diff --git "a/checkpoint-1908000/trainer_state.json" "b/checkpoint-1908000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1908000/trainer_state.json"
@@ -0,0 +1,22581 @@
+{
+  "best_metric": 3.8433103561401367,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/existential-there-quantifier/transformer/2/checkpoints/checkpoint-915840",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 1908000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 11.016,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8135,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1935,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.989,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8212,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7126,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6089,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5426,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4525,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.408,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.357,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3155,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2618,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2164,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1741,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1411,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1112,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0959,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0536,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0335,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0219,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9772,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9516,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9374,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9288,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.8946,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782030879941405e-05,
+      "loss": 4.8738,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773644932430885e-05,
+      "loss": 4.8655,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765258984920364e-05,
+      "loss": 4.8441,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756873037409844e-05,
+      "loss": 4.8186,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.8018,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7983,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7853,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 4.7813,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 4.7579,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7416,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7452,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.7076,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7035,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6897,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6887,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6667,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6773,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6477,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6527,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6424,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6343,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6304,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.6062,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.5929,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958085003370758e-05,
+      "loss": 4.6059,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957246408619706e-05,
+      "loss": 4.5919,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 4.574,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5635,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5697,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5521,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5653,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5385,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5381,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5403,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5218,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948863736869932e-05,
+      "loss": 4.505,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5098,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4938,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4941,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5034,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 4.4972,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 4.4825,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 4.471,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.4648,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4748,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4702,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4627,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4621,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4649,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4566,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.4309,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.4426,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4366,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.4238,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.435,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.424,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4328,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 4.421,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.3936,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.405,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.4037,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 4.4055,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.3969,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925392911122715e-05,
+      "loss": 4.39,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 4.3905,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 4.3802,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 4.3719,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922038532118507e-05,
+      "loss": 4.3742,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921199937367455e-05,
+      "loss": 4.371,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.3809,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.3577,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.3551,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.3533,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.3604,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.3592,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.344,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3449,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.3266,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.3403,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911980308747002e-05,
+      "loss": 4.3264,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.3239,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.3297,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3276,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.3201,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.3192,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.31,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3188,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3059,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.3059,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.3119,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.3078,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.3053,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.297,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.3023,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994063011223415e-05,
+      "loss": 4.2978,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985677063712895e-05,
+      "loss": 4.2825,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977291116202375e-05,
+      "loss": 4.3009,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968905168691855e-05,
+      "loss": 4.2869,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.2934,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.275,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2702,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2688,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.2702,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2636,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2789,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2755,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2598,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.2641,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887670888248733e-05,
+      "loss": 4.2669,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886832293497681e-05,
+      "loss": 4.2629,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.2667,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.2648,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2509,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.2567,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.252,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2433,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8809654060010626e-05,
+      "loss": 4.2538,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8801268112500106e-05,
+      "loss": 4.2426,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8792882164989586e-05,
+      "loss": 4.2303,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2548,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877612664877228e-05,
+      "loss": 4.2328,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876774070126176e-05,
+      "loss": 4.2315,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875935475375124e-05,
+      "loss": 4.2346,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875096880624072e-05,
+      "loss": 4.2284,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.198453903198242,
+      "eval_runtime": 303.027,
+      "eval_samples_per_second": 1259.264,
+      "eval_steps_per_second": 39.353,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874259923753393e-05,
