diff --git "a/checkpoint-763190/trainer_state.json" "b/checkpoint-763190/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-763190/trainer_state.json"
@@ -0,0 +1,9045 @@
+{
+  "best_metric": 3.897075891494751,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/pp-mod-subj/transformer/4/checkpoints/checkpoint-763190",
+  "epoch": 1.0250002784396635,
+  "eval_steps": 10,
+  "global_step": 763190,
+  "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": 11.0005,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.849,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1928,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.969,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.818,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.6938,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6036,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.528,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4669,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.3968,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3531,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3092,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.276,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2132,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1737,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1452,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1033,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0758,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0546,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0218,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983231380739706e-05,
+      "loss": 5.0072,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9812,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9481,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9428,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.913,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.8886,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8751,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.866,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8306,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756873037409844e-05,
+      "loss": 4.8223,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.7984,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7886,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.79,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7668,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7559,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7431,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7353,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.723,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.708,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6985,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6794,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6781,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964790485618428e-05,
+      "loss": 4.6498,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963951890867376e-05,
+      "loss": 4.6591,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 4.6542,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6418,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9614377444945934e-05,
+      "loss": 4.6241,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9605991497435414e-05,
+      "loss": 4.6233,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9597605549924894e-05,
+      "loss": 4.618,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.6003,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958085003370758e-05,
+      "loss": 4.588,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957248046500079e-05,
+      "loss": 4.5828,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956409451749027e-05,
+      "loss": 4.5931,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955570856997975e-05,
+      "loss": 4.5609,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 4.5819,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 4.5458,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 4.5538,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5441,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5404,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5321,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949702331620985e-05,
+      "loss": 4.526,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948863736869932e-05,
+      "loss": 4.5087,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5206,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.5004,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946349590497149e-05,
+      "loss": 4.5066,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945510995746097e-05,
+      "loss": 4.498,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.4885,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4713,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9429968493733145e-05,
+      "loss": 4.4851,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9421598925026354e-05,
+      "loss": 4.475,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941321297751584e-05,
+      "loss": 4.4599,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940482703000532e-05,
+      "loss": 4.466,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396441082494794e-05,
+      "loss": 4.4575,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388055134984274e-05,
+      "loss": 4.4561,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379669187473754e-05,
+      "loss": 4.4584,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 4.4502,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 4.4416,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 4.4269,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 4.4338,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 4.4308,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 4.4331,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 4.4258,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.4131,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 4.3969,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.4015,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.4024,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.4031,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 4.3985,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926231505873767e-05,
+      "loss": 4.3939,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925392911122715e-05,
+      "loss": 4.3819,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924555954252036e-05,
+      "loss": 4.3787,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923717359500984e-05,
+      "loss": 4.3765,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922878764749932e-05,
+      "loss": 4.3835,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922041807879253e-05,
+      "loss": 4.3755,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921203213128201e-05,
+      "loss": 4.352,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920364618377149e-05,
+      "loss": 4.3571,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919526023626097e-05,
+      "loss": 4.3629,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918687428875045e-05,
+      "loss": 4.3642,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917848834123993e-05,
+      "loss": 4.357,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917010239372941e-05,
+      "loss": 4.3497,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.3493,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.3437,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3472,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913657498249106e-05,
+      "loss": 4.3456,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912818903498054e-05,
+      "loss": 4.3269,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911980308747002e-05,
+      "loss": 4.3363,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.314,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.3246,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.3239,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.3164,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.3199,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3052,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3023,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3077,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90443623174828e-05,
+      "loss": 4.3207,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.2925,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.2872,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901920447495124e-05,
+      "loss": 4.2898,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.3038,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.2917,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899407939002715e-05,
+      "loss": 4.2946,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985693442516624e-05,
+      "loss": 4.288,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977307495006104e-05,
+      "loss": 4.2907,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968921547495584e-05,
+      "loss": 4.281,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.2814,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2757,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.286,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2639,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.2767,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2724,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2598,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2711,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889348077750837e-05,
+      "loss": 4.2657,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888509482999785e-05,
+      "loss": 4.2699,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887670888248733e-05,
+      "loss": 4.2659,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886832293497681e-05,
+      "loss": 4.2481,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.2447,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.2488,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2588,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.2597,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.2515,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2403,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.2414,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8801268112500106e-05,
+      "loss": 4.2444,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8792882164989586e-05,
+      "loss": 4.2404,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2354,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877611026996855e-05,
+      "loss": 4.2504,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876774070126176e-05,
+      "loss": 4.2379,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875935475375124e-05,
+      "loss": 4.2296,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875096880624072e-05,
+      "loss": 4.2271,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.243750095367432,
+      "eval_runtime": 568.7385,
+      "eval_samples_per_second": 670.943,
+      "eval_steps_per_second": 20.967,
+      "step": 76319
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.87425828587302e-05,
+      "loss": 4.2337,
+      "step": 76800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.873419691121968e-05,
+      "loss": 4.2297,
+      "step": 77312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.872581096370916e-05,
+      "loss": 4.2167,
+      "step": 77824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8717425016198635e-05,
+      "loss": 4.2169,
+      "step": 78336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8709039068688115e-05,
+      "loss": 4.2103,
+      "step": 78848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8700653121177595e-05,
+      "loss": 4.2082,
+      "step": 79360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8692267173667075e-05,
+      "loss": 4.2012,
+      "step": 79872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8683881226156555e-05,
+      "loss": 4.2031,
+      "step": 80384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.867549527864604e-05,
+      "loss": 4.2089,
+      "step": 80896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.2071,
+      "step": 81408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.2035,
+      "step": 81920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.2064,
+      "step": 82432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.864196786740769e-05,
+      "loss": 4.1982,
+      "step": 82944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.863358191989717e-05,
+      "loss": 4.1931,
+      "step": 83456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.862519597238665e-05,
+      "loss": 4.1904,
+      "step": 83968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.861681002487613e-05,
+      "loss": 4.1642,
+      "step": 84480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860844045616934e-05,
+      "loss": 4.1869,
+      "step": 84992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860005450865882e-05,
+      "loss": 4.1843,
+      "step": 85504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.85916685611483e-05,
+      "loss": 4.1838,
+      "step": 86016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.858328261363778e-05,
+      "loss": 4.1949,
+      "step": 86528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.1788,
+      "step": 87040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.1812,
+      "step": 87552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.1798,
+      "step": 88064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8549755202399435e-05,
+      "loss": 4.179,
+      "step": 88576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.1698,
+      "step": 89088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8532999686182124e-05,
+      "loss": 4.1698,
+      "step": 89600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8524613738671604e-05,
+      "loss": 4.1761,
+      "step": 90112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8516227791161084e-05,
+      "loss": 4.1589,
+      "step": 90624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.164,
+      "step": 91136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.1519,
+      "step": 91648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.1558,
+      "step": 92160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.1671,
+      "step": 92672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.847433081121595e-05,
+      "loss": 4.1595,
+      "step": 93184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.846594486370543e-05,
+      "loss": 4.1597,
+      "step": 93696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.845755891619491e-05,
+      "loss": 4.158,
+      "step": 94208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.844917296868439e-05,
+      "loss": 4.1576,
+      "step": 94720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.84408033999776e-05,
+      "loss": 4.1567,
+      "step": 95232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.1539,
+      "step": 95744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.842403150495656e-05,
+      "loss": 4.1513,
+      "step": 96256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.841564555744604e-05,
+      "loss": 4.1398,
+      "step": 96768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.840727598873925e-05,
+      "loss": 4.1477,
+      "step": 97280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839889004122873e-05,
+      "loss": 4.1316,
+      "step": 97792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839050409371821e-05,
+      "loss": 4.1432,
+      "step": 98304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8382118146207687e-05,
+      "loss": 4.1481,
+      "step": 98816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.83737485775009e-05,
+      "loss": 4.1433,
+      "step": 99328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.836536262999038e-05,
+      "loss": 4.1356,
+      "step": 99840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.835697668247986e-05,
+      "loss": 4.1377,
+      "step": 100352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834859073496934e-05,
+      "loss": 4.1425,
+      "step": 100864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834022116626255e-05,
+      "loss": 4.1222,
+      "step": 101376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.833183521875203e-05,
+      "loss": 4.1254,
+      "step": 101888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.832344927124151e-05,
+      "loss": 4.1251,
+      "step": 102400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.831506332373099e-05,
+      "loss": 4.1396,
+      "step": 102912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.83066937550242e-05,
+      "loss": 4.1173,
+      "step": 103424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.829830780751368e-05,
+      "loss": 4.1383,
+      "step": 103936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828992186000316e-05,
+      "loss": 4.1073,
+      "step": 104448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828153591249264e-05,
+      "loss": 4.1229,
+      "step": 104960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8273166343785856e-05,
+      "loss": 4.1215,
+      "step": 105472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8264780396275336e-05,
+      "loss": 4.1249,
+      "step": 105984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8256394448764816e-05,
+      "loss": 4.1156,
+      "step": 106496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8248008501254296e-05,
+      "loss": 4.1191,
+      "step": 107008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8239638932547505e-05,
+      "loss": 4.1065,
