": 3.8494,
+      "step": 2413056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.048284385824342e-05,
+      "loss": 3.857,
+      "step": 2413568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.04744579107329e-05,
+      "loss": 3.8508,
+      "step": 2414080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.046607196322238e-05,
+      "loss": 3.8614,
+      "step": 2414592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.045768601571186e-05,
+      "loss": 3.8429,
+      "step": 2415104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.044930006820134e-05,
+      "loss": 3.8508,
+      "step": 2415616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0440914120690818e-05,
+      "loss": 3.8542,
+      "step": 2416128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.04325281731803e-05,
+      "loss": 3.8473,
+      "step": 2416640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.042414222566978e-05,
+      "loss": 3.8585,
+      "step": 2417152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.041577265696299e-05,
+      "loss": 3.8457,
+      "step": 2417664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.040738670945247e-05,
+      "loss": 3.8448,
+      "step": 2418176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0399017140745681e-05,
+      "loss": 3.8419,
+      "step": 2418688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0390631193235163e-05,
+      "loss": 3.8551,
+      "step": 2419200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0382245245724641e-05,
+      "loss": 3.8365,
+      "step": 2419712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0373859298214121e-05,
+      "loss": 3.8334,
+      "step": 2420224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.03654733507036e-05,
+      "loss": 3.8507,
+      "step": 2420736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0357103781996813e-05,
+      "loss": 3.8517,
+      "step": 2421248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0348717834486292e-05,
+      "loss": 3.8455,
+      "step": 2421760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0340331886975772e-05,
+      "loss": 3.8508,
+      "step": 2422272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0331945939465253e-05,
+      "loss": 3.8421,
+      "step": 2422784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0323559991954731e-05,
+      "loss": 3.8554,
+      "step": 2423296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0315174044444211e-05,
+      "loss": 3.8383,
+      "step": 2423808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0306804475737424e-05,
+      "loss": 3.8506,
+      "step": 2424320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0298418528226904e-05,
+      "loss": 3.854,
+      "step": 2424832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0290032580716382e-05,
+      "loss": 3.8578,
+      "step": 2425344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0281646633205864e-05,
+      "loss": 3.8493,
+      "step": 2425856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0273260685695344e-05,
+      "loss": 3.8436,
+      "step": 2426368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0264874738184822e-05,
+      "loss": 3.8503,
+      "step": 2426880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0256488790674302e-05,
+      "loss": 3.8501,
+      "step": 2427392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0248102843163784e-05,
+      "loss": 3.8416,
+      "step": 2427904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0239716895653262e-05,
+      "loss": 3.8559,
+      "step": 2428416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0231330948142742e-05,
+      "loss": 3.846,
+      "step": 2428928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0222945000632222e-05,
+      "loss": 3.8472,
+      "step": 2429440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0214559053121703e-05,
+      "loss": 3.8431,
+      "step": 2429952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0206205863218645e-05,
+      "loss": 3.8449,
+      "step": 2430464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0197819915708125e-05,
+      "loss": 3.8489,
+      "step": 2430976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0189433968197605e-05,
+      "loss": 3.8434,
+      "step": 2431488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0181064399490817e-05,
+      "loss": 3.8523,
+      "step": 2432000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0172678451980296e-05,
+      "loss": 3.8566,
+      "step": 2432512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0164292504469776e-05,
+      "loss": 3.8583,
+      "step": 2433024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0155906556959256e-05,
+      "loss": 3.8403,
+      "step": 2433536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0147520609448736e-05,
+      "loss": 3.8543,
+      "step": 2434048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0139134661938216e-05,
+      "loss": 3.8573,
+      "step": 2434560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0130748714427695e-05,
+      "loss": 3.8555,
+      "step": 2435072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0122362766917175e-05,
+      "loss": 3.8482,
+      "step": 2435584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0113976819406655e-05,
+      "loss": 3.8457,
+      "step": 2436096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0105607250699866e-05,
+      "loss": 3.8538,
+      "step": 2436608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0097221303189348e-05,
+      "loss": 3.8669,
+      "step": 2437120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0088835355678826e-05,
+      "loss": 3.8454,
+      "step": 2437632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0080449408168306e-05,
+      "loss": 3.8403,
+      "step": 2438144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0072079839461519e-05,
+      "loss": 3.8514,
+      "step": 2438656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0063693891950998e-05,
+      "loss": 3.8433,
+      "step": 2439168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0055307944440477e-05,
+      "loss": 3.847,
+      "step": 2439680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0046921996929957e-05,
+      "loss": 3.8546,
+      "step": 2440192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.003855242822317e-05,
+      "loss": 3.8409,
+      "step": 2440704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0030166480712649e-05,
+      "loss": 3.8582,
+      "step": 2441216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0021780533202129e-05,
+      "loss": 3.8429,
+      "step": 2441728
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.0013394585691609e-05,
+      "loss": 3.8457,
+      "step": 2442240
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9743807315826416,
+      "eval_runtime": 293.4406,
+      "eval_samples_per_second": 1300.403,
+      "eval_steps_per_second": 40.639,
+      "step": 2442240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0005008638181089e-05,
+      "loss": 3.8434,
+      "step": 2442752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.996622690670567e-06,
+      "loss": 3.8504,
+      "step": 2443264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.988236743160049e-06,
+      "loss": 3.8585,
+      "step": 2443776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.979850795649529e-06,
+      "loss": 3.8526,
+      "step": 2444288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.971464848139007e-06,
+      "loss": 3.8533,
+      "step": 2444800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.963078900628487e-06,
+      "loss": 3.8531,
+      "step": 2445312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.954692953117969e-06,
+      "loss": 3.8494,
+      "step": 2445824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.946307005607449e-06,
+      "loss": 3.8383,
+      "step": 2446336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.937921058096927e-06,
+      "loss": 3.8687,
+      "step": 2446848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.929535110586407e-06,
+      "loss": 3.8525,
+      "step": 2447360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.921149163075889e-06,
+      "loss": 3.8613,
+      "step": 2447872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.912779594369098e-06,
+      "loss": 3.864,
+      "step": 2448384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.90439364685858e-06,
+      "loss": 3.8369,
+      "step": 2448896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.89600769934806e-06,
+      "loss": 3.8509,
+      "step": 2449408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.88762175183754e-06,
+      "loss": 3.8418,
+      "step": 2449920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.879235804327017e-06,
+      "loss": 3.847,
+      "step": 2450432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.87086623562023e-06,
+      "loss": 3.8487,
+      "step": 2450944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.86248028810971e-06,
+      "loss": 3.8438,
+      "step": 2451456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.854094340599188e-06,
+      "loss": 3.8434,
+      "step": 2451968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.84570839308867e-06,
+      "loss": 3.8635,
+      "step": 2452480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.83732244557815e-06,
+      "loss": 3.8549,
+      "step": 2452992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.82893649806763e-06,
+      "loss": 3.8478,
+      "step": 2453504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.820550550557108e-06,
+      "loss": 3.8459,
+      "step": 2454016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.81216460304659e-06,
+      "loss": 3.859,
+      "step": 2454528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.8037950343398e-06,
+      "loss": 3.8418,
+      "step": 2455040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.79540908682928e-06,
+      "loss": 3.8377,
+      "step": 2455552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.78702313931876e-06,
+      "loss": 3.8515,
+      "step": 2456064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.778653570611971e-06,
+      "loss": 3.8449,
+      "step": 2456576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.770267623101453e-06,
+      "loss": 3.8433,
+      "step": 2457088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.761881675590931e-06,
+      "loss": 3.8459,
+      "step": 2457600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.753495728080411e-06,
+      "loss": 3.8472,
+      "step": 2458112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.745109780569891e-06,
+      "loss": 3.8537,
+      "step": 2458624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.736723833059371e-06,
+      "loss": 3.8446,
+      "step": 2459136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.728337885548851e-06,
+      "loss": 3.8508,
+      "step": 2459648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.71995193803833e-06,
+      "loss": 3.8594,
+      "step": 2460160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.71156599052781e-06,
+      "loss": 3.8522,
+      "step": 2460672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.70318004301729e-06,
+      "loss": 3.8307,
+      "step": 2461184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.69479409550677e-06,
+      "loss": 3.846,
+      "step": 2461696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.68640814799625e-06,
+      "loss": 3.8336,
+      "step": 2462208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.678038579289461e-06,
+      "loss": 3.848,
+      "step": 2462720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.669652631778941e-06,
+      "loss": 3.8456,
+      "step": 2463232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.661283063072154e-06,
+      "loss": 3.853,
+      "step": 2463744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.652913494365365e-06,
+      "loss": 3.8428,
+      "step": 2464256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.644543925658575e-06,
+      "loss": 3.8511,
+      "step": 2464768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.636157978148055e-06,
+      "loss": 3.8493,
+      "step": 2465280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.627772030637535e-06,
+      "loss": 3.8515,
+      "step": 2465792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.619386083127015e-06,
+      "loss": 3.8472,
+      "step": 2466304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.611000135616495e-06,
+      "loss": 3.8319,
+      "step": 2466816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.602614188105975e-06,
+      "loss": 3.8488,
+      "step": 2467328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.594228240595455e-06,
+      "loss": 3.8583,
+      "step": 2467840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.585842293084935e-06,
+      "loss": 3.8457,
+      "step": 2468352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.577456345574415e-06,
+      "loss": 3.8391,
+      "step": 2468864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.569070398063895e-06,
+      "loss": 3.8427,
+      "step": 2469376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.560684450553373e-06,
+      "loss": 3.8427,
+      "step": 2469888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.552298503042855e-06,
+      "loss": 3.8329,
+      "step": 2470400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.543928934336066e-06,
+      "loss": 3.8598,
+      "step": 2470912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.535542986825546e-06,
+      "loss": 3.8438,
+      "step": 2471424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.527157039315026e-06,
+      "loss": 3.8561,
+      "step": 2471936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.518771091804506e-06,
+      "loss": 3.8512,
+      "step": 2472448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.510401523097716e-06,
+      "loss": 3.8315,
+      "step": 2472960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.502015575587198e-06,
+      "loss": 3.8456,
+      "step": 2473472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.493629628076676e-06,
