diff --git "a/checkpoint-610560/trainer_state.json" "b/checkpoint-610560/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-610560/trainer_state.json"
@@ -0,0 +1,7241 @@
+{
+  "best_metric": 3.854868173599243,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/binding-case/transformer/2/checkpoints/checkpoint-610560",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 610560,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.9575,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8163,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1923,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9769,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8209,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7141,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6128,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5406,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.459,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4141,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3519,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3101,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2647,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2178,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1785,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1386,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1091,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0779,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0521,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0219,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0079,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.98,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9504,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9376,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9266,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.8884,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8712,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.859,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8438,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8202,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.804,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7996,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7845,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7783,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7637,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7439,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7372,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.7037,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.711,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6823,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6828,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6697,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6727,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6507,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6475,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6369,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6209,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6323,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.5989,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.5909,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.6016,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.598,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.5725,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5671,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 4.5663,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 4.5485,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 4.5686,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5298,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5394,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5305,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.515,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5236,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5041,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.5024,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4918,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5076,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.4788,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4822,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4619,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9421582546222625e-05,
+      "loss": 4.4715,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4816,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4713,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4535,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4665,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4661,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4423,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.4472,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.4286,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4195,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.417,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.4191,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.4248,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4312,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 4.4192,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.3957,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.4024,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.4029,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 4.3938,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.3876,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925392911122715e-05,
+      "loss": 4.383,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 4.3772,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 4.3863,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 4.3715,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922038532118507e-05,
+      "loss": 4.3788,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921199937367455e-05,
+      "loss": 4.3683,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920361342616403e-05,
+      "loss": 4.368,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919522747865351e-05,
+      "loss": 4.3522,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918684153114299e-05,
+      "loss": 4.3541,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917845558363247e-05,
+      "loss": 4.3515,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917006963612195e-05,
+      "loss": 4.3581,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.3588,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.3347,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.3427,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136542224883596e-05,
+      "loss": 4.3336,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.337,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.3251,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.3227,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.3279,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.3178,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.3161,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907787334991742e-05,
+      "loss": 4.3094,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90694874024069e-05,
+      "loss": 4.3053,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3189,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3045,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.305,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.3091,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.3136,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901920447495124e-05,
+      "loss": 4.2986,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901081852744072e-05,
+      "loss": 4.2873,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.2933,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.2994,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.2851,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977291116202375e-05,
+      "loss": 4.284,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968905168691855e-05,
+      "loss": 4.2826,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.2827,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.2755,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.2801,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.2719,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.2742,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918605862432463e-05,
+      "loss": 4.2663,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.2841,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.2673,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.2602,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.2543,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.2647,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886830655617308e-05,
+      "loss": 4.2594,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.266,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.2528,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884316509244525e-05,
+      "loss": 4.2583,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883477914493473e-05,
+      "loss": 4.2514,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.2514,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.2431,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.2512,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.2393,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.2275,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878447983867534e-05,
+      "loss": 4.2481,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.2362,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.2334,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.2397,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8750952427436986e-05,
+      "loss": 4.2236,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.196439266204834,
+      "eval_runtime": 309.473,
+      "eval_samples_per_second": 1233.035,
+      "eval_steps_per_second": 38.533,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8742566479926466e-05,
+      "loss": 4.2129,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.2179,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8725794584905426e-05,
+      "loss": 4.2222,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.2174,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.2258,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700636742373866e-05,
+      "loss": 4.2104,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.2133,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683881226156555e-05,
+      "loss": 4.2036,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867551165744977e-05,
+      "loss": 4.2189,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866712570993925e-05,
+      "loss": 4.1999,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865873976242873e-05,
+      "loss": 4.199,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865035381491821e-05,
+      "loss": 4.2005,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864198424621142e-05,
+      "loss": 4.1963,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86335982987009e-05,
+      "loss": 4.1888,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862521235119038e-05,
+      "loss": 4.189,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.1737,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860844045616934e-05,
+      "loss": 4.1849,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860005450865882e-05,
+      "loss": 4.1797,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85916685611483e-05,
+      "loss": 4.1727,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858328261363778e-05,
+      "loss": 4.2005,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.1775,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.1798,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.1756,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549771581203164e-05,
+      "loss": 4.1893,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.162,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532999686182124e-05,
+      "loss": 4.1664,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524613738671604e-05,
+      "loss": 4.169,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8516227791161084e-05,
+      "loss": 4.1677,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.1568,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.1577,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.1647,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.1673,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847433081121595e-05,
+      "loss": 4.1675,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846594486370543e-05,
+      "loss": 4.1581,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845755891619491e-05,
+      "loss": 4.1516,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844917296868439e-05,
+      "loss": 4.1619,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84408033999776e-05,
+      "loss": 4.1349,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.1557,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842403150495656e-05,
+      "loss": 4.1332,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841564555744604e-05,
+      "loss": 4.1436,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840727598873925e-05,
+      "loss": 4.1384,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839889004122873e-05,
+      "loss": 4.148,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839050409371821e-05,
+      "loss": 4.1409,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382118146207687e-05,
+      "loss": 4.1353,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83737485775009e-05,