+      "loss": 4.2117,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873421329002341e-05,
+      "loss": 4.2218,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872582734251289e-05,
+      "loss": 4.2385,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871744139500237e-05,
+      "loss": 4.2172,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870907182629558e-05,
+      "loss": 4.2315,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870068587878506e-05,
+      "loss": 4.2041,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869229993127454e-05,
+      "loss": 4.2185,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683913983764027e-05,
+      "loss": 4.2012,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675560793860965e-05,
+      "loss": 4.2125,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667174846350445e-05,
+      "loss": 4.2101,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658788898839925e-05,
+      "loss": 4.2129,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650402951329405e-05,
+      "loss": 4.2092,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642017003818885e-05,
+      "loss": 4.1947,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633631056308365e-05,
+      "loss": 4.1867,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8625245108797845e-05,
+      "loss": 4.1886,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8616859161287324e-05,
+      "loss": 4.1906,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608489592580534e-05,
+      "loss": 4.2008,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8600103645070014e-05,
+      "loss": 4.1851,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591717697559493e-05,
+      "loss": 4.1933,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858333175004898e-05,
+      "loss": 4.2079,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857496218134219e-05,
+      "loss": 4.1831,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856657623383167e-05,
+      "loss": 4.1781,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855819028632115e-05,
+      "loss": 4.1814,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854980433881063e-05,
+      "loss": 4.1889,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854143477010384e-05,
+      "loss": 4.1717,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853304882259332e-05,
+      "loss": 4.175,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85246628750828e-05,
+      "loss": 4.1748,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851627692757228e-05,
+      "loss": 4.1707,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850790735886549e-05,
+      "loss": 4.1557,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849952141135497e-05,
+      "loss": 4.161,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849113546384445e-05,
+      "loss": 4.1598,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482749516333934e-05,
+      "loss": 4.17,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847437994762714e-05,
+      "loss": 4.1745,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846599400011662e-05,
+      "loss": 4.1584,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84576080526061e-05,
+      "loss": 4.154,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844922210509558e-05,
+      "loss": 4.1671,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844085253638879e-05,
+      "loss": 4.1422,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843246658887827e-05,
+      "loss": 4.1501,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842408064136775e-05,
+      "loss": 4.1418,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415694693857225e-05,
+      "loss": 4.148,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840732512515044e-05,
+      "loss": 4.1397,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839893917763992e-05,
+      "loss": 4.1563,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83905532301294e-05,
+      "loss": 4.143,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838216728261888e-05,
+      "loss": 4.1412,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83737977139121e-05,
+      "loss": 4.1408,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836541176640158e-05,
+      "loss": 4.1442,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8357025818891057e-05,
+      "loss": 4.1371,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834863987138053e-05,
+      "loss": 4.1277,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8340270302673746e-05,
+      "loss": 4.1198,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331884355163226e-05,
+      "loss": 4.1429,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83234984076527e-05,
+      "loss": 4.1316,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831511246014218e-05,
+      "loss": 4.1192,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306742891435395e-05,
+      "loss": 4.1205,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298356943924875e-05,
+      "loss": 4.1282,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289970996414354e-05,
+      "loss": 4.1148,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281585048903834e-05,
+      "loss": 4.1332,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827321548019705e-05,
+      "loss": 4.1158,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826482953268653e-05,
+      "loss": 4.1182,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256443585176003e-05,
+      "loss": 4.1291,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248057637665483e-05,
+      "loss": 4.1052,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82396880689587e-05,
+      "loss": 4.1042,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823130212144817e-05,
+      "loss": 4.1127,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822291617393765e-05,
+      "loss": 4.0963,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821453022642713e-05,
+      "loss": 4.1005,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820616065772035e-05,
+      "loss": 4.1137,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819777471020983e-05,
+      "loss": 4.1182,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818938876269931e-05,
+      "loss": 4.1043,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818100281518879e-05,
+      "loss": 4.0986,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817264962528573e-05,
+      "loss": 4.0891,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816426367777521e-05,
+      "loss": 4.1087,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815587773026469e-05,
+      "loss": 4.1055,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814749178275417e-05,
+      "loss": 4.1048,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8139105835243646e-05,
+      "loss": 4.1056,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8130719887733126e-05,
+      "loss": 4.1076,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812235031902634e-05,
+      "loss": 4.1069,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811396437151582e-05,
+      "loss": 4.0849,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81055784240053e-05,
+      "loss": 4.0975,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809719247649478e-05,
+      "loss": 4.0992,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808880652898426e-05,
+      "loss": 4.0835,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808042058147374e-05,
+      "loss": 4.1053,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807205101276695e-05,
+      "loss": 4.0973,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806366506525643e-05,
+      "loss": 4.1022,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805527911774591e-05,
+      "loss": 4.0972,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804689317023539e-05,
+      "loss": 4.0753,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803850722272487e-05,
+      "loss": 4.0879,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803012127521435e-05,
+      "loss": 4.087,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802173532770383e-05,
+      "loss": 4.091,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801334938019331e-05,
+      "loss": 4.0865,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800497981148652e-05,
+      "loss": 4.0781,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996593863976e-05,
+      "loss": 4.0868,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988207916465486e-05,
+      "loss": 4.0824,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979821968954966e-05,
+      "loss": 4.0668,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971436021444446e-05,
+      "loss": 4.0761,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7963050073933926e-05,
+      "loss": 4.0812,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954680505227135e-05,
+      "loss": 4.0876,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946294557716615e-05,
+      "loss": 4.0705,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937908610206095e-05,
+      "loss": 4.0605,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929522662695575e-05,
+      "loss": 4.0688,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7921153093988784e-05,
+      "loss": 4.0757,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912767146478264e-05,
+      "loss": 4.0754,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7904381198967744e-05,
+      "loss": 4.0679,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895995251457224e-05,
+      "loss": 4.0637,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887609303946704e-05,
+      "loss": 4.0532,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879223356436184e-05,
+      "loss": 4.067,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78708537877294e-05,
+      "loss": 4.0539,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786246784021888e-05,
+      "loss": 4.0584,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785408189270836e-05,
+      "loss": 4.063,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784569594519783e-05,
+      "loss": 4.0623,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783730999768731e-05,
+      "loss": 4.0563,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782892405017679e-05,