+      "loss": 3.8434,
+      "step": 2473984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.485243680566156e-06,
+      "loss": 3.8336,
+      "step": 2474496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.476857733055638e-06,
+      "loss": 3.8331,
+      "step": 2475008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.468471785545116e-06,
+      "loss": 3.8542,
+      "step": 2475520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.460085838034596e-06,
+      "loss": 3.8414,
+      "step": 2476032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.451699890524076e-06,
+      "loss": 3.8403,
+      "step": 2476544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.443330321817289e-06,
+      "loss": 3.8311,
+      "step": 2477056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.434944374306767e-06,
+      "loss": 3.8501,
+      "step": 2477568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.42657480559998e-06,
+      "loss": 3.8551,
+      "step": 2478080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.41818885808946e-06,
+      "loss": 3.8486,
+      "step": 2478592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.409802910578938e-06,
+      "loss": 3.8505,
+      "step": 2479104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.40141696306842e-06,
+      "loss": 3.8544,
+      "step": 2479616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.3930310155579e-06,
+      "loss": 3.8533,
+      "step": 2480128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.384645068047379e-06,
+      "loss": 3.8545,
+      "step": 2480640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.37627549934059e-06,
+      "loss": 3.8419,
+      "step": 2481152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.36788955183007e-06,
+      "loss": 3.8532,
+      "step": 2481664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.35950360431955e-06,
+      "loss": 3.8469,
+      "step": 2482176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.351117656809028e-06,
+      "loss": 3.8501,
+      "step": 2482688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.34274808810224e-06,
+      "loss": 3.8418,
+      "step": 2483200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.33436214059172e-06,
+      "loss": 3.8511,
+      "step": 2483712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.3259761930812e-06,
+      "loss": 3.8612,
+      "step": 2484224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.31759024557068e-06,
+      "loss": 3.8526,
+      "step": 2484736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.30920429806016e-06,
+      "loss": 3.8292,
+      "step": 2485248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.30081835054964e-06,
+      "loss": 3.8406,
+      "step": 2485760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.29243240303912e-06,
+      "loss": 3.8529,
+      "step": 2486272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.2840464555286e-06,
+      "loss": 3.8488,
+      "step": 2486784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.275676886821811e-06,
+      "loss": 3.8474,
+      "step": 2487296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.267307318115024e-06,
+      "loss": 3.8475,
+      "step": 2487808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.258921370604502e-06,
+      "loss": 3.8462,
+      "step": 2488320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.250535423093982e-06,
+      "loss": 3.8365,
+      "step": 2488832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.242149475583463e-06,
+      "loss": 3.8444,
+      "step": 2489344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.233763528072943e-06,
+      "loss": 3.8527,
+      "step": 2489856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.225377580562422e-06,
+      "loss": 3.8467,
+      "step": 2490368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.217008011855634e-06,
+      "loss": 3.8596,
+      "step": 2490880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.208622064345114e-06,
+      "loss": 3.8404,
+      "step": 2491392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.200236116834592e-06,
+      "loss": 3.8474,
+      "step": 2491904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.191850169324074e-06,
+      "loss": 3.8478,
+      "step": 2492416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.183496979421016e-06,
+      "loss": 3.8472,
+      "step": 2492928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.175111031910497e-06,
+      "loss": 3.853,
+      "step": 2493440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.166725084399976e-06,
+      "loss": 3.8428,
+      "step": 2493952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.158339136889455e-06,
+      "loss": 3.8398,
+      "step": 2494464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.149969568182668e-06,
+      "loss": 3.8411,
+      "step": 2494976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.141583620672148e-06,
+      "loss": 3.8511,
+      "step": 2495488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.133197673161626e-06,
+      "loss": 3.8349,
+      "step": 2496000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.124811725651108e-06,
+      "loss": 3.8308,
+      "step": 2496512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.116425778140588e-06,
+      "loss": 3.8419,
+      "step": 2497024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.108039830630066e-06,
+      "loss": 3.8506,
+      "step": 2497536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.099653883119546e-06,
+      "loss": 3.8397,
+      "step": 2498048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.091267935609028e-06,
+      "loss": 3.848,
+      "step": 2498560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.082881988098506e-06,
+      "loss": 3.838,
+      "step": 2499072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.074512419391717e-06,
+      "loss": 3.8501,
+      "step": 2499584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.066126471881198e-06,
+      "loss": 3.8329,
+      "step": 2500096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.057740524370678e-06,
+      "loss": 3.8476,
+      "step": 2500608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.049354576860157e-06,
+      "loss": 3.8515,
+      "step": 2501120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.040985008153369e-06,
+      "loss": 3.8543,
+      "step": 2501632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.03261543944658e-06,
+      "loss": 3.8428,
+      "step": 2502144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.024229491936061e-06,
+      "loss": 3.844,
+      "step": 2502656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.01584354442554e-06,
+      "loss": 3.8463,
+      "step": 2503168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.00745759691502e-06,
+      "loss": 3.8489,
+      "step": 2503680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.9990716494045e-06,
+      "loss": 3.8384,
+      "step": 2504192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.99068570189398e-06,
+      "loss": 3.8481,
+      "step": 2504704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.98229975438346e-06,
+      "loss": 3.8456,
+      "step": 2505216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.97391380687294e-06,
+      "loss": 3.8504,
+      "step": 2505728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.96552785936242e-06,
+      "loss": 3.8388,
+      "step": 2506240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.9571419118519e-06,
+      "loss": 3.837,
+      "step": 2506752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.94875596434138e-06,
+      "loss": 3.8414,
+      "step": 2507264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.94038639563459e-06,
+      "loss": 3.8473,
+      "step": 2507776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.93200044812407e-06,
+      "loss": 3.8418,
+      "step": 2508288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.92361450061355e-06,
+      "loss": 3.8564,
+      "step": 2508800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.915244931906763e-06,
+      "loss": 3.8512,
+      "step": 2509312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.906858984396243e-06,
+      "loss": 3.8372,
+      "step": 2509824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.89847303688572e-06,
+      "loss": 3.8483,
+      "step": 2510336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.8900870893752e-06,
+      "loss": 3.8562,
+      "step": 2510848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.881701141864682e-06,
+      "loss": 3.8536,
+      "step": 2511360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.87331519435416e-06,
+      "loss": 3.8457,
+      "step": 2511872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.864945625647373e-06,
+      "loss": 3.8429,
+      "step": 2512384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.856559678136853e-06,
+      "loss": 3.8531,
+      "step": 2512896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.848173730626333e-06,
+      "loss": 3.8582,
+      "step": 2513408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.839787783115811e-06,
+      "loss": 3.8431,
+      "step": 2513920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.831401835605293e-06,
+      "loss": 3.8402,
+      "step": 2514432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.823015888094773e-06,
+      "loss": 3.8452,
+      "step": 2514944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.814629940584251e-06,
+      "loss": 3.8442,
+      "step": 2515456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.806243993073731e-06,
+      "loss": 3.8397,
+      "step": 2515968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.797858045563213e-06,
+      "loss": 3.851,
+      "step": 2516480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.789488476856424e-06,
+      "loss": 3.8413,
+      "step": 2516992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.781102529345902e-06,
+      "loss": 3.8515,
+      "step": 2517504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.772716581835383e-06,
+      "loss": 3.8419,
+      "step": 2518016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.764330634324863e-06,
+      "loss": 3.8399,
+      "step": 2518528
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.973440647125244,
+      "eval_runtime": 291.3433,
+      "eval_samples_per_second": 1309.764,
+      "eval_steps_per_second": 40.931,
+      "step": 2518560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.755944686814342e-06,
+      "loss": 3.8427,
+      "step": 2519040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.747558739303822e-06,
+      "loss": 3.8514,
+      "step": 2519552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.739189170597034e-06,
+      "loss": 3.8546,
+      "step": 2520064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.730803223086514e-06,
+      "loss": 3.847,
+      "step": 2520576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.722417275575994e-06,
+      "loss": 3.8514,
+      "step": 2521088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.714031328065474e-06,
+      "loss": 3.8502,
+      "step": 2521600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.705645380554954e-06,
+      "loss": 3.8451,
+      "step": 2522112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.697259433044432e-06,
+      "loss": 3.8384,
+      "step": 2522624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.688889864337645e-06,
+      "loss": 3.8623,
+      "step": 2523136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.680503916827125e-06,
+      "loss": 3.8475,
+      "step": 2523648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.672117969316605e-06,
+      "loss": 3.8587,
+      "step": 2524160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.663732021806085e-06,
+      "loss": 3.8573,
+      "step": 2524672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.655362453099295e-06,
+      "loss": 3.8424,
+      "step": 2525184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.646976505588775e-06,
+      "loss": 3.8441,
+      "step": 2525696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.638590558078257e-06,
+      "loss": 3.8428,
+      "step": 2526208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.630204610567735e-06,
+      "loss": 3.8377,
+      "step": 2526720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.621818663057215e-06,
+      "loss": 3.846,
+      "step": 2527232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.613432715546695e-06,
+      "loss": 3.8422,
+      "step": 2527744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.605046768036175e-06,
+      "loss": 3.8377,
+      "step": 2528256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.596660820525655e-06,
+      "loss": 3.8652,
+      "step": 2528768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.588274873015135e-06,
+      "loss": 3.8508,
+      "step": 2529280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.579905304308346e-06,
+      "loss": 3.8472,
+      "step": 2529792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.571519356797826e-06,
+      "loss": 3.8447,
+      "step": 2530304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.563133409287306e-06,
+      "loss": 3.8572,
+      "step": 2530816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.554747461776787e-06,
+      "loss": 3.8338,
+      "step": 2531328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.546377893069996e-06,
+      "loss": 3.8356,
+      "step": 2531840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.537991945559478e-06,
+      "loss": 3.8433,
+      "step": 2532352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.529605998048958e-06,
+      "loss": 3.8457,
+      "step": 2532864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.521220050538436e-06,
+      "loss": 3.8396,
+      "step": 2533376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.512834103027916e-06,
+      "loss": 3.8431,
+      "step": 2533888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.504448155517398e-06,
+      "loss": 3.8468,
+      "step": 2534400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.496062208006878e-06,