+      "loss": 4.1402,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836536262999038e-05,
+      "loss": 4.1325,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835697668247986e-05,
+      "loss": 4.1446,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834859073496934e-05,
+      "loss": 4.1184,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834022116626255e-05,
+      "loss": 4.1189,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833183521875203e-05,
+      "loss": 4.1362,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832344927124151e-05,
+      "loss": 4.1353,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831506332373099e-05,
+      "loss": 4.121,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830671013382793e-05,
+      "loss": 4.1154,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829832418631741e-05,
+      "loss": 4.1243,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828993823880689e-05,
+      "loss": 4.1074,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828155229129637e-05,
+      "loss": 4.1329,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273166343785856e-05,
+      "loss": 4.1113,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264780396275336e-05,
+      "loss": 4.1208,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256394448764816e-05,
+      "loss": 4.1154,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248008501254296e-05,
+      "loss": 4.1034,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239638932547505e-05,
+      "loss": 4.1184,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231252985036985e-05,
+      "loss": 4.104,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222867037526465e-05,
+      "loss": 4.1039,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214481090015945e-05,
+      "loss": 4.0994,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206111521309154e-05,
+      "loss": 4.1166,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197725573798634e-05,
+      "loss": 4.0986,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189339626288114e-05,
+      "loss": 4.0993,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180953678777594e-05,
+      "loss": 4.0893,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172567731267074e-05,
+      "loss": 4.0989,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164181783756554e-05,
+      "loss": 4.1119,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155795836246034e-05,
+      "loss": 4.1059,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147409888735514e-05,
+      "loss": 4.0958,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813904032002872e-05,
+      "loss": 4.107,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81306543725182e-05,
+      "loss": 4.1093,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812226842500768e-05,
+      "loss": 4.0949,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811388247749716e-05,
+      "loss": 4.0996,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810551290879037e-05,
+      "loss": 4.0892,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809712696127985e-05,
+      "loss": 4.0808,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808874101376933e-05,
+      "loss": 4.0827,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808035506625881e-05,
+      "loss": 4.086,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807198549755203e-05,
+      "loss": 4.0913,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806359955004151e-05,
+      "loss": 4.1052,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805521360253099e-05,
+      "loss": 4.095,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804682765502047e-05,
+      "loss": 4.0718,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8038458086313677e-05,
+      "loss": 4.0822,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8030072138803156e-05,
+      "loss": 4.0873,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8021686191292636e-05,
+      "loss": 4.0833,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8013300243782116e-05,
+      "loss": 4.0743,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8004930675075325e-05,
+      "loss": 4.0733,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996544727564805e-05,
+      "loss": 4.0748,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988175158858015e-05,
+      "loss": 4.0787,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979789211347494e-05,
+      "loss": 4.0708,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797140326383698e-05,
+      "loss": 4.0804,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796301731632646e-05,
+      "loss": 4.073,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795464774761967e-05,
+      "loss": 4.0775,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794626180010915e-05,
+      "loss": 4.0609,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793787585259863e-05,
+      "loss": 4.0701,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792948990508811e-05,
+      "loss": 4.0661,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792110395757759e-05,
+      "loss": 4.0726,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791271801006707e-05,
+      "loss": 4.0729,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790433206255655e-05,
+      "loss": 4.0603,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789594611504603e-05,
+      "loss": 4.0616,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788757654633924e-05,
+      "loss": 4.0621,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787919059882872e-05,
+      "loss": 4.0639,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870821030121935e-05,
+      "loss": 4.0503,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862435082611415e-05,
+      "loss": 4.0549,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854049135100895e-05,
+      "loss": 4.0632,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845663187590375e-05,
+      "loss": 4.0515,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837293618883584e-05,
+      "loss": 4.0513,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828907671373064e-05,
+      "loss": 4.0477,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820521723862544e-05,
+      "loss": 4.0453,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812135776352024e-05,
+      "loss": 4.0653,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803749828841504e-05,
+      "loss": 4.0457,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795363881330984e-05,
+      "loss": 4.0568,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786977933820464e-05,
+      "loss": 4.0522,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777860836511367e-05,
+      "loss": 4.0664,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777022241760315e-05,
+      "loss": 4.0522,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776183647009263e-05,
+      "loss": 4.0384,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775345052258212e-05,
+      "loss": 4.0475,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77450645750716e-05,
+      "loss": 4.0526,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773669500636481e-05,
+      "loss": 4.0372,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772830905885429e-05,
+      "loss": 4.0446,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771992311134377e-05,
+      "loss": 4.0379,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771153716383325e-05,
+      "loss": 4.0496,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770316759512646e-05,
+      "loss": 4.0348,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769478164761594e-05,
+      "loss": 4.0468,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768639570010542e-05,
+      "loss": 4.0389,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76780097525949e-05,
+      "loss": 4.0422,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766962380508437e-05,
+      "loss": 4.0327,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766123785757386e-05,
+      "loss": 4.0555,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765286828886707e-05,
+      "loss": 4.0427,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764448234135655e-05,
+      "loss": 4.0289,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7636096393846026e-05,
+      "loss": 4.0308,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7627710446335506e-05,
+      "loss": 4.0373,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761934087762872e-05,
+      "loss": 4.0389,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610954930118195e-05,
+      "loss": 4.0461,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7602568982607675e-05,
+      "loss": 4.0297,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7594183035097155e-05,
+      "loss": 4.0397,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758581346639037e-05,
+      "loss": 4.0307,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577427518879844e-05,
+      "loss": 4.0368,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569041571369324e-05,
+      "loss": 4.0298,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756065562385881e-05,
+      "loss": 4.0345,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755228605515202e-05,
+      "loss": 4.0268,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75439001076415e-05,
+      "loss": 4.0195,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753551416013098e-05,
+      "loss": 4.04,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752712821262046e-05,
+      "loss": 4.0292,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751874226510994e-05,
+      "loss": 4.0257,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751037269640315e-05,
+      "loss": 4.0288,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750198674889263e-05,
+      "loss": 4.0182,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.027224063873291,
+      "eval_runtime": 305.3242,
+      "eval_samples_per_second": 1249.789,
+      "eval_steps_per_second": 39.057,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749360080138211e-05,
+      "loss": 4.0098,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748521485387159e-05,
+      "loss": 4.0111,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747682890636107e-05,
+      "loss": 4.0251,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746844295885055e-05,
+      "loss": 4.0158,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746005701134003e-05,
+      "loss": 4.0294,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745167106382951e-05,
+      "loss": 4.0134,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744328511631899e-05,
+      "loss": 4.0152,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434899168808475e-05,
+      "loss": 4.0082,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426529600101684e-05,
+      "loss": 4.0265,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418143652591164e-05,
+      "loss": 4.0078,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409757705080644e-05,
+      "loss": 4.0053,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401371757570124e-05,
+      "loss": 4.0077,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739300218886333e-05,
+      "loss": 4.0081,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738461624135281e-05,
+      "loss": 4.0003,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737623029384229e-05,
+      "loss": 4.0039,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736784434633177e-05,
+      "loss": 3.9932,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735947477762498e-05,
+      "loss": 3.9976,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735108883011446e-05,
+      "loss": 3.9945,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734270288260394e-05,
+      "loss": 3.9884,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733431693509343e-05,
+      "loss": 4.0224,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732594736638664e-05,
+      "loss": 3.9986,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731756141887612e-05,
+      "loss": 3.9929,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73091754713656e-05,
+      "loss": 3.9999,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730078952385508e-05,
+      "loss": 4.0076,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729241995514829e-05,
+      "loss": 3.9851,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728403400763777e-05,
+      "loss": 3.9938,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727564806012725e-05,
+      "loss": 3.9928,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726726211261673e-05,
+      "loss": 3.9965,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258892543909936e-05,
+      "loss": 3.9811,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250506596399416e-05,
+      "loss": 3.9873,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7242120648888896e-05,
+      "loss": 3.9894,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723373470137838e-05,
+      "loss": 3.997,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722536513267159e-05,
+      "loss": 4.0023,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721697918516107e-05,
+      "loss": 3.9893,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720859323765055e-05,
+      "loss": 3.9825,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720020729014003e-05,
+      "loss": 3.9935,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719183772143324e-05,
+      "loss": 3.9746,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718345177392272e-05,
+      "loss": 3.9892,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71750658264122e-05,