+      "loss": 4.0602,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782053810266627e-05,
+      "loss": 4.0509,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781216853395948e-05,
+      "loss": 4.057,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780378258644896e-05,
+      "loss": 4.0504,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779539663893844e-05,
+      "loss": 4.0485,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778701069142792e-05,
+      "loss": 4.0614,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77786247439174e-05,
+      "loss": 4.0609,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777025517521062e-05,
+      "loss": 4.0558,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77618692277001e-05,
+      "loss": 4.0505,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775348328018958e-05,
+      "loss": 4.0545,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774509733267906e-05,
+      "loss": 4.0531,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773671138516854e-05,
+      "loss": 4.0391,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7728341816461747e-05,
+      "loss": 4.0586,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719955868951227e-05,
+      "loss": 4.0459,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7711569921440707e-05,
+      "loss": 4.0534,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703183973930186e-05,
+      "loss": 4.0375,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694814405223396e-05,
+      "loss": 4.0379,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686428457712876e-05,
+      "loss": 4.0351,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7678042510202355e-05,
+      "loss": 4.0353,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766965656269184e-05,
+      "loss": 4.0345,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766128699398505e-05,
+      "loss": 4.0466,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765290104647453e-05,
+      "loss": 4.0516,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764451509896401e-05,
+      "loss": 4.0379,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763612915145349e-05,
+      "loss": 4.0318,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762774320394297e-05,
+      "loss": 4.0423,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761937363523618e-05,
+      "loss": 4.0417,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761098768772566e-05,
+      "loss": 4.0467,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760260174021514e-05,
+      "loss": 4.0402,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759421579270462e-05,
+      "loss": 4.0369,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75858298451941e-05,
+      "loss": 4.0288,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757744389768358e-05,
+      "loss": 4.0403,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569074328976796e-05,
+      "loss": 4.0298,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560688381466276e-05,
+      "loss": 4.0334,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552302433955756e-05,
+      "loss": 4.0316,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543916486445236e-05,
+      "loss": 4.0212,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535546917738445e-05,
+      "loss": 4.0453,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527160970227925e-05,
+      "loss": 4.0253,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518775022717405e-05,
+      "loss": 4.0204,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510389075206885e-05,
+      "loss": 4.0248,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7502019506500094e-05,
+      "loss": 4.0216,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.029199600219727,
+      "eval_runtime": 302.4487,
+      "eval_samples_per_second": 1261.672,
+      "eval_steps_per_second": 39.428,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493633558989574e-05,
+      "loss": 4.0063,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485247611479054e-05,
+      "loss": 4.0207,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476861663968534e-05,
+      "loss": 4.0335,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468475716458014e-05,
+      "loss": 4.0191,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460089768947494e-05,
+      "loss": 4.0317,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451703821436974e-05,
+      "loss": 4.0104,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443317873926453e-05,
+      "loss": 4.024,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434931926415933e-05,
+      "loss": 4.0021,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742656235770914e-05,
+      "loss": 4.0198,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741817641019862e-05,
+      "loss": 4.0171,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74097904626881e-05,
+      "loss": 4.0152,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740140451517758e-05,
+      "loss": 4.0237,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739301856766706e-05,
+      "loss": 4.0022,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738463262015654e-05,
+      "loss": 3.9996,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737626305144975e-05,
+      "loss": 3.9997,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736787710393923e-05,
+      "loss": 4.0053,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735949115642871e-05,
+      "loss": 4.0129,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73511052089182e-05,
+      "loss": 4.0019,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734273564021141e-05,
+      "loss": 4.0069,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733434969270089e-05,
+      "loss": 4.0264,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732596374519037e-05,
+      "loss": 4.0011,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731757779767985e-05,
+      "loss": 3.9965,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7309208228973056e-05,
+      "loss": 4.0035,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300822281462536e-05,
+      "loss": 4.0099,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292436333952016e-05,
+      "loss": 3.9942,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7284050386441496e-05,
+      "loss": 3.9982,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275664438930976e-05,
+      "loss": 3.9999,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267294870224185e-05,
+      "loss": 3.9967,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258908922713665e-05,
+      "loss": 3.9824,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725052297520315e-05,
+      "loss": 3.9906,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724213702769263e-05,
+      "loss": 3.9926,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723376745898584e-05,
+      "loss": 3.9961,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722538151147532e-05,
+      "loss": 4.0089,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72169955639648e-05,
+      "loss": 3.9873,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720860961645428e-05,
+      "loss": 3.9865,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720022366894376e-05,
+      "loss": 3.9971,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719185410023697e-05,
+      "loss": 3.9808,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718346815272645e-05,
+      "loss": 3.9815,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717508220521593e-05,
+      "loss": 3.9754,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716669625770541e-05,
+      "loss": 3.9873,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715831031019489e-05,
+      "loss": 3.9742,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149940741488105e-05,
+      "loss": 3.9963,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141554793977585e-05,
+      "loss": 3.9778,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133168846467065e-05,
+      "loss": 3.9862,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124782898956545e-05,
+      "loss": 3.985,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116396951446025e-05,
+      "loss": 3.9793,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7108027382739234e-05,
+      "loss": 3.9812,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099641435228714e-05,
+      "loss": 3.9677,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091255487718194e-05,
+      "loss": 3.9628,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082869540207674e-05,
+      "loss": 3.9863,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707449997150088e-05,
+      "loss": 3.9779,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706611402399036e-05,
+      "loss": 3.9669,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705772807647984e-05,
+      "loss": 3.9704,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704934212896932e-05,
+      "loss": 3.9775,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704097256026254e-05,
+      "loss": 3.957,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703258661275202e-05,
+      "loss": 3.9793,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70242006652415e-05,
+      "loss": 3.9672,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701581471773098e-05,
+      "loss": 3.9656,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700744514902419e-05,
+      "loss": 3.9827,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699905920151367e-05,
+      "loss": 3.9603,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699067325400315e-05,
+      "loss": 3.9572,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698228730649263e-05,
+      "loss": 3.9676,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69739013589821e-05,
+      "loss": 3.9519,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696553179027532e-05,
+      "loss": 3.9552,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69571458427648e-05,
+      "loss": 3.969,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694875989525428e-05,
+      "loss": 3.9715,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694037394774376e-05,