+      "loss": 3.8515,
+      "step": 2534912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.487676260496356e-06,
+      "loss": 3.8391,
+      "step": 2535424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.479290312985836e-06,
+      "loss": 3.85,
+      "step": 2535936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.470904365475318e-06,
+      "loss": 3.8551,
+      "step": 2536448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.462518417964796e-06,
+      "loss": 3.8498,
+      "step": 2536960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.454132470454276e-06,
+      "loss": 3.8303,
+      "step": 2537472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.445762901747488e-06,
+      "loss": 3.84,
+      "step": 2537984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.437376954236968e-06,
+      "loss": 3.8353,
+      "step": 2538496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.428991006726447e-06,
+      "loss": 3.8436,
+      "step": 2539008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.420605059215927e-06,
+      "loss": 3.8391,
+      "step": 2539520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.412235490509139e-06,
+      "loss": 3.8485,
+      "step": 2540032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.403849542998617e-06,
+      "loss": 3.84,
+      "step": 2540544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.395463595488099e-06,
+      "loss": 3.847,
+      "step": 2541056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.387077647977579e-06,
+      "loss": 3.849,
+      "step": 2541568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.378724458074522e-06,
+      "loss": 3.8478,
+      "step": 2542080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.370338510564e-06,
+      "loss": 3.8454,
+      "step": 2542592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.36195256305348e-06,
+      "loss": 3.8263,
+      "step": 2543104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.35356661554296e-06,
+      "loss": 3.8474,
+      "step": 2543616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.345197046836173e-06,
+      "loss": 3.851,
+      "step": 2544128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.336811099325651e-06,
+      "loss": 3.8456,
+      "step": 2544640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.328425151815133e-06,
+      "loss": 3.8389,
+      "step": 2545152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.320039204304613e-06,
+      "loss": 3.8351,
+      "step": 2545664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.311653256794091e-06,
+      "loss": 3.8407,
+      "step": 2546176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.303267309283571e-06,
+      "loss": 3.8258,
+      "step": 2546688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.294881361773053e-06,
+      "loss": 3.857,
+      "step": 2547200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.286495414262533e-06,
+      "loss": 3.8399,
+      "step": 2547712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.278109466752011e-06,
+      "loss": 3.8543,
+      "step": 2548224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.269723519241491e-06,
+      "loss": 3.8503,
+      "step": 2548736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.261337571730973e-06,
+      "loss": 3.8279,
+      "step": 2549248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.25295162422045e-06,
+      "loss": 3.8421,
+      "step": 2549760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.244582055513663e-06,
+      "loss": 3.8411,
+      "step": 2550272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.236196108003143e-06,
+      "loss": 3.8295,
+      "step": 2550784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.227826539296354e-06,
+      "loss": 3.8299,
+      "step": 2551296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.219440591785834e-06,
+      "loss": 3.8501,
+      "step": 2551808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.211054644275314e-06,
+      "loss": 3.8404,
+      "step": 2552320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.202668696764794e-06,
+      "loss": 3.8379,
+      "step": 2552832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.194282749254272e-06,
+      "loss": 3.8274,
+      "step": 2553344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.185896801743754e-06,
+      "loss": 3.8481,
+      "step": 2553856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.177510854233234e-06,
+      "loss": 3.8492,
+      "step": 2554368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.169124906722714e-06,
+      "loss": 3.8462,
+      "step": 2554880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.160755338015924e-06,
+      "loss": 3.8451,
+      "step": 2555392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.152369390505404e-06,
+      "loss": 3.8557,
+      "step": 2555904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.143983442994884e-06,
+      "loss": 3.8491,
+      "step": 2556416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.135613874288097e-06,
+      "loss": 3.8539,
+      "step": 2556928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.127244305581306e-06,
+      "loss": 3.837,
+      "step": 2557440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.118858358070788e-06,
+      "loss": 3.8498,
+      "step": 2557952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.110472410560268e-06,
+      "loss": 3.8446,
+      "step": 2558464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.102086463049746e-06,
+      "loss": 3.8444,
+      "step": 2558976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.093700515539226e-06,
+      "loss": 3.8423,
+      "step": 2559488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.085314568028707e-06,
+      "loss": 3.8448,
+      "step": 2560000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.076928620518187e-06,
+      "loss": 3.86,
+      "step": 2560512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.068559051811398e-06,
+      "loss": 3.8527,
+      "step": 2561024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.060173104300878e-06,
+      "loss": 3.8299,
+      "step": 2561536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.051787156790358e-06,
+      "loss": 3.8367,
+      "step": 2562048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.043401209279836e-06,
+      "loss": 3.8448,
+      "step": 2562560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.035015261769318e-06,
+      "loss": 3.8465,
+      "step": 2563072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.026629314258798e-06,
+      "loss": 3.8466,
+      "step": 2563584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.018243366748278e-06,
+      "loss": 3.8449,
+      "step": 2564096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.009857419237756e-06,
+      "loss": 3.8419,
+      "step": 2564608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.001471471727238e-06,
+      "loss": 3.834,
+      "step": 2565120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.993085524216718e-06,
+      "loss": 3.8426,
+      "step": 2565632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.984699576706196e-06,
+      "loss": 3.8442,
+      "step": 2566144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.976313629195676e-06,
+      "loss": 3.8449,
+      "step": 2566656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.967944060488888e-06,
+      "loss": 3.8536,
+      "step": 2567168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.959558112978367e-06,
+      "loss": 3.8448,
+      "step": 2567680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.951172165467847e-06,
+      "loss": 3.8397,
+      "step": 2568192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.942786217957328e-06,
+      "loss": 3.847,
+      "step": 2568704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.93441664925054e-06,
+      "loss": 3.8458,
+      "step": 2569216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.926030701740019e-06,
+      "loss": 3.8512,
+      "step": 2569728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.917644754229499e-06,
+      "loss": 3.8385,
+      "step": 2570240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.909258806718979e-06,
+      "loss": 3.8362,
+      "step": 2570752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.900872859208457e-06,
+      "loss": 3.8386,
+      "step": 2571264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.892486911697939e-06,
+      "loss": 3.8476,
+      "step": 2571776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.884100964187419e-06,
+      "loss": 3.8328,
+      "step": 2572288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.87573139548063e-06,
+      "loss": 3.8262,
+      "step": 2572800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.86734544797011e-06,
+      "loss": 3.8413,
+      "step": 2573312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.85895950045959e-06,
+      "loss": 3.8467,
+      "step": 2573824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.85057355294907e-06,
+      "loss": 3.8367,
+      "step": 2574336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.84218760543855e-06,
+      "loss": 3.8425,
+      "step": 2574848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.83381803673176e-06,
+      "loss": 3.8379,
+      "step": 2575360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.82543208922124e-06,
+      "loss": 3.8464,
+      "step": 2575872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.81704614171072e-06,
+      "loss": 3.832,
+      "step": 2576384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.8086601942002e-06,
+      "loss": 3.8435,
+      "step": 2576896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.80027424668968e-06,
+      "loss": 3.8487,
+      "step": 2577408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.791904677982893e-06,
+      "loss": 3.8517,
+      "step": 2577920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.783518730472373e-06,
+      "loss": 3.8414,
+      "step": 2578432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.775149161765582e-06,
+      "loss": 3.8385,
+      "step": 2578944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.766763214255063e-06,
+      "loss": 3.8441,
+      "step": 2579456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.758377266744543e-06,
+      "loss": 3.8469,
+      "step": 2579968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.749991319234022e-06,
+      "loss": 3.8328,
+      "step": 2580480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.741605371723503e-06,
+      "loss": 3.8448,
+      "step": 2580992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.733219424212983e-06,
+      "loss": 3.8381,
+      "step": 2581504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.724833476702463e-06,
+      "loss": 3.8512,
+      "step": 2582016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.716447529191941e-06,
+      "loss": 3.8325,
+      "step": 2582528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.708061581681423e-06,
+      "loss": 3.8356,
+      "step": 2583040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.699675634170903e-06,
+      "loss": 3.8383,
+      "step": 2583552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.691306065464112e-06,
+      "loss": 3.8414,
+      "step": 2584064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.682920117953594e-06,
+      "loss": 3.8395,
+      "step": 2584576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.674534170443074e-06,
+      "loss": 3.8571,
+      "step": 2585088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.666148222932554e-06,
+      "loss": 3.8469,
+      "step": 2585600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.657762275422032e-06,
+      "loss": 3.8374,
+      "step": 2586112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.649376327911514e-06,
+      "loss": 3.8413,
+      "step": 2586624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.640990380400993e-06,
+      "loss": 3.8547,
+      "step": 2587136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.632620811694204e-06,
+      "loss": 3.8507,
+      "step": 2587648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.624234864183684e-06,
+      "loss": 3.8428,
+      "step": 2588160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.615848916673164e-06,
+      "loss": 3.8401,
+      "step": 2588672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.607462969162644e-06,
+      "loss": 3.8503,
+      "step": 2589184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.599093400455855e-06,
+      "loss": 3.8508,
+      "step": 2589696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.590707452945335e-06,
+      "loss": 3.8398,
+      "step": 2590208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.582321505434816e-06,
+      "loss": 3.8411,
+      "step": 2590720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.573935557924294e-06,
+      "loss": 3.8442,
+      "step": 2591232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.565565989217506e-06,
+      "loss": 3.8395,
+      "step": 2591744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.557180041706986e-06,
+      "loss": 3.8384,
+      "step": 2592256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.548794094196466e-06,
+      "loss": 3.8488,
+      "step": 2592768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.5404081466859454e-06,
+      "loss": 3.8394,
+      "step": 2593280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.532022199175426e-06,
+      "loss": 3.8477,
+      "step": 2593792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.523636251664906e-06,
+      "loss": 3.8401,
+      "step": 2594304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.515250304154385e-06,
+      "loss": 3.8385,
+      "step": 2594816
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.97273850440979,
+      "eval_runtime": 292.2564,
+      "eval_samples_per_second": 1305.672,
+      "eval_steps_per_second": 40.803,
+      "step": 2594880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.506864356643865e-06,
+      "loss": 3.8367,
+      "step": 2595328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.498478409133346e-06,