+      "loss": 3.9705,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716667987890168e-05,
+      "loss": 3.9765,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158293931391154e-05,
+      "loss": 3.9787,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714992436268437e-05,
+      "loss": 3.9888,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714153841517385e-05,
+      "loss": 3.9801,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713315246766333e-05,
+      "loss": 3.9762,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712476652015281e-05,
+      "loss": 3.9791,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116396951446025e-05,
+      "loss": 3.9752,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7108011003935505e-05,
+      "loss": 3.984,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709962505642498e-05,
+      "loss": 3.9636,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709123910891446e-05,
+      "loss": 3.9604,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082869540207674e-05,
+      "loss": 3.9823,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074483592697154e-05,
+      "loss": 3.9769,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706609764518663e-05,
+      "loss": 3.9664,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705771169767611e-05,
+      "loss": 3.9672,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704934212896932e-05,
+      "loss": 3.9708,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70409561814588e-05,
+      "loss": 3.9555,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703257023394828e-05,
+      "loss": 3.9779,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702418428643776e-05,
+      "loss": 3.9625,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701581471773098e-05,
+      "loss": 3.9699,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700742877022045e-05,
+      "loss": 3.9636,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699904282270993e-05,
+      "loss": 3.9611,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699065687519941e-05,
+      "loss": 3.9607,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698228730649263e-05,
+      "loss": 3.9604,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69739013589821e-05,
+      "loss": 3.956,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696551541147158e-05,
+      "loss": 3.9535,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695712946396106e-05,
+      "loss": 3.9703,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694875989525428e-05,
+      "loss": 3.9555,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694037394774376e-05,
+      "loss": 3.955,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693198800023324e-05,
+      "loss": 3.9474,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692360205272272e-05,
+      "loss": 3.954,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915232484015926e-05,
+      "loss": 3.9702,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906846536505406e-05,
+      "loss": 3.9651,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898460588994886e-05,
+      "loss": 3.9573,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6890074641484366e-05,
+      "loss": 3.9658,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881705072777575e-05,
+      "loss": 3.9692,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873319125267055e-05,
+      "loss": 3.9518,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864933177756535e-05,
+      "loss": 3.9608,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856547230246015e-05,
+      "loss": 3.9545,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684817766153923e-05,
+      "loss": 3.9457,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683979171402871e-05,
+      "loss": 3.9528,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683140576651819e-05,
+      "loss": 3.9461,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682301981900767e-05,
+      "loss": 3.9568,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681465025030088e-05,
+      "loss": 3.9639,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680626430279036e-05,
+      "loss": 3.9617,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679787835527984e-05,
+      "loss": 3.9356,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678949240776932e-05,
+      "loss": 3.9517,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678112283906253e-05,
+      "loss": 3.9515,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677273689155201e-05,
+      "loss": 3.9505,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676435094404149e-05,
+      "loss": 3.9458,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675596499653097e-05,
+      "loss": 3.9379,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747595427824184e-05,
+      "loss": 3.9477,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739209480313664e-05,
+      "loss": 3.9448,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730823532803144e-05,
+      "loss": 3.9417,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722437585292624e-05,
+      "loss": 3.9501,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671406801658583e-05,
+      "loss": 3.9414,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670568206907531e-05,
+      "loss": 3.9504,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669729612156479e-05,
+      "loss": 3.9367,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668891017405427e-05,
+      "loss": 3.9379,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668054060534748e-05,
+      "loss": 3.9386,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667215465783696e-05,
+      "loss": 3.949,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666376871032644e-05,
+      "loss": 3.9474,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665538276281592e-05,
+      "loss": 3.9359,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664701319410914e-05,
+      "loss": 3.9364,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663862724659862e-05,
+      "loss": 3.9363,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66302412990881e-05,
+      "loss": 3.9415,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662187173038131e-05,
+      "loss": 3.9242,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661348578287079e-05,
+      "loss": 3.9287,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660509983536027e-05,
+      "loss": 3.9433,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659671388784975e-05,
+      "loss": 3.9341,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588344319142956e-05,
+      "loss": 3.9266,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579958371632436e-05,
+      "loss": 3.9268,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571572424121916e-05,
+      "loss": 3.9248,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563186476611396e-05,
+      "loss": 3.9405,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554800529100876e-05,
+      "loss": 3.9306,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546414581590356e-05,
+      "loss": 3.9319,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6538028634079836e-05,
+      "loss": 3.9336,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652964268656932e-05,
+      "loss": 3.9483,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652127311786253e-05,
+      "loss": 3.9361,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651288717035201e-05,
+      "loss": 3.9203,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650450122284149e-05,
+      "loss": 3.9299,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649611527533097e-05,
+      "loss": 3.9321,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648774570662418e-05,
+      "loss": 3.9229,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647935975911366e-05,
+      "loss": 3.9252,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647097381160314e-05,
+      "loss": 3.9213,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646258786409262e-05,
+      "loss": 3.9332,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645421829538583e-05,
+      "loss": 3.9178,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644583234787531e-05,
+      "loss": 3.9335,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643744640036479e-05,
+      "loss": 3.9183,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429060452854276e-05,
+      "loss": 3.9283,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420690884147485e-05,
+      "loss": 3.9182,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412304936636965e-05,
+      "loss": 3.9396,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403918989126445e-05,
+      "loss": 3.9309,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395549420419654e-05,
+      "loss": 3.9203,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387163472909134e-05,
+      "loss": 3.9135,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378777525398614e-05,
+      "loss": 3.9265,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370391577888094e-05,
+      "loss": 3.9232,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6362005630377574e-05,
+      "loss": 3.9341,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635363606167078e-05,
+      "loss": 3.9164,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345266492964e-05,
+      "loss": 3.93,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633688054545348e-05,
+      "loss": 3.9227,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632849459794296e-05,
+      "loss": 3.9316,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632010865043244e-05,
+      "loss": 3.9156,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631172270292192e-05,
+      "loss": 3.9236,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63033367554114e-05,
+      "loss": 3.9186,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629495080790088e-05,
+      "loss": 3.9105,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628656486039035e-05,
+      "loss": 3.9276,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627819529168357e-05,
+      "loss": 3.9248,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626980934417305e-05,
+      "loss": 3.9113,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626142339666253e-05,
+      "loss": 3.9212,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6253037449152e-05,
+      "loss": 3.91,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.953357696533203,
+      "eval_runtime": 301.6217,
+      "eval_samples_per_second": 1265.131,
+      "eval_steps_per_second": 39.536,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244667880445217e-05,
+      "loss": 3.9062,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6236281932934697e-05,
+      "loss": 3.9047,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227895985424176e-05,
+      "loss": 3.9162,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219510037913656e-05,
+      "loss": 3.9135,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621114046920687e-05,
+      "loss": 3.9271,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620275452169635e-05,
+      "loss": 3.9031,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194368574185825e-05,
+      "loss": 3.9113,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185982626675305e-05,
+      "loss": 3.9041,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617761305796852e-05,
+      "loss": 3.925,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6169227110458e-05,
+      "loss": 3.9036,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616085754175121e-05,
+      "loss": 3.8994,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615247159424069e-05,
+      "loss": 3.9092,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614408564673017e-05,
+      "loss": 3.9013,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613569969921965e-05,
+      "loss": 3.8995,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612731375170913e-05,
+      "loss": 3.9013,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6118944183002346e-05,
+      "loss": 3.8948,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110558235491826e-05,
+      "loss": 3.8934,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61021722879813e-05,
+      "loss": 3.8953,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609378634047078e-05,
+      "loss": 3.8902,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608540039296026e-05,
+      "loss": 3.92,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607701444544974e-05,
+      "loss": 3.8973,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606862849793922e-05,
+      "loss": 3.8964,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60602425504287e-05,
+      "loss": 3.9002,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605187298172191e-05,
+      "loss": 3.9047,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604348703421139e-05,
+      "loss": 3.8897,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6035101086700875e-05,
+      "loss": 3.897,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026715139190355e-05,
+      "loss": 3.893,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018345570483564e-05,
+      "loss": 3.8964,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009959622973044e-05,
+      "loss": 3.8854,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001573675462524e-05,
+      "loss": 3.8903,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5993187727952004e-05,
+      "loss": 3.8954,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598481815924521e-05,
+      "loss": 3.899,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597643221173469e-05,