+      "step": 107520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8231252985036985e-05,
+      "loss": 4.1168,
+      "step": 108032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8222867037526465e-05,
+      "loss": 4.1083,
+      "step": 108544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8214481090015945e-05,
+      "loss": 4.1133,
+      "step": 109056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8206111521309154e-05,
+      "loss": 4.1083,
+      "step": 109568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8197725573798634e-05,
+      "loss": 4.1046,
+      "step": 110080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8189339626288114e-05,
+      "loss": 4.0879,
+      "step": 110592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8180953678777594e-05,
+      "loss": 4.1127,
+      "step": 111104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.817258411007081e-05,
+      "loss": 4.1025,
+      "step": 111616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.816419816256029e-05,
+      "loss": 4.0941,
+      "step": 112128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.815581221504977e-05,
+      "loss": 4.1016,
+      "step": 112640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.814742626753924e-05,
+      "loss": 4.0982,
+      "step": 113152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813905669883246e-05,
+      "loss": 4.0995,
+      "step": 113664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813067075132194e-05,
+      "loss": 4.1085,
+      "step": 114176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.812228480381141e-05,
+      "loss": 4.1035,
+      "step": 114688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.811389885630089e-05,
+      "loss": 4.0942,
+      "step": 115200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.810552928759411e-05,
+      "loss": 4.0869,
+      "step": 115712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.809714334008359e-05,
+      "loss": 4.0926,
+      "step": 116224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808875739257306e-05,
+      "loss": 4.095,
+      "step": 116736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808037144506255e-05,
+      "loss": 4.1045,
+      "step": 117248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.807200187635576e-05,
+      "loss": 4.0934,
+      "step": 117760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8063615928845243e-05,
+      "loss": 4.0882,
+      "step": 118272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8055229981334717e-05,
+      "loss": 4.0731,
+      "step": 118784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.804686041262793e-05,
+      "loss": 4.0785,
+      "step": 119296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.803847446511741e-05,
+      "loss": 4.0815,
+      "step": 119808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8030088517606886e-05,
+      "loss": 4.0834,
+      "step": 120320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8021702570096366e-05,
+      "loss": 4.0908,
+      "step": 120832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.801333300138958e-05,
+      "loss": 4.0829,
+      "step": 121344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.800494705387906e-05,
+      "loss": 4.0749,
+      "step": 121856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7996561106368535e-05,
+      "loss": 4.0793,
+      "step": 122368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7988175158858015e-05,
+      "loss": 4.071,
+      "step": 122880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797980559015123e-05,
+      "loss": 4.0805,
+      "step": 123392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797141964264071e-05,
+      "loss": 4.0798,
+      "step": 123904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.796303369513019e-05,
+      "loss": 4.0608,
+      "step": 124416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.795464774761967e-05,
+      "loss": 4.0623,
+      "step": 124928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7946278178912886e-05,
+      "loss": 4.0697,
+      "step": 125440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.793789223140236e-05,
+      "loss": 4.0802,
+      "step": 125952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7929522662695575e-05,
+      "loss": 4.0656,
+      "step": 126464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7921136715185055e-05,
+      "loss": 4.0741,
+      "step": 126976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7912750767674535e-05,
+      "loss": 4.0662,
+      "step": 127488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.790436482016401e-05,
+      "loss": 4.0645,
+      "step": 128000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.789597887265349e-05,
+      "loss": 4.0701,
+      "step": 128512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.788759292514297e-05,
+      "loss": 4.069,
+      "step": 129024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7879223356436184e-05,
+      "loss": 4.0562,
+      "step": 129536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7870837408925664e-05,
+      "loss": 4.0683,
+      "step": 130048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7862451461415144e-05,
+      "loss": 4.0479,
+      "step": 130560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7854065513904624e-05,
+      "loss": 4.055,
+      "step": 131072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7845679566394104e-05,
+      "loss": 4.0617,
+      "step": 131584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7837293618883584e-05,
+      "loss": 4.0502,
+      "step": 132096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7828907671373064e-05,
+      "loss": 4.0558,
+      "step": 132608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7820521723862544e-05,
+      "loss": 4.0497,
+      "step": 133120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.781215215515575e-05,
+      "loss": 4.0453,
+      "step": 133632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.780378258644896e-05,
+      "loss": 4.0511,
+      "step": 134144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.779539663893844e-05,
+      "loss": 4.0662,
+      "step": 134656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.778701069142792e-05,
+      "loss": 4.0424,
+      "step": 135168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77786247439174e-05,
+      "loss": 4.0417,
+      "step": 135680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777023879640689e-05,
+      "loss": 4.0404,
+      "step": 136192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.776185284889637e-05,
+      "loss": 4.0507,
+      "step": 136704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.775346690138585e-05,
+      "loss": 4.0486,
+      "step": 137216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.774509733267906e-05,
+      "loss": 4.0463,
+      "step": 137728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.773671138516854e-05,
+      "loss": 4.0474,
+      "step": 138240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.772832543765802e-05,
+      "loss": 4.0481,
+      "step": 138752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77199394901475e-05,
+      "loss": 4.0442,
+      "step": 139264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771155354263698e-05,
+      "loss": 4.0392,
+      "step": 139776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.770316759512646e-05,
+      "loss": 4.0412,
+      "step": 140288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.769478164761594e-05,
+      "loss": 4.0511,
+      "step": 140800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.768639570010542e-05,
+      "loss": 4.0341,
+      "step": 141312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7678026131398626e-05,
+      "loss": 4.0374,
+      "step": 141824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7669640183888106e-05,
+      "loss": 4.0487,
+      "step": 142336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7661254236377586e-05,
+      "loss": 4.0249,
+      "step": 142848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.765286828886707e-05,
+      "loss": 4.0421,
+      "step": 143360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.764451509896401e-05,
+      "loss": 4.0354,
+      "step": 143872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.763612915145349e-05,
+      "loss": 4.047,
+      "step": 144384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.762774320394297e-05,
+      "loss": 4.0394,
+      "step": 144896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761935725643245e-05,
+      "loss": 4.0266,
+      "step": 145408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761097130892193e-05,
+      "loss": 4.0248,
+      "step": 145920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.760258536141141e-05,
+      "loss": 4.0304,
+      "step": 146432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.759421579270462e-05,
+      "loss": 4.0399,
+      "step": 146944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.75858298451941e-05,
+      "loss": 4.0414,
+      "step": 147456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.757744389768358e-05,
+      "loss": 4.0377,
+      "step": 147968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756905795017306e-05,
+      "loss": 4.0228,
+      "step": 148480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756067200266254e-05,
+      "loss": 4.0271,
+      "step": 148992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.755228605515202e-05,
+      "loss": 4.0346,
+      "step": 149504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.75439001076415e-05,
+      "loss": 4.0287,
+      "step": 150016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7535530538934716e-05,
+      "loss": 4.0246,
+      "step": 150528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7527144591424196e-05,
+      "loss": 4.0425,
+      "step": 151040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751875864391367e-05,
+      "loss": 4.0232,
+      "step": 151552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751037269640315e-05,
+      "loss": 4.0267,
+      "step": 152064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.750198674889263e-05,
+      "loss": 4.0211,
+      "step": 152576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.073619365692139,
+      "eval_runtime": 572.2917,
+      "eval_samples_per_second": 666.777,
+      "eval_steps_per_second": 20.837,
+      "step": 152638
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493617180185845e-05,
+      "loss": 4.0314,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748523123267532e-05,
+      "loss": 4.027,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74768452851648e-05,
+      "loss": 4.0185,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746845933765428e-05,
+      "loss": 4.016,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460073390143764e-05,
+      "loss": 4.0105,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451687442633244e-05,
+      "loss": 4.0113,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443301495122724e-05,
+      "loss": 4.0091,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434915547612204e-05,
+      "loss": 4.0082,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426545978905413e-05,
+      "loss": 4.0131,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741816003139489e-05,
+      "loss": 4.0143,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740977408388437e-05,
+      "loss": 4.0136,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740138813637385e-05,
+      "loss": 4.014,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739301856766706e-05,
+      "loss": 4.0086,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738463262015654e-05,
+      "loss": 4.0035,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737624667264602e-05,
+      "loss": 4.0048,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73678607251355e-05,
+      "loss": 3.9782,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735947477762498e-05,
+      "loss": 4.0035,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735108883011446e-05,
+      "loss": 4.0003,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734270288260394e-05,
+      "loss": 3.9989,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733433331389716e-05,
+      "loss": 4.0165,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732594736638664e-05,
+      "loss": 3.9985,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731756141887612e-05,
+      "loss": 4.004,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73091754713656e-05,
+      "loss": 4.0009,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730078952385508e-05,
+      "loss": 3.9989,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729241995514829e-05,
+      "loss": 3.9932,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728403400763777e-05,
+      "loss": 3.9925,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727564806012725e-05,
+      "loss": 4.0005,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726726211261673e-05,
+      "loss": 3.9855,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725887616510621e-05,
+      "loss": 3.9913,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725049021759568e-05,
+      "loss": 3.9783,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7242120648888896e-05,
+      "loss": 3.9875,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723373470137838e-05,
+      "loss": 3.9964,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225348753867856e-05,
+      "loss": 3.9858,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216962806357336e-05,
+      "loss": 3.9933,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720859323765055e-05,
+      "loss": 3.9882,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720020729014003e-05,
+      "loss": 3.9907,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191821342629505e-05,
+      "loss": 3.9938,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183435395118985e-05,
+      "loss": 3.9892,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175049447608465e-05,
+      "loss": 3.9855,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716667987890168e-05,
+      "loss": 3.9746,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158293931391154e-05,
+      "loss": 3.9852,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149907983880634e-05,
+      "loss": 3.9711,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714152203637012e-05,
+      "loss": 3.9873,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71331360888596e-05,
+      "loss": 3.9839,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712475014134908e-05,
+      "loss": 3.9855,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711638057264229e-05,
+      "loss": 3.9764,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710799462513177e-05,
+      "loss": 3.9774,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709960867762125e-05,
+      "loss": 3.9849,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709122273011073e-05,
+      "loss": 3.9654,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708285316140394e-05,
+      "loss": 3.9712,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707446721389342e-05,
+      "loss": 3.9664,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70660812663829e-05,
+      "loss": 3.9847,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705769531887238e-05,
+      "loss": 3.9653,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704930937136186e-05,
+      "loss": 3.9848,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040939802655074e-05,
+      "loss": 3.9562,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032553855144554e-05,
+      "loss": 3.9683,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024167907634034e-05,
+      "loss": 3.9729,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015781960123514e-05,
+      "loss": 3.9714,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700741239141672e-05,
+      "loss": 3.9675,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69990264439062e-05,
+      "loss": 3.9809,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699064049639568e-05,
+      "loss": 3.95,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698225454888516e-05,
+      "loss": 3.9665,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697386860137464e-05,
+      "loss": 3.9617,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696549903266785e-05,
+      "loss": 3.9703,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695711308515733e-05,
+      "loss": 3.9581,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694872713764681e-05,
+      "loss": 3.9615,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694034119013629e-05,
+      "loss": 3.9449,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693197162142951e-05,
+      "loss": 3.9678,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692358567391899e-05,
+      "loss": 3.9583,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691519972640847e-05,
+      "loss": 3.9551,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690681377889795e-05,
+      "loss": 3.96,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898444210191157e-05,
+      "loss": 3.9585,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6890058262680636e-05,
+      "loss": 3.9563,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881672315170116e-05,
+      "loss": 3.971,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873286367659596e-05,
+      "loss": 3.9653,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864916798952806e-05,
+      "loss": 3.9575,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856530851442285e-05,
+      "loss": 3.9497,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848144903931765e-05,
+      "loss": 3.9519,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6839758956421245e-05,