+      "loss": 3.8484,
+      "step": 2595840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.490092461622826e-06,
+      "loss": 3.8551,
+      "step": 2596352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.481706514112305e-06,
+      "loss": 3.8442,
+      "step": 2596864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.473320566601785e-06,
+      "loss": 3.8491,
+      "step": 2597376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.464934619091266e-06,
+      "loss": 3.8432,
+      "step": 2597888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.456548671580744e-06,
+      "loss": 3.8458,
+      "step": 2598400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.448162724070225e-06,
+      "loss": 3.8343,
+      "step": 2598912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.439776776559705e-06,
+      "loss": 3.8551,
+      "step": 2599424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.431390829049184e-06,
+      "loss": 3.8517,
+      "step": 2599936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.423004881538664e-06,
+      "loss": 3.8471,
+      "step": 2600448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.414618934028145e-06,
+      "loss": 3.8563,
+      "step": 2600960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.406249365321356e-06,
+      "loss": 3.8384,
+      "step": 2601472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.3978634178108355e-06,
+      "loss": 3.8405,
+      "step": 2601984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.3894774703003155e-06,
+      "loss": 3.8482,
+      "step": 2602496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.381091522789796e-06,
+      "loss": 3.8325,
+      "step": 2603008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.3727055752792745e-06,
+      "loss": 3.8408,
+      "step": 2603520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.364319627768755e-06,
+      "loss": 3.8368,
+      "step": 2604032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.355933680258235e-06,
+      "loss": 3.837,
+      "step": 2604544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.347547732747714e-06,
+      "loss": 3.8625,
+      "step": 2605056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.339161785237194e-06,
+      "loss": 3.8453,
+      "step": 2605568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.330792216530406e-06,
+      "loss": 3.8493,
+      "step": 2606080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.322406269019887e-06,
+      "loss": 3.8382,
+      "step": 2606592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.314020321509366e-06,
+      "loss": 3.8549,
+      "step": 2607104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.305634373998846e-06,
+      "loss": 3.8287,
+      "step": 2607616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.297248426488326e-06,
+      "loss": 3.8374,
+      "step": 2608128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.288862478977805e-06,
+      "loss": 3.8373,
+      "step": 2608640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.2804929102710166e-06,
+      "loss": 3.8446,
+      "step": 2609152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.272106962760497e-06,
+      "loss": 3.8378,
+      "step": 2609664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.263721015249977e-06,
+      "loss": 3.8369,
+      "step": 2610176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.2553350677394564e-06,
+      "loss": 3.846,
+      "step": 2610688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.246949120228936e-06,
+      "loss": 3.8483,
+      "step": 2611200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.238563172718417e-06,
+      "loss": 3.8317,
+      "step": 2611712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.2301772252078955e-06,
+      "loss": 3.8503,
+      "step": 2612224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.221791277697376e-06,
+      "loss": 3.8499,
+      "step": 2612736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.213421708990588e-06,
+      "loss": 3.8454,
+      "step": 2613248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.205035761480068e-06,
+      "loss": 3.8275,
+      "step": 2613760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.196649813969547e-06,
+      "loss": 3.8407,
+      "step": 2614272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.188263866459028e-06,
+      "loss": 3.8286,
+      "step": 2614784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.179894297752239e-06,
+      "loss": 3.8427,
+      "step": 2615296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.171508350241719e-06,
+      "loss": 3.839,
+      "step": 2615808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.1631224027311985e-06,
+      "loss": 3.8453,
+      "step": 2616320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.154736455220678e-06,
+      "loss": 3.8359,
+      "step": 2616832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.146383265317622e-06,
+      "loss": 3.8428,
+      "step": 2617344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.138013696610833e-06,
+      "loss": 3.8475,
+      "step": 2617856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.1296277491003125e-06,
+      "loss": 3.8462,
+      "step": 2618368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.1212418015897924e-06,
+      "loss": 3.8465,
+      "step": 2618880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.112855854079272e-06,
+      "loss": 3.8223,
+      "step": 2619392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.1044699065687515e-06,
+      "loss": 3.8395,
+      "step": 2619904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.096083959058232e-06,
+      "loss": 3.8528,
+      "step": 2620416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.087698011547712e-06,
+      "loss": 3.8395,
+      "step": 2620928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.079312064037193e-06,
+      "loss": 3.8369,
+      "step": 2621440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.070942495330403e-06,
+      "loss": 3.835,
+      "step": 2621952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.062556547819883e-06,
+      "loss": 3.8386,
+      "step": 2622464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.054170600309364e-06,
+      "loss": 3.8272,
+      "step": 2622976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.045784652798843e-06,
+      "loss": 3.8485,
+      "step": 2623488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.037398705288323e-06,
+      "loss": 3.8404,
+      "step": 2624000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.029012757777803e-06,
+      "loss": 3.8455,
+      "step": 2624512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.020626810267284e-06,
+      "loss": 3.8521,
+      "step": 2625024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.012240862756763e-06,
+      "loss": 3.8236,
+      "step": 2625536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.003871294049974e-06,
+      "loss": 3.8394,
+      "step": 2626048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.995501725343186e-06,
+      "loss": 3.8371,
+      "step": 2626560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.987115777832666e-06,
+      "loss": 3.8291,
+      "step": 2627072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.978729830322145e-06,
+      "loss": 3.8278,
+      "step": 2627584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.970343882811625e-06,
+      "loss": 3.8441,
+      "step": 2628096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.961957935301106e-06,
+      "loss": 3.8403,
+      "step": 2628608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.953588366594316e-06,
+      "loss": 3.8353,
+      "step": 2629120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.945202419083796e-06,
+      "loss": 3.8221,
+      "step": 2629632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.9368164715732765e-06,
+      "loss": 3.8445,
+      "step": 2630144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.9284305240627565e-06,
+      "loss": 3.848,
+      "step": 2630656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.9200445765522356e-06,
+      "loss": 3.8416,
+      "step": 2631168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.911675007845447e-06,
+      "loss": 3.8429,
+      "step": 2631680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.903289060334928e-06,
+      "loss": 3.8529,
+      "step": 2632192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.894903112824406e-06,
+      "loss": 3.8462,
+      "step": 2632704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.886517165313887e-06,
+      "loss": 3.8524,
+      "step": 2633216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.878147596607099e-06,
+      "loss": 3.831,
+      "step": 2633728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.869761649096579e-06,
+      "loss": 3.8481,
+      "step": 2634240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.861375701586058e-06,
+      "loss": 3.8451,
+      "step": 2634752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.852989754075538e-06,
+      "loss": 3.8393,
+      "step": 2635264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.8446038065650185e-06,
+      "loss": 3.8431,
+      "step": 2635776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.836217859054498e-06,
+      "loss": 3.8404,
+      "step": 2636288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.827831911543978e-06,
+      "loss": 3.8561,
+      "step": 2636800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.8194459640334575e-06,
+      "loss": 3.8459,
+      "step": 2637312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.811060016522937e-06,
+      "loss": 3.828,
+      "step": 2637824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.8026740690124175e-06,
+      "loss": 3.8326,
+      "step": 2638336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.794288121501897e-06,
+      "loss": 3.8412,
+      "step": 2638848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.785902173991377e-06,
+      "loss": 3.8462,
+      "step": 2639360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.777532605284588e-06,
+      "loss": 3.8446,
+      "step": 2639872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.7691630365778e-06,
+      "loss": 3.8392,
+      "step": 2640384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.76077708906728e-06,
+      "loss": 3.8381,
+      "step": 2640896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.7523911415567606e-06,
+      "loss": 3.8305,
+      "step": 2641408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.74400519404624e-06,
+      "loss": 3.8441,
+      "step": 2641920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.735635625339451e-06,
+      "loss": 3.8363,
+      "step": 2642432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.727249677828931e-06,
+      "loss": 3.8435,
+      "step": 2642944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.71886373031841e-06,
+      "loss": 3.852,
+      "step": 2643456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.71047778280789e-06,
+      "loss": 3.8447,
+      "step": 2643968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.702091835297371e-06,
+      "loss": 3.8345,
+      "step": 2644480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.693705887786851e-06,
+      "loss": 3.8432,
+      "step": 2644992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.68531994027633e-06,
+      "loss": 3.8413,
+      "step": 2645504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.67693399276581e-06,
+      "loss": 3.8475,
+      "step": 2646016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.668564424059022e-06,
+      "loss": 3.8373,
+      "step": 2646528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.660194855352233e-06,
+      "loss": 3.8359,
+      "step": 2647040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.651825286645444e-06,
+      "loss": 3.8338,
+      "step": 2647552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.643439339134924e-06,
+      "loss": 3.8445,
+      "step": 2648064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.635053391624404e-06,
+      "loss": 3.832,
+      "step": 2648576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.626667444113883e-06,
+      "loss": 3.825,
+      "step": 2649088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.618281496603364e-06,
+      "loss": 3.8333,
+      "step": 2649600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.609895549092844e-06,
+      "loss": 3.8409,
+      "step": 2650112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.601509601582325e-06,
+      "loss": 3.8317,
+      "step": 2650624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.593123654071803e-06,
+      "loss": 3.8446,
+      "step": 2651136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.584754085365015e-06,
+      "loss": 3.8314,
+      "step": 2651648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.5763681378544955e-06,
+      "loss": 3.8438,
+      "step": 2652160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.567982190343975e-06,
+      "loss": 3.8319,
+      "step": 2652672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.559612621637186e-06,
+      "loss": 3.836,
+      "step": 2653184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.551226674126666e-06,
+      "loss": 3.8501,
+      "step": 2653696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.542840726616146e-06,
+      "loss": 3.8482,
+      "step": 2654208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.534454779105625e-06,
+      "loss": 3.8388,
+      "step": 2654720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.526068831595106e-06,
+      "loss": 3.837,
+      "step": 2655232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.517682884084586e-06,
+      "loss": 3.8379,
+      "step": 2655744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.509296936574065e-06,
+      "loss": 3.845,
+      "step": 2656256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.500910989063545e-06,
+      "loss": 3.8316,
+      "step": 2656768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.492541420356757e-06,
+      "loss": 3.8477,