+      "loss": 3.9568,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693198800023324e-05,
+      "loss": 3.9577,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692361843152645e-05,
+      "loss": 3.9451,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915232484015926e-05,
+      "loss": 3.9646,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906846536505406e-05,
+      "loss": 3.9609,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898460588994886e-05,
+      "loss": 3.9657,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6890091020288095e-05,
+      "loss": 3.9628,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881705072777575e-05,
+      "loss": 3.9668,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873319125267055e-05,
+      "loss": 3.9724,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864933177756535e-05,
+      "loss": 3.9456,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685656360904975e-05,
+      "loss": 3.9581,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684817766153923e-05,
+      "loss": 3.9628,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683979171402871e-05,
+      "loss": 3.9459,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683140576651819e-05,
+      "loss": 3.9665,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682301981900767e-05,
+      "loss": 3.9611,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681465025030088e-05,
+      "loss": 3.9673,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680626430279036e-05,
+      "loss": 3.9621,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679787835527984e-05,
+      "loss": 3.941,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678949240776932e-05,
+      "loss": 3.9536,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678112283906253e-05,
+      "loss": 3.9507,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677273689155201e-05,
+      "loss": 3.9602,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676435094404149e-05,
+      "loss": 3.95,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675596499653097e-05,
+      "loss": 3.9497,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747595427824184e-05,
+      "loss": 3.9538,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739209480313664e-05,
+      "loss": 3.9515,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730823532803144e-05,
+      "loss": 3.9367,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722437585292624e-05,
+      "loss": 3.9433,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6714051637782104e-05,
+      "loss": 3.9533,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670568206907531e-05,
+      "loss": 3.9585,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669729612156479e-05,
+      "loss": 3.9472,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668891017405427e-05,
+      "loss": 3.9302,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668052422654375e-05,
+      "loss": 3.9418,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667215465783696e-05,
+      "loss": 3.9488,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666376871032644e-05,
+      "loss": 3.9468,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665538276281592e-05,
+      "loss": 3.9437,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66469968153054e-05,
+      "loss": 3.9397,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663862724659862e-05,
+      "loss": 3.927,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66302412990881e-05,
+      "loss": 3.9476,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662185535157758e-05,
+      "loss": 3.9244,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661346940406706e-05,
+      "loss": 3.9383,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660509983536027e-05,
+      "loss": 3.9329,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659671388784975e-05,
+      "loss": 3.9467,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658832794033923e-05,
+      "loss": 3.9278,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579941992828707e-05,
+      "loss": 3.9417,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571572424121916e-05,
+      "loss": 3.9311,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563186476611396e-05,
+      "loss": 3.9308,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554800529100876e-05,
+      "loss": 3.9355,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546414581590356e-05,
+      "loss": 3.9271,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653804501288357e-05,
+      "loss": 3.942,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652965906537305e-05,
+      "loss": 3.9408,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652127311786253e-05,
+      "loss": 3.936,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651288717035201e-05,
+      "loss": 3.9309,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650451760164522e-05,
+      "loss": 3.9346,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64961316541347e-05,
+      "loss": 3.9386,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648774570662418e-05,
+      "loss": 3.9186,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647935975911366e-05,
+      "loss": 3.9408,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647099019040687e-05,
+      "loss": 3.931,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646260424289635e-05,
+      "loss": 3.9372,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645421829538583e-05,
+      "loss": 3.9248,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644583234787531e-05,
+      "loss": 3.927,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643744640036479e-05,
+      "loss": 3.9156,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429076831658005e-05,
+      "loss": 3.923,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420690884147485e-05,
+      "loss": 3.9197,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412304936636965e-05,
+      "loss": 3.9328,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403918989126445e-05,
+      "loss": 3.9383,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395549420419654e-05,
+      "loss": 3.9264,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387163472909134e-05,
+      "loss": 3.9191,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378777525398614e-05,
+      "loss": 3.9265,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370391577888094e-05,
+      "loss": 3.9291,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63620220091813e-05,
+      "loss": 3.934,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635363606167078e-05,
+      "loss": 3.9263,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634525011416026e-05,
+      "loss": 3.9212,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633686416664974e-05,
+      "loss": 3.9261,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632849459794296e-05,
+      "loss": 3.9309,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632010865043244e-05,
+      "loss": 3.9169,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631172270292192e-05,
+      "loss": 3.9246,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63033367554114e-05,
+      "loss": 3.9203,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629496718670461e-05,
+      "loss": 3.9127,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628658123919409e-05,
+      "loss": 3.9349,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627819529168357e-05,
+      "loss": 3.9181,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626980934417305e-05,
+      "loss": 3.913,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626142339666253e-05,
+      "loss": 3.9176,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6253053827955737e-05,
+      "loss": 3.9117,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9546828269958496,
+      "eval_runtime": 303.1541,
+      "eval_samples_per_second": 1258.736,
+      "eval_steps_per_second": 39.336,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244667880445217e-05,
+      "loss": 3.9057,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6236281932934697e-05,
+      "loss": 3.9127,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227895985424176e-05,
+      "loss": 3.9289,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621952641671739e-05,
+      "loss": 3.9132,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621114046920687e-05,
+      "loss": 3.9329,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620275452169635e-05,
+      "loss": 3.9018,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194368574185825e-05,
+      "loss": 3.9203,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185982626675305e-05,
+      "loss": 3.8985,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617761305796852e-05,
+      "loss": 3.9157,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6169227110458e-05,
+      "loss": 3.9158,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160841162947474e-05,
+      "loss": 3.9105,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152455215436954e-05,
+      "loss": 3.9187,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614408564673017e-05,
+      "loss": 3.8988,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613569969921965e-05,
+      "loss": 3.8987,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612731375170913e-05,
+      "loss": 3.8969,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611892780419861e-05,
+      "loss": 3.9014,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611054185668809e-05,
+      "loss": 3.9105,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610215590917757e-05,
+      "loss": 3.9028,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609376996166705e-05,
+      "loss": 3.9073,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608538401415653e-05,
+      "loss": 3.9235,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607701444544974e-05,
+      "loss": 3.9055,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606862849793922e-05,
+      "loss": 3.9023,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60602425504287e-05,
+      "loss": 3.904,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605185660291818e-05,