+      "loss": 3.9051,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59680626430279e-05,
+      "loss": 3.8968,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595967669551738e-05,
+      "loss": 3.8849,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595129074800686e-05,
+      "loss": 3.8964,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594290480049634e-05,
+      "loss": 3.883,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593451885298583e-05,
+      "loss": 3.8916,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592613290547531e-05,
+      "loss": 3.8777,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591776333676852e-05,
+      "loss": 3.8825,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5909377389258e-05,
+      "loss": 3.8812,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590099144174748e-05,
+      "loss": 3.898,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589260549423696e-05,
+      "loss": 3.8904,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588421954672644e-05,
+      "loss": 3.8821,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587583359921592e-05,
+      "loss": 3.886,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58674476517054e-05,
+      "loss": 3.8833,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585906170419488e-05,
+      "loss": 3.8889,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5850692135488086e-05,
+      "loss": 3.873,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5842322566781295e-05,
+      "loss": 3.8667,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5833936619270775e-05,
+      "loss": 3.8904,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582555067176026e-05,
+      "loss": 3.8861,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581716472424974e-05,
+      "loss": 3.8723,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580877877673922e-05,
+      "loss": 3.8786,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580040920803243e-05,
+      "loss": 3.8827,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579202326052191e-05,
+      "loss": 3.8661,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578363731301139e-05,
+      "loss": 3.8822,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577525136550087e-05,
+      "loss": 3.8713,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576688179679408e-05,
+      "loss": 3.8842,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575849584928356e-05,
+      "loss": 3.8719,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575010990177304e-05,
+      "loss": 3.8804,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574172395426252e-05,
+      "loss": 3.8679,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733338006752e-05,
+      "loss": 3.8745,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572495205924148e-05,
+      "loss": 3.8641,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716582490534696e-05,
+      "loss": 3.8689,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5708196543024176e-05,
+      "loss": 3.8849,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699810595513656e-05,
+      "loss": 3.8659,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691424648003135e-05,
+      "loss": 3.8705,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568303870049261e-05,
+      "loss": 3.8608,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567465275298209e-05,
+      "loss": 3.8688,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566626680547157e-05,
+      "loss": 3.8824,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565789723676478e-05,
+      "loss": 3.8771,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649527668057994e-05,
+      "loss": 3.8752,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5641141720547474e-05,
+      "loss": 3.8805,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632755773036953e-05,
+      "loss": 3.8786,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624369825526433e-05,
+      "loss": 3.8691,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561600025681965e-05,
+      "loss": 3.8746,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560761430930913e-05,
+      "loss": 3.8719,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559922836179861e-05,
+      "loss": 3.863,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559084241428808e-05,
+      "loss": 3.8651,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55824728455813e-05,
+      "loss": 3.8636,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557408689807078e-05,
+      "loss": 3.8696,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556570095056025e-05,
+      "loss": 3.8793,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555731500304973e-05,
+      "loss": 3.8767,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554892905553921e-05,
+      "loss": 3.8517,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554055948683243e-05,
+      "loss": 3.8726,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553217353932191e-05,
+      "loss": 3.8671,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552378759181139e-05,
+      "loss": 3.8674,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551540164430087e-05,
+      "loss": 3.8622,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550701569679035e-05,
+      "loss": 3.8541,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498646128083556e-05,
+      "loss": 3.8689,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5490260180573036e-05,
+      "loss": 3.861,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481874233062516e-05,
+      "loss": 3.8615,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5473488285551996e-05,
+      "loss": 3.8657,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5465102338041476e-05,
+      "loss": 3.8605,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5456732769334685e-05,
+      "loss": 3.8724,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5448346821824165e-05,
+      "loss": 3.8577,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5439960874313645e-05,
+      "loss": 3.8544,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431574926803125e-05,
+      "loss": 3.8573,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542320535809634e-05,
+      "loss": 3.8709,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541481941058582e-05,
+      "loss": 3.8635,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54064334630753e-05,
+      "loss": 3.8569,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539804751556478e-05,
+      "loss": 3.8548,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538966156805426e-05,
+      "loss": 3.8575,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538129199934747e-05,
+      "loss": 3.8659,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537290605183695e-05,
+      "loss": 3.8505,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536452010432643e-05,
+      "loss": 3.8445,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535613415681591e-05,
+      "loss": 3.8643,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534774820930539e-05,
+      "loss": 3.852,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53393786405986e-05,
+      "loss": 3.8482,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533099269308808e-05,
+      "loss": 3.8463,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532260674557756e-05,
+      "loss": 3.8507,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314220798067045e-05,
+      "loss": 3.8642,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305834850556525e-05,
+      "loss": 3.8487,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297465281849734e-05,
+      "loss": 3.8543,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5289079334339214e-05,
+      "loss": 3.8568,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5280693386828694e-05,
+      "loss": 3.8679,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5272307439318174e-05,
+      "loss": 3.8627,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526393787061138e-05,
+      "loss": 3.8415,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525555192310086e-05,
+      "loss": 3.8542,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524716597559034e-05,
+      "loss": 3.8508,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523878002807982e-05,
+      "loss": 3.8501,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52303940805693e-05,
+      "loss": 3.8499,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522200813305878e-05,
+      "loss": 3.844,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213638564352e-05,
+      "loss": 3.8566,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520525261684148e-05,
+      "loss": 3.8446,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519686666933096e-05,
+      "loss": 3.8557,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518848072182044e-05,
+      "loss": 3.8461,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518009477430992e-05,
+      "loss": 3.8532,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517172520560313e-05,
+      "loss": 3.8416,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516333925809261e-05,
+      "loss": 3.8682,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515495331058209e-05,
+      "loss": 3.8497,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514656736307156e-05,
+      "loss": 3.8494,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513819779436478e-05,
+      "loss": 3.8377,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512981184685426e-05,
+      "loss": 3.8546,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512142589934374e-05,
+      "loss": 3.8474,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511303995183322e-05,
+      "loss": 3.8576,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510467038312643e-05,
+      "loss": 3.8435,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509628443561591e-05,
+      "loss": 3.8546,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508789848810539e-05,
+      "loss": 3.8506,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079512540594866e-05,
+      "loss": 3.8611,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507114297188808e-05,
+      "loss": 3.8391,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506275702437756e-05,
+      "loss": 3.8524,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5054371076867035e-05,
+      "loss": 3.845,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5045985129356515e-05,
+      "loss": 3.8419,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5037599181845995e-05,
+      "loss": 3.8521,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5029213234335474e-05,
+      "loss": 3.8522,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020827286824954e-05,
+      "loss": 3.842,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501245771811817e-05,
+      "loss": 3.8457,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500407177060765e-05,
+      "loss": 3.8414,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.911860942840576,
+      "eval_runtime": 302.0292,
+      "eval_samples_per_second": 1263.424,
+      "eval_steps_per_second": 39.483,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499568582309713e-05,
+      "loss": 3.8349,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498729987558661e-05,
+      "loss": 3.831,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497891392807609e-05,
+      "loss": 3.8433,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497052798056557e-05,
+      "loss": 3.842,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496215841185878e-05,
+      "loss": 3.8543,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495377246434826e-05,
+      "loss": 3.8353,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494538651683774e-05,
+      "loss": 3.8369,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493700056932722e-05,
+      "loss": 3.8346,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49286146218167e-05,
+      "loss": 3.8518,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492022867430618e-05,
+      "loss": 3.8372,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491185910559939e-05,
+      "loss": 3.8278,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490347315808887e-05,
+      "loss": 3.8365,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4895087210578355e-05,
+      "loss": 3.8304,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886701263067835e-05,
+      "loss": 3.8302,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878315315557315e-05,
+      "loss": 3.8351,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869945746850524e-05,
+      "loss": 3.8205,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861559799340004e-05,
+      "loss": 3.8269,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4853173851829484e-05,
+      "loss": 3.8261,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844787904318964e-05,
+      "loss": 3.8217,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4836401956808444e-05,
+      "loss": 3.8478,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482803238810165e-05,
+      "loss": 3.8333,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481964644059113e-05,
+      "loss": 3.8265,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481126049308061e-05,
+      "loss": 3.8317,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480287454557009e-05,
+      "loss": 3.8388,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479448859805957e-05,
+      "loss": 3.8223,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478611902935279e-05,
+      "loss": 3.8308,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477773308184227e-05,