+      "step": 2657280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.484171851649968e-06,
+      "loss": 3.8373,
+      "step": 2657792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.4757859041394475e-06,
+      "loss": 3.8477,
+      "step": 2658304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.4673999566289274e-06,
+      "loss": 3.8282,
+      "step": 2658816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.459014009118408e-06,
+      "loss": 3.8359,
+      "step": 2659328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.450628061607888e-06,
+      "loss": 3.8334,
+      "step": 2659840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.442258492901099e-06,
+      "loss": 3.8384,
+      "step": 2660352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.433872545390579e-06,
+      "loss": 3.8349,
+      "step": 2660864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.42548659788006e-06,
+      "loss": 3.8554,
+      "step": 2661376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.417100650369538e-06,
+      "loss": 3.8429,
+      "step": 2661888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.408714702859019e-06,
+      "loss": 3.8393,
+      "step": 2662400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.400328755348499e-06,
+      "loss": 3.8321,
+      "step": 2662912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.3919428078379796e-06,
+      "loss": 3.8499,
+      "step": 2663424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.383556860327458e-06,
+      "loss": 3.8474,
+      "step": 2663936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.375170912816939e-06,
+      "loss": 3.845,
+      "step": 2664448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.36680134411015e-06,
+      "loss": 3.8406,
+      "step": 2664960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.358415396599629e-06,
+      "loss": 3.8441,
+      "step": 2665472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.350045827892841e-06,
+      "loss": 3.8493,
+      "step": 2665984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.341659880382321e-06,
+      "loss": 3.8392,
+      "step": 2666496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.333273932871801e-06,
+      "loss": 3.837,
+      "step": 2667008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.32488798536128e-06,
+      "loss": 3.8412,
+      "step": 2667520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.316502037850761e-06,
+      "loss": 3.8386,
+      "step": 2668032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.308116090340241e-06,
+      "loss": 3.834,
+      "step": 2668544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.29973014282972e-06,
+      "loss": 3.8425,
+      "step": 2669056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.2913441953192e-06,
+      "loss": 3.8436,
+      "step": 2669568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.282958247808681e-06,
+      "loss": 3.8393,
+      "step": 2670080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.274588679101892e-06,
+      "loss": 3.8353,
+      "step": 2670592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.266219110395102e-06,
+      "loss": 3.8308,
+      "step": 2671104
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.971803665161133,
+      "eval_runtime": 298.2173,
+      "eval_samples_per_second": 1279.574,
+      "eval_steps_per_second": 39.988,
+      "step": 2671200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.257833162884583e-06,
+      "loss": 3.8419,
+      "step": 2671616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.249447215374063e-06,
+      "loss": 3.8411,
+      "step": 2672128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.241061267863543e-06,
+      "loss": 3.8508,
+      "step": 2672640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.232675320353022e-06,
+      "loss": 3.8397,
+      "step": 2673152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.224289372842503e-06,
+      "loss": 3.8489,
+      "step": 2673664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.215903425331982e-06,
+      "loss": 3.8403,
+      "step": 2674176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.207517477821463e-06,
+      "loss": 3.8419,
+      "step": 2674688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.199131530310942e-06,
+      "loss": 3.8299,
+      "step": 2675200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.1907619616041535e-06,
+      "loss": 3.8525,
+      "step": 2675712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.1823760140936335e-06,
+      "loss": 3.8469,
+      "step": 2676224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.1739900665831126e-06,
+      "loss": 3.8459,
+      "step": 2676736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.165604119072593e-06,
+      "loss": 3.8519,
+      "step": 2677248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.157234550365804e-06,
+      "loss": 3.8332,
+      "step": 2677760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.148848602855285e-06,
+      "loss": 3.8375,
+      "step": 2678272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.140462655344764e-06,
+      "loss": 3.8465,
+      "step": 2678784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.132076707834245e-06,
+      "loss": 3.8338,
+      "step": 2679296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.123707139127456e-06,
+      "loss": 3.8341,
+      "step": 2679808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.115321191616936e-06,
+      "loss": 3.8318,
+      "step": 2680320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.106935244106416e-06,
+      "loss": 3.8354,
+      "step": 2680832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.098549296595895e-06,
+      "loss": 3.8625,
+      "step": 2681344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.0901633490853755e-06,
+      "loss": 3.8419,
+      "step": 2681856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.081777401574855e-06,
+      "loss": 3.842,
+      "step": 2682368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.073391454064335e-06,
+      "loss": 3.8385,
+      "step": 2682880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.0650055065538145e-06,
+      "loss": 3.8444,
+      "step": 2683392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.0566195590432945e-06,
+      "loss": 3.8296,
+      "step": 2683904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.0482336115327744e-06,
+      "loss": 3.8322,
+      "step": 2684416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.039864042825986e-06,
+      "loss": 3.8369,
+      "step": 2684928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.031478095315466e-06,
+      "loss": 3.8444,
+      "step": 2685440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.023092147804946e-06,
+      "loss": 3.8311,
+      "step": 2685952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.014706200294426e-06,
+      "loss": 3.8389,
+      "step": 2686464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.006320252783906e-06,
+      "loss": 3.839,
+      "step": 2686976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.9979506840771175e-06,
+      "loss": 3.8449,
+      "step": 2687488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.989581115370328e-06,
+      "loss": 3.8374,
+      "step": 2688000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.981195167859808e-06,
+      "loss": 3.8431,
+      "step": 2688512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.972809220349288e-06,
+      "loss": 3.8452,
+      "step": 2689024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.964423272838768e-06,
+      "loss": 3.8429,
+      "step": 2689536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.956037325328248e-06,
+      "loss": 3.823,
+      "step": 2690048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.947651377817728e-06,
+      "loss": 3.8396,
+      "step": 2690560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.939265430307207e-06,
+      "loss": 3.8251,
+      "step": 2691072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.930879482796688e-06,
+      "loss": 3.8416,
+      "step": 2691584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.922509914089899e-06,
+      "loss": 3.8325,
+      "step": 2692096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.914123966579379e-06,
+      "loss": 3.8437,
+      "step": 2692608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.905738019068859e-06,
+      "loss": 3.8374,
+      "step": 2693120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.897352071558339e-06,
+      "loss": 3.8372,
+      "step": 2693632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.8889825028515495e-06,
+      "loss": 3.847,
+      "step": 2694144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.88059655534103e-06,
+      "loss": 3.8407,
+      "step": 2694656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.872210607830509e-06,
+      "loss": 3.8425,
+      "step": 2695168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.863824660319989e-06,
+      "loss": 3.8268,
+      "step": 2695680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.855438712809469e-06,
+      "loss": 3.8323,
+      "step": 2696192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.847069144102681e-06,
+      "loss": 3.8492,
+      "step": 2696704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.838683196592161e-06,
+      "loss": 3.8381,
+      "step": 2697216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.830297249081641e-06,
+      "loss": 3.8318,
+      "step": 2697728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.821911301571121e-06,
+      "loss": 3.8343,
+      "step": 2698240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.8135417328643316e-06,
+      "loss": 3.8393,
+      "step": 2698752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.805155785353812e-06,
+      "loss": 3.8239,
+      "step": 2699264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.7967698378432915e-06,
+      "loss": 3.8426,
+      "step": 2699776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.7883838903327714e-06,
+      "loss": 3.8363,
+      "step": 2700288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.779997942822251e-06,
+      "loss": 3.8431,
+      "step": 2700800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.771611995311731e-06,
+      "loss": 3.8455,
+      "step": 2701312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.763226047801211e-06,
+      "loss": 3.8294,
+      "step": 2701824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.754840100290691e-06,
+      "loss": 3.8308,
+      "step": 2702336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.746470531583903e-06,
+      "loss": 3.8376,
+      "step": 2702848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.738084584073383e-06,
+      "loss": 3.8258,
+      "step": 2703360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.7297150153665945e-06,
+      "loss": 3.8278,
+      "step": 2703872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.721329067856074e-06,
+      "loss": 3.8375,
+      "step": 2704384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.7129431203455536e-06,
+      "loss": 3.8423,
+      "step": 2704896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.7045571728350335e-06,
+      "loss": 3.8294,
+      "step": 2705408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.6961712253245135e-06,
+      "loss": 3.8224,
+      "step": 2705920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.687801656617725e-06,
+      "loss": 3.8377,
+      "step": 2706432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.679415709107205e-06,
+      "loss": 3.8464,
+      "step": 2706944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.671029761596685e-06,
+      "loss": 3.836,
+      "step": 2707456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.662660192889896e-06,
+      "loss": 3.8432,
+      "step": 2707968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.654274245379377e-06,
+      "loss": 3.8474,
+      "step": 2708480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.645888297868856e-06,
+      "loss": 3.8461,
+      "step": 2708992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.637502350358336e-06,
+      "loss": 3.8511,
+      "step": 2709504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.629116402847816e-06,
+      "loss": 3.8265,
+      "step": 2710016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.620730455337296e-06,
+      "loss": 3.8437,
+      "step": 2710528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.6123445078267755e-06,
+      "loss": 3.8459,
+      "step": 2711040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.6039585603162555e-06,
+      "loss": 3.8319,
+      "step": 2711552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.595572612805735e-06,
+      "loss": 3.8444,
+      "step": 2712064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.587186665295215e-06,
+      "loss": 3.8359,
+      "step": 2712576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5788007177846945e-06,
+      "loss": 3.8514,
+      "step": 2713088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.570431149077906e-06,
+      "loss": 3.8449,
+      "step": 2713600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.562045201567386e-06,
+      "loss": 3.8238,
+      "step": 2714112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.553659254056866e-06,
+      "loss": 3.8336,
+      "step": 2714624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.545273306546346e-06,
+      "loss": 3.8346,
+      "step": 2715136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.536887359035826e-06,
+      "loss": 3.8418,
+      "step": 2715648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.528501411525306e-06,
+      "loss": 3.8432,
+      "step": 2716160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.520148221622249e-06,
+      "loss": 3.8357,
+      "step": 2716672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.511762274111728e-06,
+      "loss": 3.8355,
+      "step": 2717184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.503376326601208e-06,
+      "loss": 3.8288,
+      "step": 2717696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.494990379090688e-06,