+      "loss": 3.9594,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831373008910725e-05,
+      "loss": 3.9669,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682300344020394e-05,
+      "loss": 3.9544,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681461749269342e-05,
+      "loss": 3.9572,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68062315451829e-05,
+      "loss": 3.9387,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679784559767238e-05,
+      "loss": 3.9484,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678945965016186e-05,
+      "loss": 3.9406,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678109008145507e-05,
+      "loss": 3.9499,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677270413394455e-05,
+      "loss": 3.9578,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676431818643403e-05,
+      "loss": 3.9499,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675593223892351e-05,
+      "loss": 3.9422,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674756267021672e-05,
+      "loss": 3.948,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.67391767227062e-05,
+      "loss": 3.9404,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673079077519568e-05,
+      "loss": 3.9511,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672240482768516e-05,
+      "loss": 3.9478,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6714035258978375e-05,
+      "loss": 3.9334,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705649311467855e-05,
+      "loss": 3.9339,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697263363957335e-05,
+      "loss": 3.9408,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688877416446815e-05,
+      "loss": 3.9506,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668049146893629e-05,
+      "loss": 3.9399,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6672121900229504e-05,
+      "loss": 3.9491,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663735952718984e-05,
+      "loss": 3.9375,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665535000520846e-05,
+      "loss": 3.9383,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664696405769794e-05,
+      "loss": 3.9432,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663859448899115e-05,
+      "loss": 3.9436,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663020854148063e-05,
+      "loss": 3.9319,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662182259397011e-05,
+      "loss": 3.9479,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661343664645959e-05,
+      "loss": 3.9241,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660506707775281e-05,
+      "loss": 3.9281,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659668113024229e-05,
+      "loss": 3.9403,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658829518273176e-05,
+      "loss": 3.9251,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657990923522124e-05,
+      "loss": 3.9378,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657153966651446e-05,
+      "loss": 3.9273,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656315371900393e-05,
+      "loss": 3.9222,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655476777149341e-05,
+      "loss": 3.927,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654638182398289e-05,
+      "loss": 3.9477,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6538012255276106e-05,
+      "loss": 3.9236,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529626307765586e-05,
+      "loss": 3.9264,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6521240360255066e-05,
+      "loss": 3.9186,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512854412744546e-05,
+      "loss": 3.9278,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650448484403776e-05,
+      "loss": 3.9354,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6496098896527235e-05,
+      "loss": 3.9288,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487712949016715e-05,
+      "loss": 3.9266,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479327001506195e-05,
+      "loss": 3.928,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470957432799404e-05,
+      "loss": 3.9269,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462571485288884e-05,
+      "loss": 3.9262,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454185537778364e-05,
+      "loss": 3.9242,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445799590267844e-05,
+      "loss": 3.9338,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437413642757324e-05,
+      "loss": 3.9209,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642904407405054e-05,
+      "loss": 3.9204,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642065812654002e-05,
+      "loss": 3.9352,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64122721790295e-05,
+      "loss": 3.912,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640388623151898e-05,
+      "loss": 3.9253,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639551666281219e-05,
+      "loss": 3.9236,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638713071530167e-05,
+      "loss": 3.934,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637874476779115e-05,
+      "loss": 3.9249,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637035882028063e-05,
+      "loss": 3.9197,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636198925157384e-05,
+      "loss": 3.9108,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635360330406332e-05,
+      "loss": 3.9162,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63452173565528e-05,
+      "loss": 3.9275,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633683140904228e-05,
+      "loss": 3.9305,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328461840335494e-05,
+      "loss": 3.9262,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6320075892824974e-05,
+      "loss": 3.915,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311689945314454e-05,
+      "loss": 3.9144,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303303997803934e-05,
+      "loss": 3.9286,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629493442909714e-05,
+      "loss": 3.9156,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628654848158662e-05,
+      "loss": 3.9166,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62781625340761e-05,
+      "loss": 3.9303,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626977658656558e-05,
+      "loss": 3.9154,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626140701785879e-05,
+      "loss": 3.9199,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625302107034827e-05,
+      "loss": 3.9159,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0001678466796875,
+      "eval_runtime": 585.2038,
+      "eval_samples_per_second": 652.065,
+      "eval_steps_per_second": 20.378,
+      "step": 228957
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.624463512283775e-05,
+      "loss": 3.9246,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.623624917532723e-05,
+      "loss": 3.9203,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.622786322781671e-05,
+      "loss": 3.9119,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.62194772803062e-05,
+      "loss": 3.9105,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621109133279568e-05,
+      "loss": 3.9081,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.620270538528516e-05,
+      "loss": 3.9044,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.619431943777464e-05,
+      "loss": 3.9074,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.618593349026412e-05,
+      "loss": 3.9021,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.617756392155733e-05,
+      "loss": 3.9096,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616917797404681e-05,
+      "loss": 3.9161,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616079202653629e-05,
+      "loss": 3.9076,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.615240607902577e-05,
+      "loss": 3.9071,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6144036510318976e-05,
+      "loss": 3.9075,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6135650562808456e-05,
+      "loss": 3.9042,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6127264615297936e-05,
+      "loss": 3.901,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6118878667787416e-05,
+      "loss": 3.8802,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611050909908063e-05,
+      "loss": 3.8977,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.610212315157011e-05,
+      "loss": 3.899,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.609373720405959e-05,
+      "loss": 3.9005,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6085351256549065e-05,
+      "loss": 3.9142,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.607698168784228e-05,
+      "loss": 3.8995,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606859574033176e-05,
+      "loss": 3.9076,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606020979282124e-05,
+      "loss": 3.8965,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.605184022411445e-05,
+      "loss": 3.8974,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.604345427660393e-05,
+      "loss": 3.896,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.603506832909341e-05,
+      "loss": 3.8981,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.602668238158289e-05,
+      "loss": 3.8992,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.60183128128761e-05,
+      "loss": 3.8885,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6009926865365585e-05,
+      "loss": 3.8909,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6001540917855065e-05,
+      "loss": 3.8819,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599315497034454e-05,
+      "loss": 3.8913,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598476902283402e-05,
+      "loss": 3.899,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5976399454127234e-05,
+      "loss": 3.8937,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5968013506616714e-05,
+      "loss": 3.8955,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595962755910619e-05,
+      "loss": 3.8936,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5951257990399403e-05,
+      "loss": 3.8949,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.594287204288888e-05,
+      "loss": 3.8959,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.593448609537836e-05,
+      "loss": 3.8968,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5926100147867836e-05,
+      "loss": 3.8883,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.591773057916105e-05,
+      "loss": 3.8805,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590934463165054e-05,
+      "loss": 3.8909,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590095868414001e-05,
+      "loss": 3.8738,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.589257273662949e-05,
+      "loss": 3.8977,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.588420316792271e-05,
+      "loss": 3.8916,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.587581722041219e-05,
+      "loss": 3.8904,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.586743127290166e-05,
+      "loss": 3.8835,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585904532539114e-05,
+      "loss": 3.8835,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585067575668436e-05,
+      "loss": 3.8957,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584228980917384e-05,
+      "loss": 3.8722,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.583390386166331e-05,
+      "loss": 3.8779,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.582551791415279e-05,
+      "loss": 3.8791,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5817148345446006e-05,
+      "loss": 3.8902,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5808762397935486e-05,
+      "loss": 3.8723,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5800376450424966e-05,
+      "loss": 3.8931,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5791990502914446e-05,
+      "loss": 3.8707,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.578362093420766e-05,
+      "loss": 3.8716,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5775234986697135e-05,
+      "loss": 3.8867,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5766849039186615e-05,
+      "loss": 3.874,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5758463091676095e-05,
+      "loss": 3.8863,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.575009352296931e-05,
+      "loss": 3.8918,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5741707575458784e-05,
+      "loss": 3.8653,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5733321627948264e-05,
+      "loss": 3.8725,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5724935680437744e-05,
+      "loss": 3.8754,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.571656611173096e-05,
+      "loss": 3.8805,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.570818016422044e-05,
+      "loss": 3.8697,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569979421670992e-05,
+      "loss": 3.8735,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.56914082691994e-05,
+      "loss": 3.8568,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.568303870049261e-05,
+      "loss": 3.8805,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.567465275298209e-05,
+      "loss": 3.8711,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.566626680547157e-05,
+      "loss": 3.8672,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.565788085796105e-05,
+      "loss": 3.8717,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564951128925426e-05,
+      "loss": 3.8724,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564112534174374e-05,
+      "loss": 3.8714,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.563273939423322e-05,
+      "loss": 3.8829,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.56243534467227e-05,
+      "loss": 3.8802,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561598387801591e-05,
+      "loss": 3.874,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560759793050539e-05,
+      "loss": 3.867,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559921198299487e-05,
+      "loss": 3.8681,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559082603548435e-05,
+      "loss": 3.871,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.558245646677756e-05,
+      "loss": 3.8833,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.557407051926704e-05,
+      "loss": 3.8726,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.556568457175652e-05,
+      "loss": 3.8659,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5557298624246e-05,
+      "loss": 3.86,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554892905553921e-05,
+      "loss": 3.8646,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554054310802869e-05,
+      "loss": 3.8588,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.553215716051817e-05,
+      "loss": 3.8657,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.552377121300765e-05,
+      "loss": 3.8681,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.551540164430087e-05,
+      "loss": 3.8672,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.550701569679035e-05,
+      "loss": 3.8613,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549862974927983e-05,
+      "loss": 3.8701,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549024380176931e-05,
+      "loss": 3.859,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5481874233062516e-05,
+      "loss": 3.8636,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5473488285551996e-05,
+      "loss": 3.8643,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5465102338041476e-05,
+      "loss": 3.8522,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5456716390530956e-05,
+      "loss": 3.8508,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5448346821824165e-05,
+      "loss": 3.8648,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5439960874313645e-05,
+      "loss": 3.862,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5431574926803125e-05,
+      "loss": 3.868,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5423188979292605e-05,
+      "loss": 3.8632,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.541481941058582e-05,
+      "loss": 3.8568,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54064334630753e-05,
+      "loss": 3.8565,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.539804751556478e-05,
+      "loss": 3.8654,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538966156805426e-05,
+      "loss": 3.8646,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538129199934747e-05,
+      "loss": 3.8516,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.537290605183695e-05,
+      "loss": 3.869,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.536452010432643e-05,
+      "loss": 3.8431,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.535613415681591e-05,
+      "loss": 3.8524,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.534776458810912e-05,
+      "loss": 3.8605,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.53393786405986e-05,
+      "loss": 3.847,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533099269308808e-05,
+      "loss": 3.8602,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.532260674557756e-05,
+      "loss": 3.8454,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5314237176870774e-05,
+      "loss": 3.8418,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5305851229360254e-05,
+      "loss": 3.8496,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5297465281849734e-05,
+      "loss": 3.8695,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5289079334339214e-05,
+      "loss": 3.8486,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528070976563242e-05,