+      "loss": 3.8242,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476934713433175e-05,
+      "loss": 3.8267,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476096118682122e-05,
+      "loss": 3.8183,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47525752393107e-05,
+      "loss": 3.8253,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474418929180018e-05,
+      "loss": 3.8287,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47358197230934e-05,
+      "loss": 3.8319,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472743377558287e-05,
+      "loss": 3.8371,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471904782807235e-05,
+      "loss": 3.8315,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471066188056183e-05,
+      "loss": 3.8178,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4702292311855046e-05,
+      "loss": 3.8291,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693906364344526e-05,
+      "loss": 3.8209,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685520416834006e-05,
+      "loss": 3.824,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4677134469323486e-05,
+      "loss": 3.8136,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668764900616695e-05,
+      "loss": 3.818,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660378953106175e-05,
+      "loss": 3.8179,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651993005595655e-05,
+      "loss": 3.8309,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643607058085135e-05,
+      "loss": 3.8228,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4635221110574615e-05,
+      "loss": 3.8183,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626835163064095e-05,
+      "loss": 3.8196,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618465594357304e-05,
+      "loss": 3.8202,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4610079646846784e-05,
+      "loss": 3.8255,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4601693699336264e-05,
+      "loss": 3.8105,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4593307751825744e-05,
+      "loss": 3.7996,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458493818311896e-05,
+      "loss": 3.8249,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457655223560844e-05,
+      "loss": 3.8213,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456816628809792e-05,
+      "loss": 3.8074,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45597803405874e-05,
+      "loss": 3.8142,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455141077188061e-05,
+      "loss": 3.8205,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454302482437009e-05,
+      "loss": 3.8004,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453463887685957e-05,
+      "loss": 3.8193,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452625292934905e-05,
+      "loss": 3.81,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451786698183853e-05,
+      "loss": 3.8203,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450949741313174e-05,
+      "loss": 3.8087,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450111146562122e-05,
+      "loss": 3.8149,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44927255181107e-05,
+      "loss": 3.8045,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448433957060018e-05,
+      "loss": 3.8143,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475953623089664e-05,
+      "loss": 3.7994,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446758405438287e-05,
+      "loss": 3.8092,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445919810687235e-05,
+      "loss": 3.819,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445081215936183e-05,
+      "loss": 3.8076,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444242621185131e-05,
+      "loss": 3.8061,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443404026434079e-05,
+      "loss": 3.7989,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442565431683027e-05,
+      "loss": 3.8034,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441726836931975e-05,
+      "loss": 3.8155,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440889880061296e-05,
+      "loss": 3.8187,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440051285310244e-05,
+      "loss": 3.8136,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439212690559192e-05,
+      "loss": 3.8205,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43837409580814e-05,
+      "loss": 3.8195,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437537138937462e-05,
+      "loss": 3.8058,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43669854418641e-05,
+      "loss": 3.811,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435859949435358e-05,
+      "loss": 3.8145,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435021354684306e-05,
+      "loss": 3.8009,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434182759933253e-05,
+      "loss": 3.8039,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433345803062575e-05,
+      "loss": 3.8069,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432507208311523e-05,
+      "loss": 3.8067,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431668613560471e-05,
+      "loss": 3.8206,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430830018809418e-05,
+      "loss": 3.8162,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299930619387396e-05,
+      "loss": 3.792,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291544671876876e-05,
+      "loss": 3.8061,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4283158724366356e-05,
+      "loss": 3.8104,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274772776855836e-05,
+      "loss": 3.8091,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426640320814905e-05,
+      "loss": 3.804,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425801726063853e-05,
+      "loss": 3.7925,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249631313128005e-05,
+      "loss": 3.8065,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4241245365617485e-05,
+      "loss": 3.8014,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42328757969107e-05,
+      "loss": 3.8024,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422448984940018e-05,
+      "loss": 3.8028,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216103901889654e-05,
+      "loss": 3.8051,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207717954379134e-05,
+      "loss": 3.8079,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419934838567235e-05,
+      "loss": 3.8056,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419096243816182e-05,
+      "loss": 3.791,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418257649065131e-05,
+      "loss": 3.7997,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417419054314079e-05,
+      "loss": 3.8066,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416580459563027e-05,
+      "loss": 3.8099,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415741864811975e-05,
+      "loss": 3.7962,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414904907941296e-05,
+      "loss": 3.7985,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414066313190244e-05,
+      "loss": 3.7995,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413227718439192e-05,
+      "loss": 3.8069,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41238912368814e-05,
+      "loss": 3.7947,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411552166817461e-05,
+      "loss": 3.7915,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410713572066409e-05,
+      "loss": 3.7997,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409874977315357e-05,
+      "loss": 3.7957,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409036382564305e-05,
+      "loss": 3.7905,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408199425693626e-05,
+      "loss": 3.7916,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407360830942574e-05,
+      "loss": 3.7908,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406522236191522e-05,
+      "loss": 3.8077,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40568364144047e-05,
+      "loss": 3.7952,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404846684569791e-05,
+      "loss": 3.7946,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404008089818739e-05,
+      "loss": 3.801,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403169495067687e-05,
+      "loss": 3.8133,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402330900316635e-05,
+      "loss": 3.8047,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401493943445956e-05,
+      "loss": 3.7874,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400655348694904e-05,
+      "loss": 3.7943,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399816753943852e-05,
+      "loss": 3.7948,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3989781591928e-05,
+      "loss": 3.7897,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398141202322122e-05,
+      "loss": 3.7949,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39730260757107e-05,
+      "loss": 3.7865,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396464012820018e-05,
+      "loss": 3.799,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3956254180689657e-05,
+      "loss": 3.793,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947884611982866e-05,
+      "loss": 3.7984,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3939498664472346e-05,
+      "loss": 3.788,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3931112716961826e-05,
+      "loss": 3.7929,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922726769451306e-05,
+      "loss": 3.7937,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3914357200744515e-05,
+      "loss": 3.8066,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905971253233995e-05,
+      "loss": 3.7925,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897585305723475e-05,
+      "loss": 3.7981,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3889199358212955e-05,
+      "loss": 3.7837,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388082978950617e-05,
+      "loss": 3.8004,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387244384199565e-05,
+      "loss": 3.7934,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386405789448513e-05,
+      "loss": 3.8012,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385567194697461e-05,
+      "loss": 3.788,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384730237826782e-05,
+      "loss": 3.8031,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38389164307573e-05,
+      "loss": 3.7907,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383053048324678e-05,
+      "loss": 3.8082,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382214453573626e-05,
+      "loss": 3.786,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381377496702947e-05,
+      "loss": 3.7955,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380538901951895e-05,
+      "loss": 3.7912,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379700307200843e-05,
+      "loss": 3.7843,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378861712449791e-05,
+      "loss": 3.7981,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378023117698739e-05,
+      "loss": 3.7974,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771861608280604e-05,
+      "loss": 3.7874,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763475660770084e-05,
+      "loss": 3.7891,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3755089713259564e-05,
+      "loss": 3.7918,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8872928619384766,
+      "eval_runtime": 308.5774,
+      "eval_samples_per_second": 1236.614,
+      "eval_steps_per_second": 38.645,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3746703765749044e-05,
+      "loss": 3.7814,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3738317818238524e-05,
+      "loss": 3.7791,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3729931870728004e-05,
+      "loss": 3.7898,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3721545923217484e-05,
+      "loss": 3.7882,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371317635451069e-05,
+      "loss": 3.7981,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370479040700017e-05,
+      "loss": 3.7829,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369640445948965e-05,
+      "loss": 3.7835,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368801851197913e-05,
+      "loss": 3.783,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367964894327234e-05,
+      "loss": 3.7977,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367126299576182e-05,
+      "loss": 3.7811,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36628770482513e-05,
+      "loss": 3.7785,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365449110074079e-05,
+      "loss": 3.7823,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3646121532034e-05,
+      "loss": 3.7792,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363773558452348e-05,
+      "loss": 3.7773,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362934963701296e-05,
+      "loss": 3.783,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362096368950243e-05,
+      "loss": 3.7693,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3612594120795647e-05,
+      "loss": 3.7703,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3604208173285126e-05,
+      "loss": 3.7767,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3595822225774606e-05,
+      "loss": 3.7678,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358743627826408e-05,
+      "loss": 3.7933,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357905033075356e-05,