+      "loss": 3.8383,
+      "step": 2718208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.486604431580168e-06,
+      "loss": 3.8318,
+      "step": 2718720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.478218484069648e-06,
+      "loss": 3.8428,
+      "step": 2719232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.469832536559128e-06,
+      "loss": 3.85,
+      "step": 2719744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.461446589048608e-06,
+      "loss": 3.8414,
+      "step": 2720256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.45307702034182e-06,
+      "loss": 3.8306,
+      "step": 2720768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.444691072831299e-06,
+      "loss": 3.8403,
+      "step": 2721280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.43630512532078e-06,
+      "loss": 3.8376,
+      "step": 2721792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.427919177810259e-06,
+      "loss": 3.8464,
+      "step": 2722304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.41954960910347e-06,
+      "loss": 3.8395,
+      "step": 2722816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.41116366159295e-06,
+      "loss": 3.8332,
+      "step": 2723328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.40277771408243e-06,
+      "loss": 3.829,
+      "step": 2723840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.39439176657191e-06,
+      "loss": 3.8448,
+      "step": 2724352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.386005819061389e-06,
+      "loss": 3.8242,
+      "step": 2724864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.37761987155087e-06,
+      "loss": 3.8259,
+      "step": 2725376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.369250302844081e-06,
+      "loss": 3.8298,
+      "step": 2725888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.360864355333562e-06,
+      "loss": 3.8383,
+      "step": 2726400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.352478407823041e-06,
+      "loss": 3.8315,
+      "step": 2726912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.344092460312521e-06,
+      "loss": 3.8468,
+      "step": 2727424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.335706512802001e-06,
+      "loss": 3.8317,
+      "step": 2727936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.327320565291481e-06,
+      "loss": 3.8347,
+      "step": 2728448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.318934617780961e-06,
+      "loss": 3.8321,
+      "step": 2728960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.3105650490741715e-06,
+      "loss": 3.8275,
+      "step": 2729472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.302179101563652e-06,
+      "loss": 3.8491,
+      "step": 2729984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.293809532856863e-06,
+      "loss": 3.8443,
+      "step": 2730496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.285423585346343e-06,
+      "loss": 3.8389,
+      "step": 2731008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.277054016639555e-06,
+      "loss": 3.8363,
+      "step": 2731520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.268668069129035e-06,
+      "loss": 3.8307,
+      "step": 2732032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.260282121618515e-06,
+      "loss": 3.8426,
+      "step": 2732544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.251896174107994e-06,
+      "loss": 3.8294,
+      "step": 2733056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.2435102265974745e-06,
+      "loss": 3.8489,
+      "step": 2733568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.235124279086954e-06,
+      "loss": 3.8306,
+      "step": 2734080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.226738331576434e-06,
+      "loss": 3.8398,
+      "step": 2734592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.2183523840659135e-06,
+      "loss": 3.8314,
+      "step": 2735104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.2099664365553935e-06,
+      "loss": 3.8326,
+      "step": 2735616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.201580489044873e-06,
+      "loss": 3.8295,
+      "step": 2736128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.193210920338085e-06,
+      "loss": 3.8361,
+      "step": 2736640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.184824972827565e-06,
+      "loss": 3.8298,
+      "step": 2737152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.176439025317045e-06,
+      "loss": 3.8502,
+      "step": 2737664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.168053077806525e-06,
+      "loss": 3.8421,
+      "step": 2738176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.159667130296005e-06,
+      "loss": 3.836,
+      "step": 2738688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.151281182785484e-06,
+      "loss": 3.8274,
+      "step": 2739200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.142911614078696e-06,
+      "loss": 3.8485,
+      "step": 2739712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.134525666568176e-06,
+      "loss": 3.8438,
+      "step": 2740224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.1261397190576555e-06,
+      "loss": 3.8465,
+      "step": 2740736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.1177537715471355e-06,
+      "loss": 3.836,
+      "step": 2741248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.1093678240366154e-06,
+      "loss": 3.8399,
+      "step": 2741760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.100981876526095e-06,
+      "loss": 3.845,
+      "step": 2742272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.0925959290155745e-06,
+      "loss": 3.8412,
+      "step": 2742784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.084226360308786e-06,
+      "loss": 3.8306,
+      "step": 2743296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.075840412798266e-06,
+      "loss": 3.8397,
+      "step": 2743808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.067454465287746e-06,
+      "loss": 3.8348,
+      "step": 2744320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.059068517777226e-06,
+      "loss": 3.8306,
+      "step": 2744832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.050682570266707e-06,
+      "loss": 3.8396,
+      "step": 2745344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.042313001559918e-06,
+      "loss": 3.8444,
+      "step": 2745856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.0339270540493976e-06,
+      "loss": 3.8353,
+      "step": 2746368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.0255411065388775e-06,
+      "loss": 3.83,
+      "step": 2746880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.017155159028357e-06,
+      "loss": 3.8376,
+      "step": 2747392
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9708855152130127,
+      "eval_runtime": 297.9043,
+      "eval_samples_per_second": 1280.918,
+      "eval_steps_per_second": 40.03,
+      "step": 2747520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.0087692115178374e-06,
+      "loss": 3.8339,
+      "step": 2747904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.0003832640073165e-06,
+      "loss": 3.8397,
+      "step": 2748416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991997316496797e-06,
+      "loss": 3.8467,
+      "step": 2748928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9836113689862765e-06,
+      "loss": 3.8375,
+      "step": 2749440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.975225421475756e-06,
+      "loss": 3.8489,
+      "step": 2749952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.966839473965236e-06,
+      "loss": 3.8351,
+      "step": 2750464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.958453526454716e-06,
+      "loss": 3.8389,
+      "step": 2750976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.950067578944196e-06,
+      "loss": 3.8299,
+      "step": 2751488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.941681631433676e-06,
+      "loss": 3.8526,
+      "step": 2752000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.933295683923156e-06,
+      "loss": 3.8445,
+      "step": 2752512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.924909736412636e-06,
+      "loss": 3.8406,
+      "step": 2753024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.916523788902116e-06,
+      "loss": 3.8505,
+      "step": 2753536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.908154220195328e-06,
+      "loss": 3.83,
+      "step": 2754048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.899768272684807e-06,
+      "loss": 3.8383,
+      "step": 2754560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.891382325174288e-06,
+      "loss": 3.842,
+      "step": 2755072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.882996377663767e-06,
+      "loss": 3.8302,
+      "step": 2755584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.874610430153247e-06,
+      "loss": 3.8315,
+      "step": 2756096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.866224482642727e-06,
+      "loss": 3.8321,
+      "step": 2756608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.857854913935938e-06,
+      "loss": 3.8354,
+      "step": 2757120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.849468966425418e-06,
+      "loss": 3.8527,
+      "step": 2757632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.841083018914898e-06,
+      "loss": 3.8403,
+      "step": 2758144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.832697071404378e-06,
+      "loss": 3.8401,
+      "step": 2758656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.824311123893858e-06,
+      "loss": 3.8365,
+      "step": 2759168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.815925176383337e-06,
+      "loss": 3.8452,
+      "step": 2759680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.807539228872818e-06,
+      "loss": 3.8303,
+      "step": 2760192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.799153281362297e-06,
+      "loss": 3.8321,
+      "step": 2760704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.790767333851777e-06,
+      "loss": 3.8298,
+      "step": 2761216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.782397765144989e-06,
+      "loss": 3.8414,
+      "step": 2761728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.774011817634468e-06,
+      "loss": 3.8278,
+      "step": 2762240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.765625870123949e-06,
+      "loss": 3.8372,
+      "step": 2762752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.757239922613428e-06,
+      "loss": 3.8389,
+      "step": 2763264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7488539751029086e-06,
+      "loss": 3.8433,
+      "step": 2763776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.740484406396119e-06,
+      "loss": 3.8347,
+      "step": 2764288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7320984588856e-06,
+      "loss": 3.8392,
+      "step": 2764800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723712511375079e-06,
+      "loss": 3.8431,
+      "step": 2765312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715326563864559e-06,
+      "loss": 3.8408,
+      "step": 2765824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706956995157771e-06,
+      "loss": 3.8241,
+      "step": 2766336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69857104764725e-06,
+      "loss": 3.8332,
+      "step": 2766848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690185100136731e-06,
+      "loss": 3.8245,
+      "step": 2767360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68179915262621e-06,
+      "loss": 3.8377,
+      "step": 2767872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.673413205115691e-06,
+      "loss": 3.8311,
+      "step": 2768384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.66502725760517e-06,
+      "loss": 3.843,
+      "step": 2768896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.656657688898381e-06,
+      "loss": 3.8347,
+      "step": 2769408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.648271741387861e-06,
+      "loss": 3.839,
+      "step": 2769920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.639885793877341e-06,
+      "loss": 3.8409,
+      "step": 2770432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.631499846366821e-06,
+      "loss": 3.8365,
+      "step": 2770944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.623130277660032e-06,
+      "loss": 3.8427,
+      "step": 2771456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.614744330149513e-06,
+      "loss": 3.8266,
+      "step": 2771968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.606358382638992e-06,
+      "loss": 3.824,
+      "step": 2772480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.597972435128473e-06,
+      "loss": 3.8467,
+      "step": 2772992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.589586487617952e-06,
+      "loss": 3.8371,
+      "step": 2773504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.581200540107432e-06,
+      "loss": 3.8285,
+      "step": 2774016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5728309714006435e-06,
+      "loss": 3.8316,
+      "step": 2774528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5644450238901235e-06,
+      "loss": 3.8365,
+      "step": 2775040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.556059076379603e-06,
+      "loss": 3.821,
+      "step": 2775552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.547673128869083e-06,
+      "loss": 3.8409,
+      "step": 2776064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.539287181358563e-06,
+      "loss": 3.8335,
+      "step": 2776576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.530901233848043e-06,
+      "loss": 3.838,
+      "step": 2777088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.522531665141255e-06,
+      "loss": 3.8444,
+      "step": 2777600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.514145717630734e-06,
+      "loss": 3.8298,
+      "step": 2778112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.505759770120214e-06,
+      "loss": 3.8234,