+      "loss": 3.8484,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.52723238181219e-05,
+      "loss": 3.8419,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.526393787061138e-05,
+      "loss": 3.8506,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.525555192310086e-05,
+      "loss": 3.857,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.524718235439407e-05,
+      "loss": 3.8533,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523879640688355e-05,
+      "loss": 3.8446,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523041045937303e-05,
+      "loss": 3.8561,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.522202451186251e-05,
+      "loss": 3.8503,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.521365494315573e-05,
+      "loss": 3.851,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.520526899564521e-05,
+      "loss": 3.8456,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.519688304813469e-05,
+      "loss": 3.8576,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.518849710062417e-05,
+      "loss": 3.8462,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.518012753191738e-05,
+      "loss": 3.8458,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517174158440686e-05,
+      "loss": 3.8599,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.516335563689634e-05,
+      "loss": 3.8362,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515496968938582e-05,
+      "loss": 3.8494,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5146600120679026e-05,
+      "loss": 3.8517,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5138214173168506e-05,
+      "loss": 3.8567,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5129828225657986e-05,
+      "loss": 3.849,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5121442278147466e-05,
+      "loss": 3.8474,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.511307270944068e-05,
+      "loss": 3.8374,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.510468676193016e-05,
+      "loss": 3.8415,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.509630081441964e-05,
+      "loss": 3.8513,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.508791486690912e-05,
+      "loss": 3.8564,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507954529820233e-05,
+      "loss": 3.8545,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507115935069181e-05,
+      "loss": 3.8439,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.506277340318129e-05,
+      "loss": 3.8378,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.505438745567077e-05,
+      "loss": 3.854,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.504601788696398e-05,
+      "loss": 3.843,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503763193945346e-05,
+      "loss": 3.8447,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502924599194294e-05,
+      "loss": 3.8556,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5020876423236155e-05,
+      "loss": 3.8464,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5012490475725635e-05,
+      "loss": 3.8461,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5004104528215115e-05,
+      "loss": 3.8439,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9584410190582275,
+      "eval_runtime": 567.4578,
+      "eval_samples_per_second": 672.457,
+      "eval_steps_per_second": 21.015,
+      "step": 305276
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995718580704595e-05,
+      "loss": 3.8527,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4987332633194075e-05,
+      "loss": 3.8476,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978946685683555e-05,
+      "loss": 3.8394,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970560738173035e-05,
+      "loss": 3.8381,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496217479066251e-05,
+      "loss": 3.838,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495378884315199e-05,
+      "loss": 3.8335,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494540289564147e-05,
+      "loss": 3.8376,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493701694813095e-05,
+      "loss": 3.83,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492864737942416e-05,
+      "loss": 3.8396,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492027781071737e-05,
+      "loss": 3.8461,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491189186320685e-05,
+      "loss": 3.8329,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490350591569633e-05,
+      "loss": 3.8342,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489513634698955e-05,
+      "loss": 3.8393,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488675039947903e-05,
+      "loss": 3.835,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487836445196851e-05,
+      "loss": 3.8332,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486997850445798e-05,
+      "loss": 3.8105,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486159255694746e-05,
+      "loss": 3.8275,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485320660943694e-05,
+      "loss": 3.8306,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484482066192642e-05,
+      "loss": 3.8311,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48364347144159e-05,
+      "loss": 3.8493,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482804876690538e-05,
+      "loss": 3.8264,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481966281939486e-05,
+      "loss": 3.8397,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481127687188434e-05,
+      "loss": 3.8257,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480290730317756e-05,
+      "loss": 3.8324,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479452135566704e-05,
+      "loss": 3.8238,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478613540815652e-05,
+      "loss": 3.8305,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4777749460646e-05,
+      "loss": 3.8314,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4769379891939207e-05,
+      "loss": 3.8244,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4760993944428687e-05,
+      "loss": 3.8225,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752607996918166e-05,
+      "loss": 3.8166,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4744222049407646e-05,
+      "loss": 3.8183,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4735852480700856e-05,
+      "loss": 3.8307,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4727466533190335e-05,
+      "loss": 3.8272,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4719080585679815e-05,
+      "loss": 3.8313,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710694638169295e-05,
+      "loss": 3.8261,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470232506946251e-05,
+      "loss": 3.8296,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469393912195199e-05,
+      "loss": 3.828,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468555317444147e-05,
+      "loss": 3.8298,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467716722693095e-05,
+      "loss": 3.8245,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466879765822416e-05,
+      "loss": 3.8148,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466041171071364e-05,
+      "loss": 3.8295,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465202576320312e-05,
+      "loss": 3.8086,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46436398156926e-05,
+      "loss": 3.8284,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463527024698581e-05,
+      "loss": 3.8243,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462688429947529e-05,
+      "loss": 3.8302,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461849835196477e-05,
+      "loss": 3.8193,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461011240445425e-05,
+      "loss": 3.8167,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4601742835747465e-05,
+      "loss": 3.8301,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4593356888236945e-05,
+      "loss": 3.811,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584970940726425e-05,
+      "loss": 3.8128,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576584993215905e-05,
+      "loss": 3.8157,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4568215424509114e-05,
+      "loss": 3.8203,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559829476998594e-05,
+      "loss": 3.8119,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551443529488074e-05,
+      "loss": 3.8246,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4543057581977554e-05,
+      "loss": 3.8073,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453468801327076e-05,
+      "loss": 3.8109,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452630206576024e-05,
+      "loss": 3.8213,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451791611824972e-05,
+      "loss": 3.8071,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45095301707392e-05,
+      "loss": 3.8261,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450116060203242e-05,
+      "loss": 3.831,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44927746545219e-05,
+      "loss": 3.7996,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448438870701138e-05,
+      "loss": 3.8079,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447600275950086e-05,
+      "loss": 3.8134,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446763319079407e-05,
+      "loss": 3.8226,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445924724328355e-05,
+      "loss": 3.8069,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445086129577303e-05,
+      "loss": 3.8078,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444247534826251e-05,
+      "loss": 3.7981,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4434105779555717e-05,
+      "loss": 3.8145,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425719832045196e-05,
+      "loss": 3.8098,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4417333884534676e-05,
+      "loss": 3.8025,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4408947937024156e-05,
+      "loss": 3.8112,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440057836831737e-05,
+      "loss": 3.8099,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439219242080685e-05,
+      "loss": 3.8091,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438380647329633e-05,
+      "loss": 3.8213,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437543690458954e-05,
+      "loss": 3.8183,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436705095707902e-05,
+      "loss": 3.8131,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43586650095685e-05,
+      "loss": 3.8078,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435027906205798e-05,
+      "loss": 3.8104,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434190949335119e-05,
+      "loss": 3.805,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433352354584067e-05,
+      "loss": 3.8279,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432513759833015e-05,
+      "loss": 3.8094,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431675165081963e-05,
+      "loss": 3.807,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430838208211284e-05,
+      "loss": 3.7959,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4300012513406055e-05,
+      "loss": 3.8106,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291626565895535e-05,
+      "loss": 3.7961,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4283240618385015e-05,
+      "loss": 3.8083,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274854670874495e-05,
+      "loss": 3.8067,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266468723363975e-05,
+      "loss": 3.8056,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4258082775853455e-05,
+      "loss": 3.8024,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249696828342935e-05,
+      "loss": 3.8116,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4241327259636144e-05,
+      "loss": 3.7984,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232941312125624e-05,
+      "loss": 3.8047,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224555364615104e-05,
+      "loss": 3.8078,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216169417104584e-05,
+      "loss": 3.7899,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420779984839779e-05,
+      "loss": 3.7936,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419941390088728e-05,
+      "loss": 3.8082,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419102795337676e-05,
+      "loss": 3.7993,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418264200586624e-05,
+      "loss": 3.8147,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417427243715945e-05,
+      "loss": 3.8011,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416590286845266e-05,
+      "loss": 3.7934,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415751692094214e-05,
+      "loss": 3.7991,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414913097343162e-05,
+      "loss": 3.8052,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41407450259211e-05,
+      "loss": 3.8063,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413235907841058e-05,
+      "loss": 3.7956,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412397313090006e-05,
+      "loss": 3.8089,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411558718338953e-05,
+      "loss": 3.785,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4107217614682747e-05,
+      "loss": 3.793,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409883166717223e-05,
+      "loss": 3.804,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409044571966171e-05,
+      "loss": 3.7861,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4082059772151186e-05,
+      "loss": 3.8018,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40736902034444e-05,
+      "loss": 3.7973,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406532063473761e-05,
+      "loss": 3.7789,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405693468722709e-05,
+      "loss": 3.7933,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404854873971657e-05,
+      "loss": 3.8113,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404016279220605e-05,
+      "loss": 3.7903,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403177684469553e-05,
+      "loss": 3.7929,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4023390897185004e-05,
+      "loss": 3.7864,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4015004949674484e-05,
+      "loss": 3.7913,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400661900216397e-05,
+      "loss": 3.8028,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399824943345719e-05,
+      "loss": 3.7962,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398986348594666e-05,
+      "loss": 3.7887,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398147753843614e-05,
+      "loss": 3.7944,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397309159092562e-05,
+      "loss": 3.7962,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3964738401022565e-05,
+      "loss": 3.7932,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3956352453512045e-05,
+      "loss": 3.7921,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947966506001525e-05,
+      "loss": 3.8002,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3939580558491005e-05,
+      "loss": 3.7864,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393119461098048e-05,
+      "loss": 3.795,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392280866346996e-05,
+      "loss": 3.799,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391442271595944e-05,
+      "loss": 3.7845,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3906036768448925e-05,
+      "loss": 3.7901,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897667199742134e-05,
+      "loss": 3.7978,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3889281252231614e-05,
+      "loss": 3.7978,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880895304721094e-05,
+      "loss": 3.7974,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872509357210574e-05,
+      "loss": 3.7909,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386413978850378e-05,
+      "loss": 3.7821,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385575384099326e-05,
+      "loss": 3.7864,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384736789348274e-05,
+      "loss": 3.7941,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383898194597222e-05,
+      "loss": 3.8014,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383061237726543e-05,
+      "loss": 3.8019,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382222642975491e-05,
+      "loss": 3.7914,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381384048224439e-05,
+      "loss": 3.7805,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380547091353761e-05,
+      "loss": 3.7986,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379708496602709e-05,
+      "loss": 3.7855,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378869901851657e-05,
+      "loss": 3.7909,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378031307100605e-05,
+      "loss": 3.8031,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377192712349553e-05,
+      "loss": 3.793,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376354117598501e-05,
+      "loss": 3.7864,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375515522847449e-05,
+      "loss": 3.7926,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9340226650238037,
+      "eval_runtime": 557.6516,
+      "eval_samples_per_second": 684.282,
+      "eval_steps_per_second": 21.384,
+      "step": 381595
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374676928096397e-05,
+      "loss": 3.8003,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373838333345345e-05,