+      "loss": 3.7855,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357066438324304e-05,
+      "loss": 3.7775,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3562294814536255e-05,
+      "loss": 3.7759,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3553908867025735e-05,
+      "loss": 3.7864,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3545522919515215e-05,
+      "loss": 3.771,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3537136972004695e-05,
+      "loss": 3.7819,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3528767403297904e-05,
+      "loss": 3.7708,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3520381455787384e-05,
+      "loss": 3.7782,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3511995508276864e-05,
+      "loss": 3.7697,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3503609560766344e-05,
+      "loss": 3.7725,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3495223613255824e-05,
+      "loss": 3.7783,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348685404454903e-05,
+      "loss": 3.7782,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347846809703851e-05,
+      "loss": 3.7848,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347008214952799e-05,
+      "loss": 3.7807,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346169620201748e-05,
+      "loss": 3.7662,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345332663331069e-05,
+      "loss": 3.7797,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344494068580017e-05,
+      "loss": 3.7709,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343655473828965e-05,
+      "loss": 3.7657,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342816879077913e-05,
+      "loss": 3.7687,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341979922207234e-05,
+      "loss": 3.767,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341141327456182e-05,
+      "loss": 3.7689,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34030273270513e-05,
+      "loss": 3.7829,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339464137954078e-05,
+      "loss": 3.7699,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338627181083399e-05,
+      "loss": 3.7692,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337788586332347e-05,
+      "loss": 3.7765,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336949991581295e-05,
+      "loss": 3.771,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3361113968302434e-05,
+      "loss": 3.7707,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335274439959564e-05,
+      "loss": 3.761,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334435845208512e-05,
+      "loss": 3.7501,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33359725045746e-05,
+      "loss": 3.778,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332758655706408e-05,
+      "loss": 3.7688,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331921698835729e-05,
+      "loss": 3.7626,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331083104084677e-05,
+      "loss": 3.7626,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330244509333625e-05,
+      "loss": 3.7692,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329405914582573e-05,
+      "loss": 3.7518,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328567319831521e-05,
+      "loss": 3.7683,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327730362960842e-05,
+      "loss": 3.7661,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32689176820979e-05,
+      "loss": 3.7686,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326053173458739e-05,
+      "loss": 3.7638,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325214578707687e-05,
+      "loss": 3.7637,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3243776218370076e-05,
+      "loss": 3.7546,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3235390270859556e-05,
+      "loss": 3.7674,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3227004323349036e-05,
+      "loss": 3.7536,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3218618375838516e-05,
+      "loss": 3.7591,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3210248807131725e-05,
+      "loss": 3.7706,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3201862859621205e-05,
+      "loss": 3.7601,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3193476912110685e-05,
+      "loss": 3.757,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3185090964600165e-05,
+      "loss": 3.7546,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3176705017089645e-05,
+      "loss": 3.755,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3168335448382854e-05,
+      "loss": 3.7667,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315994950087234e-05,
+      "loss": 3.771,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315156355336182e-05,
+      "loss": 3.7637,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31431776058513e-05,
+      "loss": 3.7736,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313480803714451e-05,
+      "loss": 3.7699,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312642208963399e-05,
+      "loss": 3.7657,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311803614212347e-05,
+      "loss": 3.7598,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310965019461295e-05,
+      "loss": 3.7643,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310128062590616e-05,
+      "loss": 3.759,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309289467839564e-05,
+      "loss": 3.7547,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308450873088512e-05,
+      "loss": 3.7646,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30761227833746e-05,
+      "loss": 3.7579,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.306775321466781e-05,
+      "loss": 3.7705,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305936726715729e-05,
+      "loss": 3.7716,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3050981319646775e-05,
+      "loss": 3.7436,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042595372136254e-05,
+      "loss": 3.7628,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3034225803429464e-05,
+      "loss": 3.7585,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025839855918944e-05,
+      "loss": 3.7626,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3017453908408424e-05,
+      "loss": 3.7574,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3009067960897903e-05,
+      "loss": 3.7505,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3000682013387383e-05,
+      "loss": 3.7599,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299231244468059e-05,
+      "loss": 3.7545,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298392649717007e-05,
+      "loss": 3.7589,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297554054965955e-05,
+      "loss": 3.7543,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296715460214903e-05,
+      "loss": 3.7625,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295878503344224e-05,
+      "loss": 3.7614,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295039908593173e-05,
+      "loss": 3.7606,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294201313842121e-05,
+      "loss": 3.7446,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293362719091069e-05,
+      "loss": 3.7576,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29252576222039e-05,
+      "loss": 3.7585,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291687167469338e-05,
+      "loss": 3.7668,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290848572718286e-05,
+      "loss": 3.7504,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290009977967234e-05,
+      "loss": 3.7549,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2891730210965546e-05,
+      "loss": 3.7491,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2883344263455026e-05,
+      "loss": 3.7622,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2874958315944506e-05,
+      "loss": 3.7468,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286657236843398e-05,
+      "loss": 3.7489,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2858202799727195e-05,
+      "loss": 3.7527,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284981685221668e-05,
+      "loss": 3.7519,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284143090470616e-05,
+      "loss": 3.7437,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2833044957195635e-05,
+      "loss": 3.7514,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282467538848885e-05,
+      "loss": 3.7466,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281628944097833e-05,
+      "loss": 3.7554,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280790349346781e-05,
+      "loss": 3.7554,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2799517545957284e-05,
+      "loss": 3.7439,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27911479772505e-05,
+      "loss": 3.7571,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278276202973998e-05,
+      "loss": 3.7676,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277437608222945e-05,
+      "loss": 3.7644,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276599013471893e-05,
+      "loss": 3.7403,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275760418720842e-05,
+      "loss": 3.7509,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2749234618501636e-05,
+      "loss": 3.7503,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274084867099111e-05,
+      "loss": 3.7444,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273246272348059e-05,
+      "loss": 3.7508,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272407677597007e-05,
+      "loss": 3.7416,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715707207263285e-05,
+      "loss": 3.7547,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270732125975276e-05,
+      "loss": 3.7463,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269893531224224e-05,
+      "loss": 3.7572,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269054936473172e-05,
+      "loss": 3.7458,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268217979602493e-05,
+      "loss": 3.7448,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267379384851441e-05,
+      "loss": 3.7542,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2665407901003887e-05,
+      "loss": 3.7605,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265702195349337e-05,
+      "loss": 3.7503,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264865238478658e-05,
+      "loss": 3.7522,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264026643727606e-05,
+      "loss": 3.7408,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263188048976554e-05,
+      "loss": 3.7567,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262349454225502e-05,
+      "loss": 3.7483,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261512497354823e-05,
+      "loss": 3.7571,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260673902603771e-05,
+      "loss": 3.7449,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.259835307852719e-05,
+      "loss": 3.7567,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258996713101667e-05,
+      "loss": 3.7463,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258159756230988e-05,
+      "loss": 3.7685,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257321161479936e-05,
+      "loss": 3.7414,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256482566728884e-05,
+      "loss": 3.7558,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255643971977833e-05,
+      "loss": 3.7456,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2548070151071536e-05,
+      "loss": 3.7447,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2539684203561016e-05,
+      "loss": 3.7516,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2531298256050496e-05,
+      "loss": 3.7503,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2522912308539976e-05,
+      "loss": 3.7518,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2514542739833185e-05,
+      "loss": 3.7417,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2506156792322665e-05,
+      "loss": 3.7502,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8711609840393066,
+      "eval_runtime": 304.5353,
+      "eval_samples_per_second": 1253.027,
+      "eval_steps_per_second": 39.158,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2497770844812145e-05,
+      "loss": 3.7413,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2489384897301625e-05,
+      "loss": 3.7345,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2481015328594834e-05,
+      "loss": 3.7429,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2472629381084314e-05,
+      "loss": 3.748,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2464243433573794e-05,
+      "loss": 3.7528,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245585748606328e-05,
+      "loss": 3.7452,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244747153855276e-05,
+      "loss": 3.7336,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243910196984597e-05,
+      "loss": 3.7429,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243071602233545e-05,
+      "loss": 3.7525,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.242233007482493e-05,
+      "loss": 3.7418,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.241394412731441e-05,
+      "loss": 3.7376,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.240557455860762e-05,
+      "loss": 3.7391,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23971886110971e-05,
+      "loss": 3.7398,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238880266358658e-05,
+      "loss": 3.7334,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238041671607606e-05,