+      "step": 2778624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.497373822609694e-06,
+      "loss": 3.8422,
+      "step": 2779136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.488987875099174e-06,
+      "loss": 3.8175,
+      "step": 2779648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.480601927588654e-06,
+      "loss": 3.827,
+      "step": 2780160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472215980078134e-06,
+      "loss": 3.838,
+      "step": 2780672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463830032567613e-06,
+      "loss": 3.8392,
+      "step": 2781184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4554604638608245e-06,
+      "loss": 3.8272,
+      "step": 2781696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447090895154036e-06,
+      "loss": 3.8237,
+      "step": 2782208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438704947643516e-06,
+      "loss": 3.8329,
+      "step": 2782720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430319000132996e-06,
+      "loss": 3.8405,
+      "step": 2783232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.421949431426208e-06,
+      "loss": 3.8365,
+      "step": 2783744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.413563483915687e-06,
+      "loss": 3.8386,
+      "step": 2784256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.405177536405168e-06,
+      "loss": 3.8453,
+      "step": 2784768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.396791588894647e-06,
+      "loss": 3.8399,
+      "step": 2785280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3884056413841276e-06,
+      "loss": 3.8539,
+      "step": 2785792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.380036072677338e-06,
+      "loss": 3.8233,
+      "step": 2786304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.371650125166818e-06,
+      "loss": 3.8419,
+      "step": 2786816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.363264177656298e-06,
+      "loss": 3.8422,
+      "step": 2787328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.354878230145778e-06,
+      "loss": 3.828,
+      "step": 2787840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346492282635258e-06,
+      "loss": 3.8456,
+      "step": 2788352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338106335124738e-06,
+      "loss": 3.8346,
+      "step": 2788864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329720387614218e-06,
+      "loss": 3.8452,
+      "step": 2789376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321334440103698e-06,
+      "loss": 3.8476,
+      "step": 2789888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312948492593177e-06,
+      "loss": 3.8222,
+      "step": 2790400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304578923886389e-06,
+      "loss": 3.8327,
+      "step": 2790912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.296192976375869e-06,
+      "loss": 3.8302,
+      "step": 2791424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.287807028865349e-06,
+      "loss": 3.8405,
+      "step": 2791936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.279421081354829e-06,
+      "loss": 3.8407,
+      "step": 2792448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.27105151264804e-06,
+      "loss": 3.8317,
+      "step": 2792960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26266556513752e-06,
+      "loss": 3.8361,
+      "step": 2793472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254279617627e-06,
+      "loss": 3.8302,
+      "step": 2793984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.24589367011648e-06,
+      "loss": 3.8343,
+      "step": 2794496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.237507722605959e-06,
+      "loss": 3.8308,
+      "step": 2795008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.22912177509544e-06,
+      "loss": 3.8436,
+      "step": 2795520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.220752206388651e-06,
+      "loss": 3.8463,
+      "step": 2796032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.212366258878131e-06,
+      "loss": 3.8381,
+      "step": 2796544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.203980311367611e-06,
+      "loss": 3.8311,
+      "step": 2797056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.195594363857091e-06,
+      "loss": 3.8355,
+      "step": 2797568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.187208416346571e-06,
+      "loss": 3.8357,
+      "step": 2798080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17882246883605e-06,
+      "loss": 3.845,
+      "step": 2798592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170452900129261e-06,
+      "loss": 3.837,
+      "step": 2799104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162066952618741e-06,
+      "loss": 3.8311,
+      "step": 2799616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153681005108221e-06,
+      "loss": 3.8261,
+      "step": 2800128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145295057597701e-06,
+      "loss": 3.8389,
+      "step": 2800640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136909110087181e-06,
+      "loss": 3.8227,
+      "step": 2801152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128539541380393e-06,
+      "loss": 3.8283,
+      "step": 2801664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.120153593869872e-06,
+      "loss": 3.8229,
+      "step": 2802176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.111767646359353e-06,
+      "loss": 3.8365,
+      "step": 2802688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.103381698848832e-06,
+      "loss": 3.8244,
+      "step": 2803200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.094995751338313e-06,
+      "loss": 3.8466,
+      "step": 2803712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.086609803827792e-06,
+      "loss": 3.8263,
+      "step": 2804224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.078223856317272e-06,
+      "loss": 3.8377,
+      "step": 2804736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.069854287610483e-06,
+      "loss": 3.8327,
+      "step": 2805248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.061468340099963e-06,
+      "loss": 3.824,
+      "step": 2805760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.053082392589443e-06,
+      "loss": 3.8455,
+      "step": 2806272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044696445078923e-06,
+      "loss": 3.8427,
+      "step": 2806784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036326876372135e-06,
+      "loss": 3.8358,
+      "step": 2807296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027940928861614e-06,
+      "loss": 3.8335,
+      "step": 2807808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.019554981351095e-06,
+      "loss": 3.8289,
+      "step": 2808320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011169033840574e-06,
+      "loss": 3.8394,
+      "step": 2808832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.002783086330054e-06,
+      "loss": 3.8228,
+      "step": 2809344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.994397138819534e-06,
+      "loss": 3.8474,
+      "step": 2809856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.986011191309014e-06,
+      "loss": 3.8338,
+      "step": 2810368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.977625243798494e-06,
+      "loss": 3.8356,
+      "step": 2810880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.969239296287974e-06,
+      "loss": 3.8302,
+      "step": 2811392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.960853348777453e-06,
+      "loss": 3.8313,
+      "step": 2811904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.952467401266934e-06,
+      "loss": 3.8246,
+      "step": 2812416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.944097832560144e-06,
+      "loss": 3.8382,
+      "step": 2812928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.935711885049625e-06,
+      "loss": 3.8284,
+      "step": 2813440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.927325937539104e-06,
+      "loss": 3.8467,
+      "step": 2813952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918939990028585e-06,
+      "loss": 3.8393,
+      "step": 2814464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910554042518064e-06,
+      "loss": 3.8383,
+      "step": 2814976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902168095007544e-06,
+      "loss": 3.824,
+      "step": 2815488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893782147497024e-06,
+      "loss": 3.8438,
+      "step": 2816000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885396199986504e-06,
+      "loss": 3.8406,
+      "step": 2816512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877026631279716e-06,
+      "loss": 3.8436,
+      "step": 2817024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8686570625729265e-06,
+      "loss": 3.8349,
+      "step": 2817536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8602711150624065e-06,
+      "loss": 3.8346,
+      "step": 2818048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8518851675518864e-06,
+      "loss": 3.8411,
+      "step": 2818560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.843499220041366e-06,
+      "loss": 3.8422,
+      "step": 2819072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.835113272530846e-06,
+      "loss": 3.8287,
+      "step": 2819584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.826743703824058e-06,
+      "loss": 3.8396,
+      "step": 2820096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.818357756313538e-06,
+      "loss": 3.8296,
+      "step": 2820608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8099718088030175e-06,
+      "loss": 3.8314,
+      "step": 2821120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8015858612924974e-06,
+      "loss": 3.8406,
+      "step": 2821632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7931999137819774e-06,
+      "loss": 3.8357,
+      "step": 2822144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7848139662714573e-06,
+      "loss": 3.8332,
+      "step": 2822656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7764280187609373e-06,
+      "loss": 3.8307,
+      "step": 2823168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768042071250417e-06,
+      "loss": 3.8319,
+      "step": 2823680
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9701929092407227,
+      "eval_runtime": 303.279,
+      "eval_samples_per_second": 1258.217,
+      "eval_steps_per_second": 39.32,
+      "step": 2823840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7596725025436285e-06,
+      "loss": 3.834,
+      "step": 2824192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.751286555033108e-06,
+      "loss": 3.8357,
+      "step": 2824704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7429006075225884e-06,
+      "loss": 3.8454,
+      "step": 2825216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.734514660012068e-06,
+      "loss": 3.8375,
+      "step": 2825728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7261287125015483e-06,
+      "loss": 3.8408,
+      "step": 2826240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.717742764991028e-06,
+      "loss": 3.8367,
+      "step": 2826752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7093568174805073e-06,
+      "loss": 3.8334,
+      "step": 2827264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.700987248773719e-06,
+      "loss": 3.8304,
+      "step": 2827776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.692601301263199e-06,
+      "loss": 3.844,
+      "step": 2828288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6842153537526793e-06,
+      "loss": 3.8464,
+      "step": 2828800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.675829406242159e-06,
+      "loss": 3.8396,
+      "step": 2829312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6674598375353705e-06,
+      "loss": 3.8487,
+      "step": 2829824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6590902688285813e-06,
+      "loss": 3.8259,
+      "step": 2830336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6507043213180617e-06,
+      "loss": 3.8331,
+      "step": 2830848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.642318373807541e-06,
+      "loss": 3.845,
+      "step": 2831360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6339324262970216e-06,
+      "loss": 3.8255,
+      "step": 2831872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.625546478786501e-06,
+      "loss": 3.8324,
+      "step": 2832384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6171605312759806e-06,
+      "loss": 3.8293,
+      "step": 2832896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608774583765461e-06,
+      "loss": 3.834,
+      "step": 2833408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600388636254941e-06,
+      "loss": 3.8508,
+      "step": 2833920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.592035446351884e-06,
+      "loss": 3.8397,
+      "step": 2834432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5836494988413634e-06,
+      "loss": 3.8392,
+      "step": 2834944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5752635513308438e-06,
+      "loss": 3.831,
+      "step": 2835456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5668776038203233e-06,
+      "loss": 3.8437,
+      "step": 2835968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.558491656309803e-06,
+      "loss": 3.8253,
+      "step": 2836480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5501057087992832e-06,
+      "loss": 3.8287,
+      "step": 2836992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5417197612887628e-06,
+      "loss": 3.8321,
+      "step": 2837504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5333501925819744e-06,
+      "loss": 3.8359,
+      "step": 2838016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5249642450714543e-06,
+      "loss": 3.8249,
+      "step": 2838528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5165782975609343e-06,
+      "loss": 3.836,
+      "step": 2839040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5081923500504143e-06,
+      "loss": 3.8377,
+      "step": 2839552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.499806402539894e-06,