+      "loss": 3.7969,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372999738594293e-05,
+      "loss": 3.7822,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372161143843241e-05,
+      "loss": 3.7822,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3713241869725616e-05,
+      "loss": 3.7876,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3704855922215096e-05,
+      "loss": 3.7809,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3696469974704576e-05,
+      "loss": 3.7832,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368808402719406e-05,
+      "loss": 3.7703,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367969807968354e-05,
+      "loss": 3.7881,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367132851097675e-05,
+      "loss": 3.7967,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366294256346623e-05,
+      "loss": 3.7782,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365455661595571e-05,
+      "loss": 3.7794,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364618704724892e-05,
+      "loss": 3.7882,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36378010997384e-05,
+      "loss": 3.7807,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362941515222788e-05,
+      "loss": 3.7828,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362102920471736e-05,
+      "loss": 3.7584,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361264325720684e-05,
+      "loss": 3.7721,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3604257309696314e-05,
+      "loss": 3.7766,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35958713621858e-05,
+      "loss": 3.7786,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358748541467528e-05,
+      "loss": 3.7968,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357909946716476e-05,
+      "loss": 3.7764,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357071351965424e-05,
+      "loss": 3.7889,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356232757214372e-05,
+      "loss": 3.7749,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355395800343693e-05,
+      "loss": 3.7805,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354557205592641e-05,
+      "loss": 3.7697,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353718610841589e-05,
+      "loss": 3.7767,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352880016090537e-05,
+      "loss": 3.7796,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352041421339485e-05,
+      "loss": 3.7732,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351202826588433e-05,
+      "loss": 3.7698,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350365869717754e-05,
+      "loss": 3.7662,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349527274966702e-05,
+      "loss": 3.762,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34868868021565e-05,
+      "loss": 3.7814,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347850085464598e-05,
+      "loss": 3.7815,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3470114907135465e-05,
+      "loss": 3.7769,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3461728959624945e-05,
+      "loss": 3.7761,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3453343012114425e-05,
+      "loss": 3.7789,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3444973443407634e-05,
+      "loss": 3.7801,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3436587495897114e-05,
+      "loss": 3.7749,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342821792719032e-05,
+      "loss": 3.7767,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34198319796798e-05,
+      "loss": 3.7649,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341144603216928e-05,
+      "loss": 3.7773,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.340306008465876e-05,
+      "loss": 3.7593,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339469051595197e-05,
+      "loss": 3.78,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338630456844145e-05,
+      "loss": 3.7716,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337791862093093e-05,
+      "loss": 3.7786,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336953267342042e-05,
+      "loss": 3.7693,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33611467259099e-05,
+      "loss": 3.7677,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335277715720311e-05,
+      "loss": 3.7837,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334439120969259e-05,
+      "loss": 3.7571,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333600526218207e-05,
+      "loss": 3.764,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332761931467155e-05,
+      "loss": 3.767,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331923336716103e-05,
+      "loss": 3.7694,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331086379845424e-05,
+      "loss": 3.765,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3302477850943717e-05,
+      "loss": 3.7715,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3294091903433197e-05,
+      "loss": 3.7625,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3285705955922677e-05,
+      "loss": 3.76,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3277320008412156e-05,
+      "loss": 3.7724,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326895043970537e-05,
+      "loss": 3.7613,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326056449219485e-05,
+      "loss": 3.7727,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3252178544684325e-05,
+      "loss": 3.7798,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3243792597173805e-05,
+      "loss": 3.7544,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3235406649663285e-05,
+      "loss": 3.7581,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3227020702152765e-05,
+      "loss": 3.7609,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3218634754642245e-05,
+      "loss": 3.7775,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3210248807131725e-05,
+      "loss": 3.7595,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3201879238424934e-05,
+      "loss": 3.7617,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3193493290914414e-05,
+      "loss": 3.749,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3185107343403894e-05,
+      "loss": 3.7681,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3176721395893374e-05,
+      "loss": 3.7582,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3168335448382854e-05,
+      "loss": 3.7554,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315996587967607e-05,
+      "loss": 3.7619,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315157993216555e-05,
+      "loss": 3.7634,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314319398465503e-05,
+      "loss": 3.7608,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313480803714451e-05,
+      "loss": 3.7722,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312642208963399e-05,
+      "loss": 3.771,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31180525209272e-05,
+      "loss": 3.7646,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310966657341668e-05,
+      "loss": 3.76,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310128062590616e-05,
+      "loss": 3.7646,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309289467839564e-05,
+      "loss": 3.7534,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308450873088512e-05,
+      "loss": 3.776,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30761227833746e-05,
+      "loss": 3.7606,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.306775321466781e-05,
+      "loss": 3.7629,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305936726715729e-05,
+      "loss": 3.7494,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3050981319646775e-05,
+      "loss": 3.7631,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042595372136254e-05,
+      "loss": 3.7462,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3034225803429464e-05,
+      "loss": 3.7624,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025839855918944e-05,
+      "loss": 3.7598,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3017453908408424e-05,
+      "loss": 3.7581,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3009067960897903e-05,
+      "loss": 3.7588,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300069839219111e-05,
+      "loss": 3.7677,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299231244468059e-05,
+      "loss": 3.7459,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298392649717007e-05,
+      "loss": 3.7554,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297554054965955e-05,
+      "loss": 3.7627,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296715460214903e-05,
+      "loss": 3.7443,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295876865463851e-05,
+      "loss": 3.7478,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295039908593173e-05,
+      "loss": 3.7611,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294201313842121e-05,
+      "loss": 3.7494,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293362719091069e-05,
+      "loss": 3.7728,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292524124340016e-05,
+      "loss": 3.7558,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291685529588964e-05,
+      "loss": 3.7469,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290846934837912e-05,
+      "loss": 3.751,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29000834008686e-05,
+      "loss": 3.7602,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289169745335808e-05,
+      "loss": 3.7617,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288332788465129e-05,
+      "loss": 3.7472,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2874958315944506e-05,
+      "loss": 3.7623,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286657236843398e-05,
+      "loss": 3.739,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2858186420923466e-05,
+      "loss": 3.7457,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2849800473412946e-05,
+      "loss": 3.7575,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284143090470616e-05,
+      "loss": 3.7443,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2833044957195635e-05,
+      "loss": 3.7542,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2824659009685115e-05,
+      "loss": 3.7568,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2816273062174595e-05,
+      "loss": 3.7299,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2807887114664075e-05,
+      "loss": 3.7498,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2799517545957284e-05,
+      "loss": 3.764,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2791131598446764e-05,
+      "loss": 3.7503,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2782745650936244e-05,
+      "loss": 3.7433,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2774359703425724e-05,
+      "loss": 3.7418,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2765973755915204e-05,
+      "loss": 3.7459,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275760418720842e-05,
+      "loss": 3.7572,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27492182396979e-05,
+      "loss": 3.7536,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274083229218738e-05,
+      "loss": 3.7415,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273244634467686e-05,
+      "loss": 3.7491,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272407677597007e-05,
+      "loss": 3.7498,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271569082845955e-05,
+      "loss": 3.7523,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270730488094903e-05,
+      "loss": 3.7452,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269891893343851e-05,
+      "loss": 3.7543,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269054936473172e-05,
+      "loss": 3.746,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26821634172212e-05,
+      "loss": 3.7496,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267377746971068e-05,
+      "loss": 3.7544,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266539152220016e-05,
+      "loss": 3.7386,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265702195349337e-05,
+      "loss": 3.7462,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264863600598285e-05,
+      "loss": 3.7556,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264025005847233e-05,
+      "loss": 3.7519,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263186411096181e-05,
+      "loss": 3.7525,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262347816345129e-05,
+      "loss": 3.7478,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26151085947445e-05,
+      "loss": 3.7348,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260672264723398e-05,
+      "loss": 3.7421,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.259833669972346e-05,
+      "loss": 3.749,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258995075221294e-05,
+      "loss": 3.7639,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258156480470242e-05,
+      "loss": 3.7548,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25731788571919e-05,
+      "loss": 3.7446,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256480928848511e-05,
+      "loss": 3.7444,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255642334097459e-05,
+      "loss": 3.7504,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254803739346407e-05,
+      "loss": 3.7437,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253965144595356e-05,
+      "loss": 3.7437,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253128187724677e-05,
+      "loss": 3.762,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252289592973625e-05,
+      "loss": 3.7486,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251450998222573e-05,
+      "loss": 3.7418,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250612403471521e-05,
+      "loss": 3.7493,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.918626546859741,
+      "eval_runtime": 569.637,
+      "eval_samples_per_second": 669.884,
+      "eval_steps_per_second": 20.934,
+      "step": 457914
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249773808720469e-05,
+      "loss": 3.7555,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248935213969417e-05,
+      "loss": 3.7508,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248096619218364e-05,
+      "loss": 3.7427,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247258024467312e-05,
+      "loss": 3.7431,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2464210675966336e-05,
+      "loss": 3.7416,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455824728455816e-05,
+      "loss": 3.7362,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2447438780945296e-05,
+      "loss": 3.741,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2439052833434776e-05,
+      "loss": 3.7294,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243068326472799e-05,
+      "loss": 3.743,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242229731721747e-05,
+      "loss": 3.7561,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2413911369706945e-05,
+      "loss": 3.737,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2405525422196424e-05,
+      "loss": 3.7361,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2397139474685904e-05,
+      "loss": 3.7456,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2388769905979114e-05,
+      "loss": 3.7376,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2380383958468593e-05,
+      "loss": 3.7439,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2371998010958073e-05,
+      "loss": 3.7161,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363612063447553e-05,
+      "loss": 3.7259,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235524249474076e-05,
+      "loss": 3.7337,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234685654723025e-05,
+      "loss": 3.7337,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233847059971973e-05,
+      "loss": 3.7588,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233008465220921e-05,
+      "loss": 3.7338,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232169870469869e-05,
+      "loss": 3.7491,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23133291359919e-05,
+      "loss": 3.7287,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230494318848138e-05,
+      "loss": 3.7379,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229655724097086e-05,
+      "loss": 3.7287,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228817129346034e-05,
+      "loss": 3.7367,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227980172475355e-05,
+      "loss": 3.7381,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227141577724303e-05,
+      "loss": 3.7299,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226302982973251e-05,
+      "loss": 3.7348,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225464388222199e-05,
+      "loss": 3.7242,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2246274313515196e-05,
+      "loss": 3.7187,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223788836600468e-05,
+      "loss": 3.7397,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222950241849416e-05,
+      "loss": 3.7379,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222111647098364e-05,
+      "loss": 3.7398,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221274690227685e-05,
+      "loss": 3.7309,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220436095476633e-05,
+      "loss": 3.7402,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219597500725581e-05,
+      "loss": 3.7346,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218758905974529e-05,
+      "loss": 3.7365,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21792194910385e-05,
+      "loss": 3.7321,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217083354352798e-05,