+      "loss": 3.91,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604348703421139e-05,
+      "loss": 3.893,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6035101086700875e-05,
+      "loss": 3.902,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026715139190355e-05,
+      "loss": 3.9037,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018329191679835e-05,
+      "loss": 3.8954,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009959622973044e-05,
+      "loss": 3.889,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001573675462524e-05,
+      "loss": 3.8893,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5993187727952004e-05,
+      "loss": 3.9009,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5984801780441484e-05,
+      "loss": 3.8986,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597643221173469e-05,
+      "loss": 3.9121,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596804626422417e-05,
+      "loss": 3.8981,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595966031671365e-05,
+      "loss": 3.8846,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595127436920313e-05,
+      "loss": 3.8999,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594290480049634e-05,
+      "loss": 3.8914,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593451885298583e-05,
+      "loss": 3.8843,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592613290547531e-05,
+      "loss": 3.8856,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591774695796479e-05,
+      "loss": 3.8902,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5909377389258e-05,
+      "loss": 3.8835,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590099144174748e-05,
+      "loss": 3.9023,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589260549423696e-05,
+      "loss": 3.8859,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588421954672644e-05,
+      "loss": 3.8918,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5875849978019646e-05,
+      "loss": 3.8945,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867464030509126e-05,
+      "loss": 3.8849,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5859078082998606e-05,
+      "loss": 3.8908,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5850692135488086e-05,
+      "loss": 3.8755,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5842322566781295e-05,
+      "loss": 3.8643,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5833936619270775e-05,
+      "loss": 3.898,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582555067176026e-05,
+      "loss": 3.8861,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581716472424974e-05,
+      "loss": 3.8805,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580879515554295e-05,
+      "loss": 3.877,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580040920803243e-05,
+      "loss": 3.884,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579202326052191e-05,
+      "loss": 3.8688,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578363731301139e-05,
+      "loss": 3.8884,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57752677443046e-05,
+      "loss": 3.881,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576688179679408e-05,
+      "loss": 3.8777,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575849584928356e-05,
+      "loss": 3.8906,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575010990177304e-05,
+      "loss": 3.8707,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574174033306625e-05,
+      "loss": 3.869,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573335438555573e-05,
+      "loss": 3.8837,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724968438045216e-05,
+      "loss": 3.8588,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716582490534696e-05,
+      "loss": 3.8641,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5708212921827905e-05,
+      "loss": 3.8838,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699826974317385e-05,
+      "loss": 3.884,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691441026806865e-05,
+      "loss": 3.8662,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5683055079296345e-05,
+      "loss": 3.8773,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674685510589554e-05,
+      "loss": 3.857,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5666299563079034e-05,
+      "loss": 3.879,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657913615568514e-05,
+      "loss": 3.8749,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649527668057994e-05,
+      "loss": 3.8801,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56411580993512e-05,
+      "loss": 3.875,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563277215184068e-05,
+      "loss": 3.8803,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562438620433017e-05,
+      "loss": 3.8885,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561600025681965e-05,
+      "loss": 3.8592,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560763068811286e-05,
+      "loss": 3.8761,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559924474060234e-05,
+      "loss": 3.8797,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559085879309182e-05,
+      "loss": 3.8575,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55824728455813e-05,
+      "loss": 3.887,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557410327687451e-05,
+      "loss": 3.877,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556571732936399e-05,
+      "loss": 3.8788,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555733138185347e-05,
+      "loss": 3.8822,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554894543434295e-05,
+      "loss": 3.8601,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554055948683243e-05,
+      "loss": 3.8638,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5532189918125636e-05,
+      "loss": 3.8699,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552380397061512e-05,
+      "loss": 3.8769,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55154180231046e-05,
+      "loss": 3.8662,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550703207559408e-05,
+      "loss": 3.8679,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549866250688729e-05,
+      "loss": 3.8668,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549027655937677e-05,
+      "loss": 3.8768,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548189061186625e-05,
+      "loss": 3.8563,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5473504664355725e-05,
+      "loss": 3.8633,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546513509564894e-05,
+      "loss": 3.8713,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545674914813842e-05,
+      "loss": 3.8786,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54483632006279e-05,
+      "loss": 3.8665,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5439977253117374e-05,
+      "loss": 3.8509,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543160768441059e-05,
+      "loss": 3.8607,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542322173690008e-05,
+      "loss": 3.8662,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541483578938956e-05,
+      "loss": 3.8678,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540644984187903e-05,
+      "loss": 3.8661,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5398080273172246e-05,
+      "loss": 3.8606,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389694325661726e-05,
+      "loss": 3.848,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53813083781512e-05,
+      "loss": 3.8649,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537292243064068e-05,
+      "loss": 3.848,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364552861933895e-05,
+      "loss": 3.8565,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5356166914423375e-05,
+      "loss": 3.8525,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534778096691285e-05,
+      "loss": 3.8675,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339411398206064e-05,
+      "loss": 3.8501,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5331025450695544e-05,
+      "loss": 3.8674,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532263950318503e-05,
+      "loss": 3.8485,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314253555674504e-05,
+      "loss": 3.8565,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305867608163983e-05,
+      "loss": 3.8565,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297481660653463e-05,
+      "loss": 3.8534,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528909571314294e-05,
+      "loss": 3.8593,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528070976563242e-05,
+      "loss": 3.8683,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527234019692563e-05,
+      "loss": 3.8601,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526395424941511e-05,
+      "loss": 3.8533,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525556830190459e-05,
+      "loss": 3.8572,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524718235439407e-05,
+      "loss": 3.861,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523881278568728e-05,
+      "loss": 3.8431,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523042683817677e-05,
+      "loss": 3.8638,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522204089066625e-05,
+      "loss": 3.8506,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521365494315573e-05,
+      "loss": 3.8629,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520528537444894e-05,
+      "loss": 3.8514,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519689942693842e-05,
+      "loss": 3.8536,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51885134794279e-05,
+      "loss": 3.8379,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518012753191738e-05,
+      "loss": 3.851,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5171757963210586e-05,
+      "loss": 3.8447,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5163372015700066e-05,
+      "loss": 3.8544,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154986068189546e-05,
+      "loss": 3.8599,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146600120679026e-05,
+      "loss": 3.8554,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5138230551972235e-05,