+      "loss": 3.7423,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237204714736927e-05,
+      "loss": 3.7266,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.236366119985875e-05,
+      "loss": 3.7289,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2355275252348234e-05,
+      "loss": 3.734,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2346889304837714e-05,
+      "loss": 3.7263,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233851973613092e-05,
+      "loss": 3.7498,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23301337886204e-05,
+      "loss": 3.7452,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232174784110988e-05,
+      "loss": 3.735,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.231336189359936e-05,
+      "loss": 3.7325,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.230499232489257e-05,
+      "loss": 3.7413,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.229660637738205e-05,
+      "loss": 3.7308,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.228822042987153e-05,
+      "loss": 3.7397,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227983448236101e-05,
+      "loss": 3.7291,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227146491365422e-05,
+      "loss": 3.7374,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.22630789661437e-05,
+      "loss": 3.7309,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.225469301863319e-05,
+      "loss": 3.7302,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.224630707112267e-05,
+      "loss": 3.737,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.223793750241588e-05,
+      "loss": 3.7399,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222955155490536e-05,
+      "loss": 3.7387,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222116560739484e-05,
+      "loss": 3.7423,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.221277965988432e-05,
+      "loss": 3.7263,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2204410091177526e-05,
+      "loss": 3.7435,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2196024143667006e-05,
+      "loss": 3.7296,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2187638196156486e-05,
+      "loss": 3.724,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2179252248645966e-05,
+      "loss": 3.7281,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2170882679939175e-05,
+      "loss": 3.7265,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2162496732428655e-05,
+      "loss": 3.7281,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2154110784918135e-05,
+      "loss": 3.7384,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.214572483740762e-05,
+      "loss": 3.73,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.213735526870083e-05,
+      "loss": 3.7323,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212896932119031e-05,
+      "loss": 3.7342,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212058337367979e-05,
+      "loss": 3.7306,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.211219742616927e-05,
+      "loss": 3.7318,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.210382785746248e-05,
+      "loss": 3.7206,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.209544190995196e-05,
+      "loss": 3.7103,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.208705596244144e-05,
+      "loss": 3.7372,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.207867001493092e-05,
+      "loss": 3.7337,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.207030044622413e-05,
+      "loss": 3.7205,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206191449871361e-05,
+      "loss": 3.7303,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.205352855120309e-05,
+      "loss": 3.727,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2045142603692575e-05,
+      "loss": 3.7133,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2036773034985784e-05,
+      "loss": 3.7248,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2028387087475264e-05,
+      "loss": 3.7291,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2020001139964744e-05,
+      "loss": 3.7274,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2011615192454224e-05,
+      "loss": 3.723,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.200324562374743e-05,
+      "loss": 3.7274,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199485967623691e-05,
+      "loss": 3.7172,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.198647372872639e-05,
+      "loss": 3.7314,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197808778121587e-05,
+      "loss": 3.7139,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196971821250908e-05,
+      "loss": 3.7184,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196133226499856e-05,
+      "loss": 3.7313,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195294631748804e-05,
+      "loss": 3.7207,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.194456036997753e-05,
+      "loss": 3.7203,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.193619080127074e-05,
+      "loss": 3.7149,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.192780485376022e-05,
+      "loss": 3.7142,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19194189062497e-05,
+      "loss": 3.723,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191103295873918e-05,
+      "loss": 3.7356,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.190266339003239e-05,
+      "loss": 3.724,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.189427744252187e-05,
+      "loss": 3.7368,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.188589149501135e-05,
+      "loss": 3.729,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.187750554750083e-05,
+      "loss": 3.7308,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1869135978794036e-05,
+      "loss": 3.7241,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1860750031283516e-05,
+      "loss": 3.7223,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1852364083772996e-05,
+      "loss": 3.7214,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.184397813626248e-05,
+      "loss": 3.712,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.183560856755569e-05,
+      "loss": 3.7268,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.182722262004517e-05,
+      "loss": 3.7215,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181883667253465e-05,
+      "loss": 3.7321,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181045072502413e-05,
+      "loss": 3.7322,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180208115631734e-05,
+      "loss": 3.7051,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.179369520880682e-05,
+      "loss": 3.7269,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.17853092612963e-05,
+      "loss": 3.7186,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1776923313785774e-05,
+      "loss": 3.7252,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.176855374507899e-05,
+      "loss": 3.7164,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.176016779756847e-05,
+      "loss": 3.7129,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.175178185005795e-05,
+      "loss": 3.7226,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.174339590254743e-05,
+      "loss": 3.718,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1735026333840645e-05,
+      "loss": 3.7198,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1726640386330125e-05,
+      "loss": 3.7172,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1718254438819605e-05,
+      "loss": 3.7231,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170986849130908e-05,
+      "loss": 3.7284,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1701498922602294e-05,
+      "loss": 3.714,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1693112975091774e-05,
+      "loss": 3.7122,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.168472702758125e-05,
+      "loss": 3.7174,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.167634108007073e-05,
+      "loss": 3.7194,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.166797151136394e-05,
+      "loss": 3.731,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165958556385342e-05,
+      "loss": 3.7149,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.16511996163429e-05,
+      "loss": 3.7168,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.164281366883238e-05,
+      "loss": 3.7137,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.16344441001256e-05,
+      "loss": 3.7268,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.162605815261508e-05,
+      "loss": 3.7051,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.161767220510455e-05,
+      "loss": 3.7137,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160928625759403e-05,
+      "loss": 3.7149,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160091668888725e-05,
+      "loss": 3.716,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.159253074137672e-05,
+      "loss": 3.706,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.15841447938662e-05,
+      "loss": 3.714,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.157575884635568e-05,
+      "loss": 3.7084,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.15673892776489e-05,
+      "loss": 3.7188,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155900333013838e-05,
+      "loss": 3.7175,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155061738262786e-05,
+      "loss": 3.7101,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154223143511734e-05,
+      "loss": 3.7156,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.153386186641055e-05,
+      "loss": 3.7298,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1525475918900026e-05,
+      "loss": 3.7292,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1517089971389506e-05,
+      "loss": 3.7063,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1508704023878986e-05,
+      "loss": 3.7113,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1500334455172195e-05,
+      "loss": 3.7139,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1491948507661675e-05,
+      "loss": 3.7063,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1483562560151155e-05,
+      "loss": 3.7182,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1475176612640635e-05,
+      "loss": 3.7,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.146680704393385e-05,
+      "loss": 3.7229,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.145842109642333e-05,
+      "loss": 3.705,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.145003514891281e-05,
+      "loss": 3.7231,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144164920140229e-05,
+      "loss": 3.7097,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14332796326955e-05,
+      "loss": 3.7114,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.142489368518498e-05,
+      "loss": 3.7153,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.141650773767446e-05,
+      "loss": 3.7239,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.140812179016394e-05,
+      "loss": 3.7117,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139975222145715e-05,
+      "loss": 3.7175,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139136627394663e-05,
+      "loss": 3.7046,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.138298032643611e-05,
+      "loss": 3.722,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.137459437892559e-05,
+      "loss": 3.7117,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1366224810218804e-05,
+      "loss": 3.718,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1357838862708284e-05,
+      "loss": 3.7131,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1349452915197764e-05,
+      "loss": 3.7206,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1341066967687244e-05,
+      "loss": 3.7125,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.133269739898045e-05,
+      "loss": 3.7311,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.132431145146993e-05,
+      "loss": 3.7058,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.131592550395941e-05,
+      "loss": 3.7186,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.130753955644889e-05,
+      "loss": 3.7123,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.12991699877421e-05,
+      "loss": 3.7118,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129078404023158e-05,
+      "loss": 3.71,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128239809272106e-05,
+      "loss": 3.7157,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.127401214521054e-05,
+      "loss": 3.7149,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.126564257650376e-05,
+      "loss": 3.7066,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.125725662899324e-05,
+      "loss": 3.7204,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8612208366394043,
+      "eval_runtime": 303.3015,
+      "eval_samples_per_second": 1258.124,
+      "eval_steps_per_second": 39.317,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124887068148272e-05,
+      "loss": 3.708,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.12404847339722e-05,
+      "loss": 3.6998,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123209878646168e-05,
+      "loss": 3.707,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.122371283895116e-05,
+      "loss": 3.7104,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.121534327024437e-05,
+      "loss": 3.7178,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.120695732273385e-05,
+      "loss": 3.7115,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119857137522333e-05,
+      "loss": 3.705,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119018542771281e-05,