+      "loss": 3.7259,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216244759601746e-05,
+      "loss": 3.738,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215406164850694e-05,
+      "loss": 3.7165,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214569207980015e-05,
+      "loss": 3.7393,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2137306132289637e-05,
+      "loss": 3.7331,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128920184779116e-05,
+      "loss": 3.737,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2120534237268596e-05,
+      "loss": 3.7307,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2112148289758076e-05,
+      "loss": 3.726,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103778721051285e-05,
+      "loss": 3.7417,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2095392773540765e-05,
+      "loss": 3.7203,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2087006826030245e-05,
+      "loss": 3.7237,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2078620878519725e-05,
+      "loss": 3.7262,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2070251309812934e-05,
+      "loss": 3.7266,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061865362302414e-05,
+      "loss": 3.7268,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2053479414791894e-05,
+      "loss": 3.7279,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2045093467281374e-05,
+      "loss": 3.7224,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203672389857459e-05,
+      "loss": 3.7249,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.202833795106407e-05,
+      "loss": 3.73,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201995200355355e-05,
+      "loss": 3.7235,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201156605604303e-05,
+      "loss": 3.7329,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200319648733624e-05,
+      "loss": 3.739,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199481053982572e-05,
+      "loss": 3.716,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19864245923152e-05,
+      "loss": 3.7175,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197803864480468e-05,
+      "loss": 3.7175,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196965269729416e-05,
+      "loss": 3.7401,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196128312858737e-05,
+      "loss": 3.722,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195289718107685e-05,
+      "loss": 3.7204,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194451123356633e-05,
+      "loss": 3.7118,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193612528605581e-05,
+      "loss": 3.7224,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1927755717349024e-05,
+      "loss": 3.7203,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1919369769838504e-05,
+      "loss": 3.7168,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1910983822327984e-05,
+      "loss": 3.724,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902597874817464e-05,
+      "loss": 3.7233,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189422830611067e-05,
+      "loss": 3.7246,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188584235860015e-05,
+      "loss": 3.7316,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187745641108963e-05,
+      "loss": 3.7324,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186907046357911e-05,
+      "loss": 3.7271,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186070089487232e-05,
+      "loss": 3.7217,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18523149473618e-05,
+      "loss": 3.7234,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184392899985128e-05,
+      "loss": 3.7167,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183554305234076e-05,
+      "loss": 3.7386,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182717348363398e-05,
+      "loss": 3.7225,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181878753612346e-05,
+      "loss": 3.7273,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181040158861294e-05,
+      "loss": 3.7083,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180201564110242e-05,
+      "loss": 3.7274,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1793646072395626e-05,
+      "loss": 3.7047,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1785260124885106e-05,
+      "loss": 3.7236,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776874177374586e-05,
+      "loss": 3.7218,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1768488229864066e-05,
+      "loss": 3.7223,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1760102282353546e-05,
+      "loss": 3.7187,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1751732713646755e-05,
+      "loss": 3.7278,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1743346766136235e-05,
+      "loss": 3.7123,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1734960818625715e-05,
+      "loss": 3.7172,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1726574871115195e-05,
+      "loss": 3.7231,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171820530240841e-05,
+      "loss": 3.708,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170981935489789e-05,
+      "loss": 3.7098,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170143340738737e-05,
+      "loss": 3.7198,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1693047459876844e-05,
+      "loss": 3.7127,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168467789117006e-05,
+      "loss": 3.7349,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167629194365954e-05,
+      "loss": 3.7158,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166790599614902e-05,
+      "loss": 3.7127,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165952004863849e-05,
+      "loss": 3.7141,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165115047993171e-05,
+      "loss": 3.7218,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164276453242119e-05,
+      "loss": 3.727,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163437858491067e-05,
+      "loss": 3.7115,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162599263740015e-05,
+      "loss": 3.723,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1617623068693365e-05,
+      "loss": 3.7054,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1609237121182845e-05,
+      "loss": 3.7055,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160085117367232e-05,
+      "loss": 3.7189,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15924652261618e-05,
+      "loss": 3.7095,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158407927865128e-05,
+      "loss": 3.7139,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1575709709944494e-05,
+      "loss": 3.7178,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156732376243397e-05,
+      "loss": 3.6948,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155893781492345e-05,
+      "loss": 3.7115,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155055186741293e-05,
+      "loss": 3.7299,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154218229870614e-05,
+      "loss": 3.7167,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153379635119562e-05,
+      "loss": 3.703,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15254104036851e-05,
+      "loss": 3.7053,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151702445617458e-05,
+      "loss": 3.7095,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150863850866406e-05,
+      "loss": 3.7198,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150026893995727e-05,
+      "loss": 3.7162,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149188299244675e-05,
+      "loss": 3.706,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148349704493623e-05,
+      "loss": 3.7106,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147511109742571e-05,
+      "loss": 3.7138,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146674152871892e-05,
+      "loss": 3.7133,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14583555812084e-05,
+      "loss": 3.708,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144996963369788e-05,
+      "loss": 3.7157,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144158368618736e-05,
+      "loss": 3.7112,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1433214117480576e-05,
+      "loss": 3.7148,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424828169970056e-05,
+      "loss": 3.715,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1416442222459536e-05,
+      "loss": 3.6998,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1408056274949016e-05,
+      "loss": 3.711,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1399686706242225e-05,
+      "loss": 3.7164,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1391300758731705e-05,
+      "loss": 3.718,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1382914811221185e-05,
+      "loss": 3.7151,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1374528863710665e-05,
+      "loss": 3.7141,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1366142916200145e-05,
+      "loss": 3.7015,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357773347493354e-05,
+      "loss": 3.701,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1349387399982834e-05,
+      "loss": 3.7122,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1341001452472314e-05,
+      "loss": 3.7263,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332615504961794e-05,
+      "loss": 3.7188,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132424593625501e-05,
+      "loss": 3.7088,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131585998874449e-05,
+      "loss": 3.7044,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130747404123397e-05,
+      "loss": 3.7164,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129908809372345e-05,
+      "loss": 3.7087,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129071852501666e-05,
+      "loss": 3.7085,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128233257750614e-05,
+      "loss": 3.7261,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127394662999562e-05,
+      "loss": 3.7133,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12655606824851e-05,
+      "loss": 3.7061,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125717473497458e-05,
+      "loss": 3.7135,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.908137559890747,
+      "eval_runtime": 567.1197,
+      "eval_samples_per_second": 672.858,
+      "eval_steps_per_second": 21.027,
+      "step": 534233
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124880516626779e-05,
+      "loss": 3.7188,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124041921875727e-05,
+      "loss": 3.7168,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123203327124675e-05,
+      "loss": 3.7096,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.122364732373623e-05,
+      "loss": 3.7064,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1215277755029444e-05,
+      "loss": 3.7042,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1206891807518923e-05,
+      "loss": 3.7054,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1198505860008403e-05,
+      "loss": 3.705,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119011991249788e-05,
+      "loss": 3.6957,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118175034379109e-05,
+      "loss": 3.7055,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117336439628057e-05,
+      "loss": 3.7173,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.116497844877005e-05,
+      "loss": 3.7008,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115659250125953e-05,
+      "loss": 3.6999,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114820655374901e-05,
+      "loss": 3.713,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113983698504222e-05,
+      "loss": 3.6993,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11314510375317e-05,
+      "loss": 3.7112,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.112306509002118e-05,
+      "loss": 3.6832,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111467914251067e-05,
+      "loss": 3.6847,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110630957380388e-05,
+      "loss": 3.7024,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109792362629336e-05,
+      "loss": 3.6968,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108953767878284e-05,
+      "loss": 3.7213,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108115173127232e-05,
+      "loss": 3.7023,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1072782162565526e-05,
+      "loss": 3.7093,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1064396215055006e-05,
+      "loss": 3.6959,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1056010267544486e-05,
+      "loss": 3.7055,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1047624320033966e-05,
+      "loss": 3.6908,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1039254751327175e-05,
+      "loss": 3.7022,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1030868803816655e-05,
+      "loss": 3.7063,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1022482856306135e-05,
+      "loss": 3.7008,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101409690879562e-05,
+      "loss": 3.6925,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100572734008883e-05,
+      "loss": 3.6927,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099734139257831e-05,
+      "loss": 3.6831,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098895544506779e-05,
+      "loss": 3.7051,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098056949755727e-05,
+      "loss": 3.7019,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097219992885048e-05,
+      "loss": 3.7018,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.096381398133996e-05,
+      "loss": 3.701,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.095542803382944e-05,
+      "loss": 3.7078,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.094704208631892e-05,
+      "loss": 3.6997,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093867251761213e-05,
+      "loss": 3.6982,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093028657010161e-05,
+      "loss": 3.698,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092190062259109e-05,
+      "loss": 3.6977,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0913514675080575e-05,
+      "loss": 3.6977,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0905145106373784e-05,
+      "loss": 3.6836,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0896759158863264e-05,
+      "loss": 3.7079,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0888373211352744e-05,
+      "loss": 3.6969,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087998726384222e-05,
+      "loss": 3.7048,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0871617695135433e-05,
+      "loss": 3.6983,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0863231747624913e-05,
+      "loss": 3.6891,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085484580011439e-05,
+      "loss": 3.7089,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0846459852603867e-05,
+      "loss": 3.6848,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083809028389708e-05,
+      "loss": 3.6924,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082970433638656e-05,
+      "loss": 3.6873,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082131838887604e-05,
+      "loss": 3.6976,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081293244136552e-05,
+      "loss": 3.6929,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080456287265874e-05,
+      "loss": 3.6941,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079617692514822e-05,
+      "loss": 3.6915,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078779097763769e-05,
+      "loss": 3.6898,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077940503012717e-05,
+      "loss": 3.6951,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077103546142039e-05,
+      "loss": 3.6907,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076264951390987e-05,
+      "loss": 3.7007,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075426356639934e-05,
+      "loss": 3.7005,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.074587761888882e-05,
+      "loss": 3.6867,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0737508050182036e-05,
+      "loss": 3.6799,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0729122102671516e-05,
+      "loss": 3.6867,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0720736155160996e-05,
+      "loss": 3.7034,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0712350207650476e-05,
+      "loss": 3.6915,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0703964260139956e-05,
+      "loss": 3.6844,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0695594691433165e-05,
+      "loss": 3.6832,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0687208743922645e-05,
+      "loss": 3.6902,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0678822796412125e-05,
+      "loss": 3.6831,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0670436848901605e-05,
+      "loss": 3.6878,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0662067280194814e-05,
+      "loss": 3.6891,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0653681332684294e-05,
+      "loss": 3.6859,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0645295385173774e-05,
+      "loss": 3.6926,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0636909437663254e-05,
+      "loss": 3.6994,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062853986895647e-05,
+      "loss": 3.6983,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062015392144595e-05,
+      "loss": 3.6911,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061176797393543e-05,
+      "loss": 3.6898,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.060338202642491e-05,