+      "loss": 3.8417,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512984460446172e-05,
+      "loss": 3.8544,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51214586569512e-05,
+      "loss": 3.8577,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511307270944068e-05,
+      "loss": 3.8575,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510470314073389e-05,
+      "loss": 3.8542,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509631719322337e-05,
+      "loss": 3.8479,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508793124571285e-05,
+      "loss": 3.8498,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507954529820233e-05,
+      "loss": 3.8595,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507117572949554e-05,
+      "loss": 3.8426,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506278978198502e-05,
+      "loss": 3.854,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50544038344745e-05,
+      "loss": 3.8428,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504601788696398e-05,
+      "loss": 3.8457,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503764831825719e-05,
+      "loss": 3.8611,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5029262370746675e-05,
+      "loss": 3.8473,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020876423236155e-05,
+      "loss": 3.8405,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012490475725635e-05,
+      "loss": 3.84,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5004120907018844e-05,
+      "loss": 3.846,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9137043952941895,
+      "eval_runtime": 304.2318,
+      "eval_samples_per_second": 1254.277,
+      "eval_steps_per_second": 39.197,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995734959508324e-05,
+      "loss": 3.8285,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4987349011997804e-05,
+      "loss": 3.8423,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978963064487284e-05,
+      "loss": 3.8529,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970577116976764e-05,
+      "loss": 3.8442,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4962191169466244e-05,
+      "loss": 3.8582,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953805221955724e-05,
+      "loss": 3.8393,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945419274445204e-05,
+      "loss": 3.8436,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4937033326934684e-05,
+      "loss": 3.8334,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492866375822789e-05,
+      "loss": 3.8398,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492027781071737e-05,
+      "loss": 3.8432,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491189186320685e-05,
+      "loss": 3.8423,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490350591569633e-05,
+      "loss": 3.8499,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489515272579328e-05,
+      "loss": 3.8273,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488676677828276e-05,
+      "loss": 3.8329,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487838083077224e-05,
+      "loss": 3.8302,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486999488326172e-05,
+      "loss": 3.8316,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48616089357512e-05,
+      "loss": 3.8413,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485322298824068e-05,
+      "loss": 3.8332,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484483704073016e-05,
+      "loss": 3.8394,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483645109321963e-05,
+      "loss": 3.8551,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482808152451285e-05,
+      "loss": 3.8381,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481969557700233e-05,
+      "loss": 3.8346,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481130962949181e-05,
+      "loss": 3.8352,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480292368198129e-05,
+      "loss": 3.8392,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47945541132745e-05,
+      "loss": 3.824,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478616816576398e-05,
+      "loss": 3.8368,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4777782218253456e-05,
+      "loss": 3.8323,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4769396270742936e-05,
+      "loss": 3.8302,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476102670203615e-05,
+      "loss": 3.8222,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475264075452563e-05,
+      "loss": 3.823,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4744254807015105e-05,
+      "loss": 3.8298,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4735868859504585e-05,
+      "loss": 3.8333,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47274992907978e-05,
+      "loss": 3.8439,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471911334328728e-05,
+      "loss": 3.831,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471072739577676e-05,
+      "loss": 3.8194,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470234144826624e-05,
+      "loss": 3.8343,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693971879559456e-05,
+      "loss": 3.8318,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468558593204893e-05,
+      "loss": 3.8156,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467719998453841e-05,
+      "loss": 3.8173,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466881403702789e-05,
+      "loss": 3.826,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660444468321105e-05,
+      "loss": 3.8162,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465205852081058e-05,
+      "loss": 3.8381,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464367257330006e-05,
+      "loss": 3.8217,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463528662578954e-05,
+      "loss": 3.8248,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626917057082754e-05,
+      "loss": 3.8294,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618531109572234e-05,
+      "loss": 3.8257,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4610145162061714e-05,
+      "loss": 3.8243,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4601759214551194e-05,
+      "loss": 3.8113,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45933896458444e-05,
+      "loss": 3.7982,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458500369833388e-05,
+      "loss": 3.8394,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457661775082336e-05,
+      "loss": 3.8181,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456823180331284e-05,
+      "loss": 3.819,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455986223460605e-05,
+      "loss": 3.817,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455147628709553e-05,
+      "loss": 3.815,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454309033958501e-05,
+      "loss": 3.8082,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453470439207449e-05,
+      "loss": 3.822,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452633482336771e-05,
+      "loss": 3.8192,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451794887585719e-05,
+      "loss": 3.8142,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450956292834667e-05,
+      "loss": 3.834,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450117698083615e-05,
+      "loss": 3.8048,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449280741212936e-05,
+      "loss": 3.8023,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448442146461884e-05,
+      "loss": 3.8256,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447603551710832e-05,
+      "loss": 3.7929,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44676495695978e-05,
+      "loss": 3.8013,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4459280000891006e-05,
+      "loss": 3.823,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4450894053380486e-05,
+      "loss": 3.8236,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442508105869966e-05,
+      "loss": 3.8069,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4434122158359446e-05,
+      "loss": 3.8155,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442575258965266e-05,
+      "loss": 3.7968,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441736664214214e-05,
+      "loss": 3.8158,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440898069463162e-05,
+      "loss": 3.8131,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44005947471211e-05,
+      "loss": 3.8166,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439222517841431e-05,
+      "loss": 3.8114,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438383923090379e-05,
+      "loss": 3.8186,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437545328339327e-05,
+      "loss": 3.8253,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436706733588275e-05,
+      "loss": 3.8027,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435869776717596e-05,
+      "loss": 3.8108,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435031181966544e-05,
+      "loss": 3.8196,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434192587215492e-05,
+      "loss": 3.7979,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43335399246444e-05,
+      "loss": 3.8228,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4325170355937615e-05,
+      "loss": 3.8196,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316784408427095e-05,
+      "loss": 3.8192,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308398460916575e-05,
+      "loss": 3.8202,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4300012513406055e-05,
+      "loss": 3.8026,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291642944699264e-05,
+      "loss": 3.8011,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4283256997188744e-05,
+      "loss": 3.8087,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274871049678224e-05,
+      "loss": 3.8198,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266485102167704e-05,
+      "loss": 3.8086,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425811553346091e-05,
+      "loss": 3.8067,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424972958595039e-05,