+      "loss": 3.706,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1181815859006016e-05,
+      "loss": 3.7182,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1173429911495496e-05,
+      "loss": 3.7069,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1165043963984976e-05,
+      "loss": 3.7011,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1156658016474456e-05,
+      "loss": 3.7055,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114828844776767e-05,
+      "loss": 3.7043,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113990250025715e-05,
+      "loss": 3.699,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113151655274663e-05,
+      "loss": 3.7061,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.112313060523611e-05,
+      "loss": 3.6911,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111476103652932e-05,
+      "loss": 3.6954,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11063750890188e-05,
+      "loss": 3.6997,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109798914150828e-05,
+      "loss": 3.6941,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108960319399776e-05,
+      "loss": 3.7137,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108123362529097e-05,
+      "loss": 3.7117,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.107284767778045e-05,
+      "loss": 3.7013,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106446173026993e-05,
+      "loss": 3.6996,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.105607578275941e-05,
+      "loss": 3.7069,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1047706214052625e-05,
+      "loss": 3.6976,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1039320266542105e-05,
+      "loss": 3.7047,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1030934319031585e-05,
+      "loss": 3.6924,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1022548371521065e-05,
+      "loss": 3.7059,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1014178802814274e-05,
+      "loss": 3.7035,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1005792855303754e-05,
+      "loss": 3.694,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0997406907793234e-05,
+      "loss": 3.7001,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0989020960282714e-05,
+      "loss": 3.7063,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098065139157592e-05,
+      "loss": 3.7022,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.09722654440654e-05,
+      "loss": 3.7122,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.096387949655488e-05,
+      "loss": 3.6925,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.095549354904436e-05,
+      "loss": 3.7034,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.094712398033758e-05,
+      "loss": 3.6985,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093873803282706e-05,
+      "loss": 3.6921,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093035208531654e-05,
+      "loss": 3.6942,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092196613780602e-05,
+      "loss": 3.6918,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.091359656909923e-05,
+      "loss": 3.6936,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.090521062158871e-05,
+      "loss": 3.7064,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.089682467407819e-05,
+      "loss": 3.6973,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.088843872656767e-05,
+      "loss": 3.6987,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.088006915786088e-05,
+      "loss": 3.7027,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087168321035036e-05,
+      "loss": 3.6955,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.086329726283984e-05,
+      "loss": 3.6946,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085491131532932e-05,
+      "loss": 3.6922,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084654174662253e-05,
+      "loss": 3.6699,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083815579911201e-05,
+      "loss": 3.7106,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082976985160149e-05,
+      "loss": 3.6944,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082138390409097e-05,
+      "loss": 3.6927,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081301433538418e-05,
+      "loss": 3.6944,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080462838787366e-05,
+      "loss": 3.6928,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079624244036314e-05,
+      "loss": 3.6808,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078785649285262e-05,
+      "loss": 3.6879,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.07794705453421e-05,
+      "loss": 3.6978,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077110097663531e-05,
+      "loss": 3.6934,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076271502912479e-05,
+      "loss": 3.6979,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075432908161427e-05,
+      "loss": 3.6863,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0745959512907486e-05,
+      "loss": 3.6846,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0737573565396966e-05,
+      "loss": 3.7025,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0729187617886446e-05,
+      "loss": 3.6785,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0720801670375926e-05,
+      "loss": 3.6862,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.07124157228654e-05,
+      "loss": 3.696,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0704046154158615e-05,
+      "loss": 3.691,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0695660206648095e-05,
+      "loss": 3.6851,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0687274259137575e-05,
+      "loss": 3.6857,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067888831162705e-05,
+      "loss": 3.6789,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0670518742920264e-05,
+      "loss": 3.6886,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0662132795409744e-05,
+      "loss": 3.702,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0653746847899224e-05,
+      "loss": 3.6911,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0645360900388704e-05,
+      "loss": 3.7032,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063699133168192e-05,
+      "loss": 3.6965,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.06286053841714e-05,
+      "loss": 3.7022,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062021943666087e-05,
+      "loss": 3.6879,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061183348915035e-05,
+      "loss": 3.6926,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.060346392044357e-05,
+      "loss": 3.6901,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.059507797293305e-05,
+      "loss": 3.6767,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.058669202542252e-05,
+      "loss": 3.6944,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0578306077912e-05,
+      "loss": 3.6898,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056993650920522e-05,
+      "loss": 3.6975,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05615505616947e-05,
+      "loss": 3.7023,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055316461418418e-05,
+      "loss": 3.6783,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054477866667366e-05,
+      "loss": 3.69,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053639271916314e-05,
+      "loss": 3.6867,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.052802315045635e-05,
+      "loss": 3.6939,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051963720294583e-05,
+      "loss": 3.6862,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0511251255435307e-05,
+      "loss": 3.6832,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0502865307924787e-05,
+      "loss": 3.688,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0494495739217996e-05,
+      "loss": 3.6841,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0486109791707476e-05,
+      "loss": 3.6883,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0477723844196956e-05,
+      "loss": 3.6849,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0469337896686436e-05,
+      "loss": 3.6957,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046096832797965e-05,
+      "loss": 3.6958,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.045258238046913e-05,
+      "loss": 3.6826,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044419643295861e-05,
+      "loss": 3.6801,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.043581048544809e-05,
+      "loss": 3.6851,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04274409167413e-05,
+      "loss": 3.6926,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041905496923078e-05,
+      "loss": 3.6935,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041066902172026e-05,
+      "loss": 3.6866,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040228307420974e-05,
+      "loss": 3.6852,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039391350550295e-05,
+      "loss": 3.6809,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038552755799243e-05,
+      "loss": 3.698,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037714161048191e-05,
+      "loss": 3.6716,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036875566297139e-05,
+      "loss": 3.683,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0360386094264605e-05,
+      "loss": 3.6785,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0352000146754085e-05,
+      "loss": 3.6864,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0343614199243565e-05,
+      "loss": 3.6745,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0335228251733045e-05,
+      "loss": 3.6884,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0326858683026254e-05,
+      "loss": 3.6749,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0318472735515734e-05,
+      "loss": 3.6881,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0310086788005214e-05,
+      "loss": 3.6881,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0301700840494694e-05,
+      "loss": 3.6768,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.02933312717879e-05,
+      "loss": 3.6859,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028494532427738e-05,
+      "loss": 3.7015,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027655937676686e-05,
+      "loss": 3.6873,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026817342925634e-05,
+      "loss": 3.6834,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025980386054955e-05,
+      "loss": 3.675,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025141791303904e-05,
+      "loss": 3.6845,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.024303196552852e-05,
+      "loss": 3.6833,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0234646018018e-05,
+      "loss": 3.6817,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.022626007050748e-05,
+      "loss": 3.6683,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.021789050180069e-05,
+      "loss": 3.6891,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020950455429017e-05,
+      "loss": 3.6778,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020111860677965e-05,
+      "loss": 3.6897,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.019273265926913e-05,
+      "loss": 3.6798,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.018436309056234e-05,
+      "loss": 3.6836,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0175977143051817e-05,
+      "loss": 3.6813,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0167591195541297e-05,
+      "loss": 3.6946,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0159205248030776e-05,
+      "loss": 3.6823,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015083567932399e-05,
+      "loss": 3.688,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014244973181347e-05,
+      "loss": 3.67,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013406378430295e-05,
+      "loss": 3.6926,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.012567783679243e-05,
+      "loss": 3.681,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011730826808564e-05,
+      "loss": 3.6884,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010892232057512e-05,
+      "loss": 3.6834,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01005363730646e-05,
+      "loss": 3.6872,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009215042555408e-05,
+      "loss": 3.678,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008378085684729e-05,
+      "loss": 3.7025,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007539490933677e-05,
+      "loss": 3.6776,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006700896182625e-05,
+      "loss": 3.6902,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005862301431573e-05,
+      "loss": 3.6846,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0050253445608946e-05,
+      "loss": 3.6805,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0041867498098426e-05,
+      "loss": 3.6773,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0033481550587906e-05,
+      "loss": 3.6874,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0025095603077386e-05,
+      "loss": 3.6856,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0016726034370595e-05,
+      "loss": 3.6767,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0008340086860075e-05,
+      "loss": 3.6869,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.854868173599243,
+      "eval_runtime": 310.6755,
+      "eval_samples_per_second": 1228.262,
+      "eval_steps_per_second": 38.384,
+      "step": 610560
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 4.21995487925846e+17,
+  "trial_name": null,
+  "trial_params": null
+}