+      "loss": 3.6914,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.059501245771812e-05,
+      "loss": 3.6849,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05866265102076e-05,
+      "loss": 3.7025,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.057824056269708e-05,
+      "loss": 3.6953,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056985461518656e-05,
+      "loss": 3.6943,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056148504647977e-05,
+      "loss": 3.6747,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055309909896925e-05,
+      "loss": 3.6901,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054471315145873e-05,
+      "loss": 3.6754,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053632720394821e-05,
+      "loss": 3.6882,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.052795763524142e-05,
+      "loss": 3.6902,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05195716877309e-05,
+      "loss": 3.6868,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051118574022038e-05,
+      "loss": 3.6896,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.050279979270986e-05,
+      "loss": 3.6904,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049441384519934e-05,
+      "loss": 3.6806,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.048604427649255e-05,
+      "loss": 3.6856,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.047765832898203e-05,
+      "loss": 3.6905,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046927238147151e-05,
+      "loss": 3.6776,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046088643396099e-05,
+      "loss": 3.6796,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04525168652542e-05,
+      "loss": 3.6794,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044413091774368e-05,
+      "loss": 3.6871,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.043574497023316e-05,
+      "loss": 3.7023,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.042737540152638e-05,
+      "loss": 3.6809,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041898945401586e-05,
+      "loss": 3.6843,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041060350650534e-05,
+      "loss": 3.681,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040221755899482e-05,
+      "loss": 3.6926,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0393847990288026e-05,
+      "loss": 3.6898,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0385462042777506e-05,
+      "loss": 3.6788,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0377076095266986e-05,
+      "loss": 3.6924,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0368690147756466e-05,
+      "loss": 3.6709,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0360304200245946e-05,
+      "loss": 3.6728,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0351934631539155e-05,
+      "loss": 3.6866,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0343548684028635e-05,
+      "loss": 3.6796,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0335162736518115e-05,
+      "loss": 3.6839,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0326776789007595e-05,
+      "loss": 3.6849,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031840722030081e-05,
+      "loss": 3.6605,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031002127279029e-05,
+      "loss": 3.6829,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030163532527977e-05,
+      "loss": 3.6956,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.029324937776925e-05,
+      "loss": 3.6837,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028487980906246e-05,
+      "loss": 3.6729,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027649386155194e-05,
+      "loss": 3.6734,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026810791404142e-05,
+      "loss": 3.6748,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.02597219665309e-05,
+      "loss": 3.6875,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025135239782411e-05,
+      "loss": 3.6861,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.024296645031359e-05,
+      "loss": 3.6736,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.023458050280307e-05,
+      "loss": 3.6763,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.022619455529255e-05,
+      "loss": 3.6889,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0217824986585764e-05,
+      "loss": 3.6821,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0209439039075244e-05,
+      "loss": 3.673,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0201053091564724e-05,
+      "loss": 3.6871,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0192667144054204e-05,
+      "loss": 3.6796,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.018429757534741e-05,
+      "loss": 3.6788,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.017591162783689e-05,
+      "loss": 3.6897,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.016752568032637e-05,
+      "loss": 3.6664,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015913973281585e-05,
+      "loss": 3.679,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015077016410906e-05,
+      "loss": 3.6876,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014238421659854e-05,
+      "loss": 3.6835,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013399826908802e-05,
+      "loss": 3.6812,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01256123215775e-05,
+      "loss": 3.6853,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011724275287072e-05,
+      "loss": 3.6724,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01088568053602e-05,
+      "loss": 3.6704,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010047085784968e-05,
+      "loss": 3.6821,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009208491033916e-05,
+      "loss": 3.6899,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008371534163237e-05,
+      "loss": 3.6901,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007532939412185e-05,
+      "loss": 3.6769,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006694344661133e-05,
+      "loss": 3.6738,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005855749910081e-05,
+      "loss": 3.6854,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0050187930394016e-05,
+      "loss": 3.6753,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0041801982883496e-05,
+      "loss": 3.6801,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0033416035372976e-05,
+      "loss": 3.691,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0025030087862456e-05,
+      "loss": 3.688,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.001666051915567e-05,
+      "loss": 3.6735,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000827457164515e-05,
+      "loss": 3.6846,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9025914669036865,
+      "eval_runtime": 586.0986,
+      "eval_samples_per_second": 651.07,
+      "eval_steps_per_second": 20.346,
+      "step": 610552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999988862413463e-05,
+      "loss": 3.6834,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999150267662411e-05,
+      "loss": 3.6901,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998311672911359e-05,
+      "loss": 3.6763,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9974730781603065e-05,
+      "loss": 3.6787,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996636121289628e-05,
+      "loss": 3.6713,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995797526538576e-05,
+      "loss": 3.6764,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994958931787524e-05,
+      "loss": 3.6795,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9941203370364714e-05,
+      "loss": 3.6604,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993283380165793e-05,
+      "loss": 3.6737,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992444785414741e-05,
+      "loss": 3.6868,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991606190663689e-05,
+      "loss": 3.6722,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990767595912637e-05,
+      "loss": 3.6679,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9899306390419585e-05,
+      "loss": 3.6827,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890920442909065e-05,
+      "loss": 3.67,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988253449539854e-05,
+      "loss": 3.6789,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987414854788802e-05,
+      "loss": 3.6548,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865778979181234e-05,
+      "loss": 3.653,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9857393031670714e-05,
+      "loss": 3.6743,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984900708416019e-05,
+      "loss": 3.6653,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984062113664967e-05,
+      "loss": 3.6917,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983225156794288e-05,
+      "loss": 3.6733,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982386562043236e-05,
+      "loss": 3.6795,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981547967292184e-05,
+      "loss": 3.6655,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980709372541132e-05,
+      "loss": 3.6779,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979872415670454e-05,
+      "loss": 3.6642,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979033820919401e-05,
+      "loss": 3.6721,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978195226168349e-05,
+      "loss": 3.6683,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977356631417297e-05,
+      "loss": 3.6753,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976519674546619e-05,
+      "loss": 3.6668,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975681079795566e-05,
+      "loss": 3.6601,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974842485044514e-05,
+      "loss": 3.6553,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974003890293462e-05,
+      "loss": 3.6725,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973166933422784e-05,
+      "loss": 3.6734,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972328338671732e-05,
+      "loss": 3.6716,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.97148974392068e-05,
+      "loss": 3.6708,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970651149169628e-05,
+      "loss": 3.6751,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9698141922989486e-05,
+      "loss": 3.6708,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9689755975478966e-05,
+      "loss": 3.6701,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9681370027968446e-05,
+      "loss": 3.6675,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967300045926166e-05,
+      "loss": 3.6665,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9664614511751135e-05,
+      "loss": 3.6707,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9656228564240615e-05,
+      "loss": 3.6554,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9647842616730095e-05,
+      "loss": 3.6725,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963947304802331e-05,
+      "loss": 3.6728,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9631087100512784e-05,
+      "loss": 3.6726,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962270115300227e-05,
+      "loss": 3.672,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.961431520549175e-05,
+      "loss": 3.6567,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960594563678496e-05,
+      "loss": 3.6795,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959755968927444e-05,
+      "loss": 3.658,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958917374176392e-05,
+      "loss": 3.6641,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95807877942534e-05,
+      "loss": 3.6566,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957241822554661e-05,
+      "loss": 3.6643,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956403227803609e-05,
+      "loss": 3.6699,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955564633052557e-05,
+      "loss": 3.6574,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954726038301505e-05,
+      "loss": 3.666,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953889081430826e-05,
+      "loss": 3.6578,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953050486679774e-05,
+      "loss": 3.6701,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9522118919287224e-05,
+      "loss": 3.66,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9513732971776704e-05,
+      "loss": 3.6732,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950536340306991e-05,
+      "loss": 3.6674,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949697745555939e-05,
+      "loss": 3.6599,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948859150804887e-05,
+      "loss": 3.6545,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948020556053835e-05,
+      "loss": 3.659,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947183599183156e-05,
+      "loss": 3.6695,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946345004432104e-05,
+      "loss": 3.6617,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945506409681052e-05,
+      "loss": 3.6592,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94466781493e-05,
+      "loss": 3.6564,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943829220178948e-05,
+      "loss": 3.6582,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942992263308269e-05,
+      "loss": 3.6567,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942153668557218e-05,
+      "loss": 3.6591,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941315073806166e-05,
+      "loss": 3.658,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940476479055114e-05,
+      "loss": 3.6591,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939639522184435e-05,
+      "loss": 3.6616,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938800927433383e-05,
+      "loss": 3.675,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937962332682331e-05,
+      "loss": 3.666,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937123737931279e-05,
+      "loss": 3.6612,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9362867810605996e-05,
+      "loss": 3.6625,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9354481863095476e-05,
+      "loss": 3.6616,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9346095915584956e-05,
+      "loss": 3.6533,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9337709968074436e-05,
+      "loss": 3.6714,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9329340399367645e-05,
+      "loss": 3.6679,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932095445185713e-05,
+      "loss": 3.6663,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931256850434661e-05,
+      "loss": 3.6478,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930419893563982e-05,
+      "loss": 3.6601,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92958129881293e-05,
+      "loss": 3.6465,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.928742704061878e-05,
+      "loss": 3.6646,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927904109310826e-05,
+      "loss": 3.6624,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927067152440147e-05,
+      "loss": 3.6556,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926228557689095e-05,
+      "loss": 3.6619,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925389962938043e-05,
+      "loss": 3.6627,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924551368186991e-05,
+      "loss": 3.6539,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923714411316312e-05,
+      "loss": 3.6558,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.92287581656526e-05,
+      "loss": 3.6623,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9220372218142085e-05,
+      "loss": 3.6501,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211986270631565e-05,
+      "loss": 3.6543,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9203616701924774e-05,
+      "loss": 3.6501,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9195230754414254e-05,
+      "loss": 3.6578,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9186844806903734e-05,
+      "loss": 3.6722,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9178458859393214e-05,
+      "loss": 3.655,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917008929068642e-05,
+      "loss": 3.6556,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91617033431759e-05,
+      "loss": 3.6549,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915331739566538e-05,
+      "loss": 3.6619,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914493144815486e-05,
+      "loss": 3.6617,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913656187944807e-05,
+      "loss": 3.651,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912817593193755e-05,
+      "loss": 3.6648,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911978998442704e-05,
+      "loss": 3.6426,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911140403691652e-05,
+      "loss": 3.6479,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910303446820973e-05,
+      "loss": 3.6596,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909464852069921e-05,
+      "loss": 3.6557,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908626257318869e-05,
+      "loss": 3.6514,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907787662567817e-05,
+      "loss": 3.6577,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906950705697138e-05,
+      "loss": 3.6332,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906112110946086e-05,
+      "loss": 3.6577,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905273516195034e-05,
+      "loss": 3.6636,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904434921443982e-05,