+      "loss": 3.8047,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424134363843987e-05,
+      "loss": 3.8183,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423297406973308e-05,
+      "loss": 3.8,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422458812222257e-05,
+      "loss": 3.8003,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421620217471205e-05,
+      "loss": 3.8097,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420781622720153e-05,
+      "loss": 3.8199,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419943027969101e-05,
+      "loss": 3.8081,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419104433218049e-05,
+      "loss": 3.7948,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418265838466997e-05,
+      "loss": 3.8012,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417427243715945e-05,
+      "loss": 3.8058,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416590286845266e-05,
+      "loss": 3.8114,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415751692094214e-05,
+      "loss": 3.8056,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414913097343162e-05,
+      "loss": 3.8071,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41407450259211e-05,
+      "loss": 3.7905,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413237545721431e-05,
+      "loss": 3.8032,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412398950970379e-05,
+      "loss": 3.7907,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115603562193267e-05,
+      "loss": 3.8039,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4107217614682747e-05,
+      "loss": 3.7871,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409884804597596e-05,
+      "loss": 3.8154,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409047847726917e-05,
+      "loss": 3.7932,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408209252975865e-05,
+      "loss": 3.8031,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407370658224813e-05,
+      "loss": 3.7953,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406532063473761e-05,
+      "loss": 3.7975,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405693468722709e-05,
+      "loss": 3.7995,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404854873971657e-05,
+      "loss": 3.7995,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404016279220605e-05,
+      "loss": 3.8025,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403177684469553e-05,
+      "loss": 3.8098,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402340727598874e-05,
+      "loss": 3.8025,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401502132847822e-05,
+      "loss": 3.7999,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40066353809677e-05,
+      "loss": 3.8005,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399824943345719e-05,
+      "loss": 3.8066,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3989879864750396e-05,
+      "loss": 3.783,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981493917239876e-05,
+      "loss": 3.8074,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3973107969729356e-05,
+      "loss": 3.7932,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396472202221883e-05,
+      "loss": 3.8078,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3956352453512045e-05,
+      "loss": 3.7935,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947966506001525e-05,
+      "loss": 3.8008,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3939580558491005e-05,
+      "loss": 3.7853,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393119461098048e-05,
+      "loss": 3.7926,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922825042273694e-05,
+      "loss": 3.7941,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3914439094763174e-05,
+      "loss": 3.7973,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3906053147252654e-05,
+      "loss": 3.8005,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897667199742134e-05,
+      "loss": 3.8058,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388929763103535e-05,
+      "loss": 3.7855,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388091168352483e-05,
+      "loss": 3.7998,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38725257360143e-05,
+      "loss": 3.8052,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386413978850378e-05,
+      "loss": 3.8028,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855770219797e-05,
+      "loss": 3.7969,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384738427228648e-05,
+      "loss": 3.7978,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383899832477595e-05,
+      "loss": 3.792,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383061237726543e-05,
+      "loss": 3.8069,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382224280855865e-05,
+      "loss": 3.7872,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381385686104813e-05,
+      "loss": 3.8001,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380547091353761e-05,
+      "loss": 3.7912,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379708496602709e-05,
+      "loss": 3.7902,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37887153973203e-05,
+      "loss": 3.8065,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3780329449809777e-05,
+      "loss": 3.7942,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771943502299257e-05,
+      "loss": 3.7896,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763557554788736e-05,
+      "loss": 3.7806,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375518798608195e-05,
+      "loss": 3.7932,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8878939151763916,
+      "eval_runtime": 380.5554,
+      "eval_samples_per_second": 1002.721,
+      "eval_steps_per_second": 31.336,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3746802038571426e-05,
+      "loss": 3.7835,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3738416091060905e-05,
+      "loss": 3.7854,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3730030143550385e-05,
+      "loss": 3.8016,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37216605748436e-05,
+      "loss": 3.7921,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371327462733308e-05,
+      "loss": 3.8049,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370488867982256e-05,
+      "loss": 3.7877,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369650273231204e-05,
+      "loss": 3.7866,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368813316360525e-05,
+      "loss": 3.7826,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367974721609473e-05,
+      "loss": 3.7849,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367136126858421e-05,
+      "loss": 3.7934,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366297532107369e-05,
+      "loss": 3.7938,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36546057523669e-05,
+      "loss": 3.7913,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3646236183660115e-05,
+      "loss": 3.7815,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3637850236149595e-05,
+      "loss": 3.7779,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362946428863907e-05,
+      "loss": 3.7783,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3621078341128555e-05,
+      "loss": 3.7783,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3612692393618035e-05,
+      "loss": 3.7876,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3604306446107515e-05,
+      "loss": 3.7836,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3595920498596995e-05,
+      "loss": 3.7885,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3587534551086475e-05,
+      "loss": 3.802,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3579164982379684e-05,
+      "loss": 3.7897,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3570779034869164e-05,
+      "loss": 3.7774,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3562393087358644e-05,
+      "loss": 3.7837,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3554007139848124e-05,
+      "loss": 3.7873,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354563757114133e-05,
+      "loss": 3.7749,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353725162363081e-05,
+      "loss": 3.7847,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352886567612029e-05,
+      "loss": 3.7813,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352047972860977e-05,
+      "loss": 3.7804,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351211015990299e-05,
+      "loss": 3.7752,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350372421239247e-05,
+      "loss": 3.7682,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349533826488195e-05,
+      "loss": 3.7788,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348695231737143e-05,
+      "loss": 3.7823,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347858274866464e-05,
+      "loss": 3.7909,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347019680115412e-05,
+      "loss": 3.7822,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34618108536436e-05,
+      "loss": 3.7674,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345342490613308e-05,
+      "loss": 3.7863,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3445055337426287e-05,
+      "loss": 3.7815,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3436669389915766e-05,
+      "loss": 3.7622,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3428283442405246e-05,
+      "loss": 3.773,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419897494894726e-05,
+      "loss": 3.7743,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341152792618794e-05,
+      "loss": 3.7695,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340314197867742e-05,
+      "loss": 3.7856,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33947560311669e-05,
+      "loss": 3.7752,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338637008365638e-05,
+      "loss": 3.774,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337800051494959e-05,
+      "loss": 3.7778,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336961456743907e-05,
+      "loss": 3.7756,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336122861992855e-05,