+      "loss": 3.8372,
+      "step": 2840064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.491420455029374e-06,
+      "loss": 3.8356,
+      "step": 2840576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4830345075188537e-06,
+      "loss": 3.8357,
+      "step": 2841088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.474648560008334e-06,
+      "loss": 3.8419,
+      "step": 2841600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4662626124978136e-06,
+      "loss": 3.8355,
+      "step": 2842112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4578930437910253e-06,
+      "loss": 3.8276,
+      "step": 2842624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4495070962805048e-06,
+      "loss": 3.8279,
+      "step": 2843136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4411211487699843e-06,
+      "loss": 3.8231,
+      "step": 2843648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4327352012594647e-06,
+      "loss": 3.8343,
+      "step": 2844160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4243492537489442e-06,
+      "loss": 3.8311,
+      "step": 2844672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.415979685042156e-06,
+      "loss": 3.8406,
+      "step": 2845184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.407593737531636e-06,
+      "loss": 3.8315,
+      "step": 2845696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3992077900211158e-06,
+      "loss": 3.832,
+      "step": 2846208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3908218425105957e-06,
+      "loss": 3.8452,
+      "step": 2846720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3824358950000753e-06,
+      "loss": 3.8331,
+      "step": 2847232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3740499474895556e-06,
+      "loss": 3.8414,
+      "step": 2847744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3656803787827664e-06,
+      "loss": 3.8236,
+      "step": 2848256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.357294431272247e-06,
+      "loss": 3.8208,
+      "step": 2848768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3489084837617263e-06,
+      "loss": 3.85,
+      "step": 2849280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3405225362512067e-06,
+      "loss": 3.8321,
+      "step": 2849792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3321365887406863e-06,
+      "loss": 3.8278,
+      "step": 2850304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3237506412301658e-06,
+      "loss": 3.828,
+      "step": 2850816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315364693719646e-06,
+      "loss": 3.837,
+      "step": 2851328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3069951250128574e-06,
+      "loss": 3.8193,
+      "step": 2851840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2986091775023378e-06,
+      "loss": 3.8364,
+      "step": 2852352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2902232299918173e-06,
+      "loss": 3.8322,
+      "step": 2852864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2818372824812977e-06,
+      "loss": 3.8311,
+      "step": 2853376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.273451334970777e-06,
+      "loss": 3.8475,
+      "step": 2853888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2650653874602567e-06,
+      "loss": 3.8245,
+      "step": 2854400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.256679439949737e-06,
+      "loss": 3.8254,
+      "step": 2854912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2482934924392166e-06,
+      "loss": 3.8383,
+      "step": 2855424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2399403025361595e-06,
+      "loss": 3.8199,
+      "step": 2855936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.231554355025639e-06,
+      "loss": 3.8229,
+      "step": 2856448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2231684075151195e-06,
+      "loss": 3.8314,
+      "step": 2856960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2147824600045994e-06,
+      "loss": 3.8362,
+      "step": 2857472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2063965124940794e-06,
+      "loss": 3.8269,
+      "step": 2857984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1980105649835593e-06,
+      "loss": 3.8221,
+      "step": 2858496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.189624617473039e-06,
+      "loss": 3.8268,
+      "step": 2859008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1812386699625192e-06,
+      "loss": 3.8365,
+      "step": 2859520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.17286910125573e-06,
+      "loss": 3.8352,
+      "step": 2860032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1644831537452104e-06,
+      "loss": 3.839,
+      "step": 2860544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.15609720623469e-06,
+      "loss": 3.8397,
+      "step": 2861056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1477112587241703e-06,
+      "loss": 3.8402,
+      "step": 2861568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.13932531121365e-06,
+      "loss": 3.8523,
+      "step": 2862080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1309721213105927e-06,
+      "loss": 3.8195,
+      "step": 2862592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1225861738000727e-06,
+      "loss": 3.8435,
+      "step": 2863104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1142002262895527e-06,
+      "loss": 3.8416,
+      "step": 2863616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1058142787790326e-06,
+      "loss": 3.8209,
+      "step": 2864128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097444710072244e-06,
+      "loss": 3.8444,
+      "step": 2864640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0890587625617238e-06,
+      "loss": 3.8312,
+      "step": 2865152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0806728150512037e-06,
+      "loss": 3.8445,
+      "step": 2865664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0722868675406833e-06,
+      "loss": 3.8452,
+      "step": 2866176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0639009200301632e-06,
+      "loss": 3.8216,
+      "step": 2866688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055514972519643e-06,
+      "loss": 3.8283,
+      "step": 2867200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0471454038128544e-06,
+      "loss": 3.8278,
+      "step": 2867712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0387594563023343e-06,
+      "loss": 3.8394,
+      "step": 2868224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0303735087918143e-06,
+      "loss": 3.8351,
+      "step": 2868736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.022003940085026e-06,
+      "loss": 3.8304,
+      "step": 2869248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.013617992574506e-06,
+      "loss": 3.8302,
+      "step": 2869760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.005232045063986e-06,
+      "loss": 3.8292,
+      "step": 2870272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9968460975534654e-06,
+      "loss": 3.8317,
+      "step": 2870784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9884601500429453e-06,
+      "loss": 3.8264,
+      "step": 2871296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9800742025324253e-06,
+      "loss": 3.8442,
+      "step": 2871808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9716882550219053e-06,
+      "loss": 3.8419,
+      "step": 2872320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.963302307511385e-06,
+      "loss": 3.841,
+      "step": 2872832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9549327388045964e-06,
+      "loss": 3.8228,
+      "step": 2873344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9465467912940764e-06,
+      "loss": 3.834,
+      "step": 2873856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.938160843783556e-06,
+      "loss": 3.8335,
+      "step": 2874368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.929774896273036e-06,
+      "loss": 3.8392,
+      "step": 2874880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.921388948762516e-06,
+      "loss": 3.8381,
+      "step": 2875392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9130193800557275e-06,
+      "loss": 3.832,
+      "step": 2875904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9046334325452074e-06,
+      "loss": 3.825,
+      "step": 2876416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8962474850346874e-06,
+      "loss": 3.8358,
+      "step": 2876928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8878615375241673e-06,
+      "loss": 3.8235,
+      "step": 2877440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.879475590013647e-06,
+      "loss": 3.8227,
+      "step": 2877952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8711060213068585e-06,
+      "loss": 3.8164,
+      "step": 2878464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.862720073796338e-06,
+      "loss": 3.8348,
+      "step": 2878976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.854334126285818e-06,
+      "loss": 3.828,
+      "step": 2879488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.845948178775298e-06,
+      "loss": 3.842,
+      "step": 2880000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.837562231264778e-06,
+      "loss": 3.8273,
+      "step": 2880512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.829176283754258e-06,
+      "loss": 3.8327,
+      "step": 2881024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8208067150474695e-06,
+      "loss": 3.8329,
+      "step": 2881536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.812420767536949e-06,
+      "loss": 3.8215,
+      "step": 2882048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.804034820026429e-06,
+      "loss": 3.8433,
+      "step": 2882560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.795648872515909e-06,
+      "loss": 3.8392,
+      "step": 2883072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.787262925005389e-06,
+      "loss": 3.8329,
+      "step": 2883584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7788933562986e-06,
+      "loss": 3.8325,
+      "step": 2884096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.77050740878808e-06,
+      "loss": 3.8291,
+      "step": 2884608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.76212146127756e-06,
+      "loss": 3.8385,
+      "step": 2885120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7537355137670395e-06,
+      "loss": 3.8173,
+      "step": 2885632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7453495662565195e-06,
+      "loss": 3.8453,
+      "step": 2886144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.736979997549731e-06,
+      "loss": 3.8278,
+      "step": 2886656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.728594050039211e-06,
+      "loss": 3.8322,
+      "step": 2887168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7202244813324223e-06,
+      "loss": 3.8316,
+      "step": 2887680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7118385338219023e-06,
+      "loss": 3.826,
+      "step": 2888192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7034525863113822e-06,
+      "loss": 3.8292,
+      "step": 2888704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.695066638800862e-06,
+      "loss": 3.8312,
+      "step": 2889216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.686680691290342e-06,
+      "loss": 3.8305,
+      "step": 2889728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6782947437798217e-06,
+      "loss": 3.8406,
+      "step": 2890240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6699087962693016e-06,
+      "loss": 3.8354,
+      "step": 2890752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6615228487587816e-06,
+      "loss": 3.841,
+      "step": 2891264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6531369012482615e-06,
+      "loss": 3.82,
+      "step": 2891776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6447509537377415e-06,
+      "loss": 3.8451,
+      "step": 2892288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.636365006227221e-06,
+      "loss": 3.8352,
+      "step": 2892800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.627979058716701e-06,
+      "loss": 3.8387,
+      "step": 2893312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6196258688136443e-06,
+      "loss": 3.8341,
+      "step": 2893824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6112399213031242e-06,
+      "loss": 3.8306,
+      "step": 2894336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6028539737926038e-06,
+      "loss": 3.8434,
+      "step": 2894848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5944680262820837e-06,
+      "loss": 3.8402,
+      "step": 2895360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5860820787715637e-06,
+      "loss": 3.8253,
+      "step": 2895872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5776961312610436e-06,
+      "loss": 3.8336,
+      "step": 2896384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.569326562554255e-06,
+      "loss": 3.8316,
+      "step": 2896896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.560940615043735e-06,
+      "loss": 3.8269,
+      "step": 2897408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5525546675332148e-06,
+      "loss": 3.8394,
+      "step": 2897920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5441687200226943e-06,
+      "loss": 3.8315,
+      "step": 2898432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5357827725121743e-06,
+      "loss": 3.8382,
+      "step": 2898944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5273968250016542e-06,
+      "loss": 3.8221,
+      "step": 2899456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5190108774911346e-06,
+      "loss": 3.8359,
+      "step": 2899968
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9695701599121094,
+      "eval_runtime": 309.9389,
+      "eval_samples_per_second": 1231.181,
+      "eval_steps_per_second": 38.475,
+      "step": 2900160
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.2032121882214426e+18,
+  "trial_name": null,
+  "trial_params": null
+}