+      "loss": 3.657,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9035979645733026e-05,
+      "loss": 3.6453,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9027593698222506e-05,
+      "loss": 3.6507,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901920775071199e-05,
+      "loss": 3.6437,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901082180320147e-05,
+      "loss": 3.6611,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900245223449468e-05,
+      "loss": 3.6553,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899406628698416e-05,
+      "loss": 3.6498,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898568033947364e-05,
+      "loss": 3.6485,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897729439196312e-05,
+      "loss": 3.6606,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896892482325633e-05,
+      "loss": 3.6545,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896053887574581e-05,
+      "loss": 3.6476,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895215292823529e-05,
+      "loss": 3.6523,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894376698072477e-05,
+      "loss": 3.6548,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893539741201798e-05,
+      "loss": 3.6516,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892701146450746e-05,
+      "loss": 3.6587,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891862551699694e-05,
+      "loss": 3.6424,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8910239569486426e-05,
+      "loss": 3.6517,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8901870000779635e-05,
+      "loss": 3.6611,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8893484053269115e-05,
+      "loss": 3.6536,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8885098105758595e-05,
+      "loss": 3.6545,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8876712158248075e-05,
+      "loss": 3.6551,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8868342589541284e-05,
+      "loss": 3.6485,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8859956642030764e-05,
+      "loss": 3.6434,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8851570694520244e-05,
+      "loss": 3.6514,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8843184747009724e-05,
+      "loss": 3.6667,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883481517830293e-05,
+      "loss": 3.6617,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882642923079241e-05,
+      "loss": 3.6493,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881804328328189e-05,
+      "loss": 3.6511,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880965733577138e-05,
+      "loss": 3.6541,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880128776706459e-05,
+      "loss": 3.6502,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879290181955407e-05,
+      "loss": 3.6542,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878451587204355e-05,
+      "loss": 3.6598,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877612992453303e-05,
+      "loss": 3.6605,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876774397702251e-05,
+      "loss": 3.6474,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875935802951198e-05,
+      "loss": 3.6569,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.899163246154785,
+      "eval_runtime": 326.7628,
+      "eval_samples_per_second": 1167.792,
+      "eval_steps_per_second": 36.494,
+      "step": 686871
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.875097208200146e-05,
+      "loss": 3.6499,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874260251329468e-05,
+      "loss": 3.6632,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.873421656578416e-05,
+      "loss": 3.6515,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.872583061827363e-05,
+      "loss": 3.6497,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871744467076312e-05,
+      "loss": 3.6459,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8709075102056333e-05,
+      "loss": 3.6476,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870068915454581e-05,
+      "loss": 3.6516,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8692303207035287e-05,
+      "loss": 3.6366,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8683917259524766e-05,
+      "loss": 3.6433,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.867554769081798e-05,
+      "loss": 3.6615,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8667161743307456e-05,
+      "loss": 3.6437,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8658775795796935e-05,
+      "loss": 3.6393,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8650389848286415e-05,
+      "loss": 3.6592,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864202027957963e-05,
+      "loss": 3.643,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8633634332069105e-05,
+      "loss": 3.6522,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8625248384558584e-05,
+      "loss": 3.6296,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861686243704807e-05,
+      "loss": 3.6274,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.860849286834128e-05,
+      "loss": 3.6474,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.860010692083076e-05,
+      "loss": 3.639,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859172097332024e-05,
+      "loss": 3.6637,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858333502580972e-05,
+      "loss": 3.6456,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857496545710293e-05,
+      "loss": 3.6581,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.856657950959241e-05,
+      "loss": 3.6395,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855819356208189e-05,
+      "loss": 3.6465,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8549823993375105e-05,
+      "loss": 3.6364,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854143804586458e-05,
+      "loss": 3.6465,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853305209835406e-05,
+      "loss": 3.6445,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852466615084354e-05,
+      "loss": 3.6492,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8516280203333025e-05,
+      "loss": 3.6404,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8507894255822505e-05,
+      "loss": 3.6338,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8499508308311985e-05,
+      "loss": 3.6281,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8491122360801465e-05,
+      "loss": 3.6473,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8482752792094674e-05,
+      "loss": 3.6446,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8474366844584154e-05,
+      "loss": 3.6456,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8465980897073634e-05,
+      "loss": 3.646,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8457594949563114e-05,
+      "loss": 3.6501,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.844922538085632e-05,
+      "loss": 3.6445,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84408394333458e-05,
+      "loss": 3.6428,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843245348583528e-05,
+      "loss": 3.6454,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.842406753832476e-05,
+      "loss": 3.6422,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.841569796961798e-05,
+      "loss": 3.6416,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840731202210746e-05,
+      "loss": 3.6359,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839892607459694e-05,
+      "loss": 3.6423,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839054012708642e-05,
+      "loss": 3.6491,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838217055837963e-05,
+      "loss": 3.645,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837378461086911e-05,
+      "loss": 3.645,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836539866335859e-05,
+      "loss": 3.6306,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835701271584807e-05,
+      "loss": 3.6516,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8348643147141276e-05,
+      "loss": 3.6341,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8340257199630756e-05,
+      "loss": 3.6381,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8331871252120236e-05,
+      "loss": 3.6327,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8323485304609716e-05,
+      "loss": 3.6337,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831511573590293e-05,
+      "loss": 3.6498,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830672978839241e-05,
+      "loss": 3.6314,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.829834384088189e-05,
+      "loss": 3.6425,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828995789337137e-05,
+      "loss": 3.6274,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828158832466458e-05,
+      "loss": 3.6436,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.827320237715406e-05,
+      "loss": 3.6352,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826481642964354e-05,
+      "loss": 3.6476,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.825643048213302e-05,
+      "loss": 3.6458,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824806091342623e-05,
+      "loss": 3.6331,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823967496591571e-05,
+      "loss": 3.63,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823128901840519e-05,
+      "loss": 3.6316,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822290307089467e-05,
+      "loss": 3.639,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8214533502187886e-05,
+      "loss": 3.6409,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8206147554677366e-05,
+      "loss": 3.6337,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8197761607166846e-05,
+      "loss": 3.6313,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8189375659656326e-05,
+      "loss": 3.6341,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8181006090949535e-05,
+      "loss": 3.6277,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8172620143439015e-05,
+      "loss": 3.6346,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8164234195928495e-05,
+      "loss": 3.6323,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8155848248417975e-05,
+      "loss": 3.6311,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8147478679711184e-05,
+      "loss": 3.641,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8139092732200664e-05,
+      "loss": 3.6485,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8130706784690144e-05,
+      "loss": 3.6402,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8122320837179624e-05,
+      "loss": 3.6393,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811395126847284e-05,
+      "loss": 3.6345,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810556532096232e-05,
+      "loss": 3.6427,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80971793734518e-05,
+      "loss": 3.6275,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808879342594128e-05,
+      "loss": 3.6432,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808042385723449e-05,
+      "loss": 3.6452,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807203790972397e-05,
+      "loss": 3.642,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806365196221345e-05,
+      "loss": 3.6254,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805526601470293e-05,
+      "loss": 3.631,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804689644599614e-05,
+      "loss": 3.6198,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803851049848562e-05,
+      "loss": 3.6416,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80301245509751e-05,
+      "loss": 3.6365,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802173860346458e-05,
+      "loss": 3.6343,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8013369034757786e-05,
+      "loss": 3.6364,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.800498308724727e-05,
+      "loss": 3.6378,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.799659713973675e-05,
+      "loss": 3.6277,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798821119222623e-05,
+      "loss": 3.6299,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797984162351944e-05,
+      "loss": 3.6409,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797145567600892e-05,
+      "loss": 3.6225,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79630697284984e-05,
+      "loss": 3.6283,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.795468378098788e-05,
+      "loss": 3.6269,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.794631421228109e-05,
+      "loss": 3.6338,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793792826477057e-05,
+      "loss": 3.6461,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792954231726005e-05,
+      "loss": 3.6278,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792115636974953e-05,
+      "loss": 3.6336,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791278680104274e-05,
+      "loss": 3.6314,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.790440085353223e-05,
+      "loss": 3.6373,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789601490602171e-05,
+      "loss": 3.6348,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788762895851119e-05,
+      "loss": 3.6296,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7879259389804396e-05,
+      "loss": 3.6347,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7870873442293876e-05,
+      "loss": 3.621,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7862487494783356e-05,
+      "loss": 3.6239,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785410154727283e-05,
+      "loss": 3.6342,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7845731978566045e-05,
+      "loss": 3.6306,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7837346031055525e-05,
+      "loss": 3.6264,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7828960083545005e-05,
+      "loss": 3.6336,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782057413603448e-05,
+      "loss": 3.6125,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7812204567327694e-05,
+      "loss": 3.6302,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780381861981718e-05,
+      "loss": 3.6377,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.779543267230666e-05,
+      "loss": 3.6331,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7787046724796134e-05,
+      "loss": 3.6225,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777867715608935e-05,
+      "loss": 3.6238,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777029120857883e-05,
+      "loss": 3.6197,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77619052610683e-05,
+      "loss": 3.6311,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.775351931355778e-05,
+      "loss": 3.6311,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7745149744851e-05,
+      "loss": 3.6308,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.773676379734048e-05,
+      "loss": 3.6176,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772837784982995e-05,
+      "loss": 3.6398,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771999190231943e-05,
+      "loss": 3.626,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771162233361265e-05,
+      "loss": 3.6258,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7703236386102134e-05,
+      "loss": 3.6247,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769485043859161e-05,
+      "loss": 3.632,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768648086988482e-05,
+      "loss": 3.6273,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76780949223743e-05,
+      "loss": 3.6319,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7669708974863776e-05,
+      "loss": 3.6204,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7661323027353256e-05,
+      "loss": 3.6279,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765295345864647e-05,
+      "loss": 3.6394,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764456751113595e-05,
+      "loss": 3.6255,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7636181563625425e-05,
+      "loss": 3.6282,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7627795616114905e-05,
+      "loss": 3.635,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761942604740812e-05,
+      "loss": 3.6236,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76110400998976e-05,
+      "loss": 3.6227,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760265415238708e-05,
+      "loss": 3.6278,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759426820487656e-05,
+      "loss": 3.6344,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758589863616978e-05,
+      "loss": 3.6406,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.757751268865925e-05,
+      "loss": 3.6246,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756912674114873e-05,
+      "loss": 3.6293,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756074079363821e-05,
+      "loss": 3.6266,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7552371224931426e-05,
+      "loss": 3.627,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75439852774209e-05,
+      "loss": 3.6309,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753559932991038e-05,
+      "loss": 3.6346,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752721338239986e-05,
+      "loss": 3.6362,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751882743488934e-05,
+      "loss": 3.6235,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7510441487378826e-05,
+      "loss": 3.6319,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.897075891494751,
+      "eval_runtime": 324.3395,
+      "eval_samples_per_second": 1176.517,
+      "eval_steps_per_second": 36.767,
+      "step": 763190
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 4.947459670824653e+17,
+  "trial_name": null,
+  "trial_params": null
+}