+      "loss": 3.7812,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335284267241803e-05,
+      "loss": 3.76,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334447310371124e-05,
+      "loss": 3.7448,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333608715620072e-05,
+      "loss": 3.7919,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33277012086902e-05,
+      "loss": 3.771,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3319331639983416e-05,
+      "loss": 3.7767,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3310945692472896e-05,
+      "loss": 3.7662,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3302559744962376e-05,
+      "loss": 3.7654,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3294173797451856e-05,
+      "loss": 3.7627,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3285804228745065e-05,
+      "loss": 3.7681,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3277418281234545e-05,
+      "loss": 3.7742,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3269032333724025e-05,
+      "loss": 3.7618,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3260646386213505e-05,
+      "loss": 3.7859,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3252276817506714e-05,
+      "loss": 3.7621,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3243890869996194e-05,
+      "loss": 3.7525,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3235504922485674e-05,
+      "loss": 3.7771,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3227118974975154e-05,
+      "loss": 3.7489,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321874940626836e-05,
+      "loss": 3.7512,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321036345875785e-05,
+      "loss": 3.7747,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320197751124733e-05,
+      "loss": 3.7726,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319359156373681e-05,
+      "loss": 3.7605,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318522199503002e-05,
+      "loss": 3.7682,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31768360475195e-05,
+      "loss": 3.748,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316845010000898e-05,
+      "loss": 3.7702,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316006415249846e-05,
+      "loss": 3.7621,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315169458379167e-05,
+      "loss": 3.7732,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314330863628115e-05,
+      "loss": 3.7645,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313492268877063e-05,
+      "loss": 3.7691,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312653674126011e-05,
+      "loss": 3.7793,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3118167172553317e-05,
+      "loss": 3.7534,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31097812250428e-05,
+      "loss": 3.7666,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310139527753228e-05,
+      "loss": 3.7706,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309300933002176e-05,
+      "loss": 3.753,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308463976131497e-05,
+      "loss": 3.7754,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307625381380445e-05,
+      "loss": 3.7755,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306786786629393e-05,
+      "loss": 3.7695,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305948191878341e-05,
+      "loss": 3.7803,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305111235007662e-05,
+      "loss": 3.7542,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30427264025661e-05,
+      "loss": 3.7531,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303434045505558e-05,
+      "loss": 3.7625,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302595450754506e-05,
+      "loss": 3.7688,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301758493883827e-05,
+      "loss": 3.7688,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300919899132776e-05,
+      "loss": 3.7589,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300081304381724e-05,
+      "loss": 3.7578,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299242709630672e-05,
+      "loss": 3.7716,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2984057527599926e-05,
+      "loss": 3.7548,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975671580089406e-05,
+      "loss": 3.7564,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2967285632578886e-05,
+      "loss": 3.7622,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2958899685068366e-05,
+      "loss": 3.774,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2950530116361575e-05,
+      "loss": 3.7657,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2942144168851055e-05,
+      "loss": 3.7472,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2933758221340535e-05,
+      "loss": 3.7548,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2925372273830015e-05,
+      "loss": 3.7618,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2917002705123224e-05,
+      "loss": 3.7638,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290861675761271e-05,
+      "loss": 3.7617,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290023081010219e-05,
+      "loss": 3.7612,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289184486259167e-05,
+      "loss": 3.7449,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288347529388488e-05,
+      "loss": 3.7542,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287508934637436e-05,
+      "loss": 3.7527,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286670339886384e-05,
+      "loss": 3.7576,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285831745135332e-05,
+      "loss": 3.7417,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284994788264653e-05,
+      "loss": 3.7722,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284157831393974e-05,
+      "loss": 3.7493,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283319236642922e-05,
+      "loss": 3.7574,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28248064189187e-05,
+      "loss": 3.7464,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281642047140818e-05,
+      "loss": 3.7582,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2808034523897664e-05,
+      "loss": 3.7547,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2799648576387144e-05,
+      "loss": 3.7501,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2791262628876624e-05,
+      "loss": 3.7583,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27828766813661e-05,
+      "loss": 3.769,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277450711265931e-05,
+      "loss": 3.7518,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276612116514879e-05,
+      "loss": 3.7647,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275773521763827e-05,
+      "loss": 3.7512,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2749349270127746e-05,
+      "loss": 3.7618,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274097970142096e-05,
+      "loss": 3.7431,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273259375391044e-05,
+      "loss": 3.7615,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272420780639992e-05,
+      "loss": 3.7476,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27158218588894e-05,
+      "loss": 3.7618,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270745229018262e-05,
+      "loss": 3.7543,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26990663426721e-05,
+      "loss": 3.7549,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269068039516157e-05,
+      "loss": 3.7397,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268229444765105e-05,
+      "loss": 3.7493,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267392487894427e-05,
+      "loss": 3.7494,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266553893143375e-05,
+      "loss": 3.7564,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265715298392322e-05,
+      "loss": 3.7501,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26487670364127e-05,
+      "loss": 3.7621,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2640397467705916e-05,
+      "loss": 3.7452,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2632011520195396e-05,
+      "loss": 3.756,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262362557268487e-05,
+      "loss": 3.7587,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2615239625174356e-05,
+      "loss": 3.7613,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260687005646757e-05,
+      "loss": 3.7522,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2598484108957045e-05,
+      "loss": 3.7545,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2590098161446525e-05,
+      "loss": 3.7471,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2581712213936005e-05,
+      "loss": 3.767,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257334264522922e-05,
+      "loss": 3.7421,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2564956697718694e-05,
+      "loss": 3.7578,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2556570750208174e-05,
+      "loss": 3.7464,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2548184802697654e-05,
+      "loss": 3.7503,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253981523399087e-05,
+      "loss": 3.7592,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253142928648034e-05,
+      "loss": 3.7487,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252304333896982e-05,
+      "loss": 3.7527,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251465739145931e-05,
+      "loss": 3.7343,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250627144394879e-05,
+      "loss": 3.7526,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.87255859375,
+      "eval_runtime": 378.9148,
+      "eval_samples_per_second": 1007.063,
+      "eval_steps_per_second": 31.471,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249788549643827e-05,
+      "loss": 3.7402,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248949954892775e-05,
+      "loss": 3.7409,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248111360141723e-05,
+      "loss": 3.7576,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247272765390671e-05,
+      "loss": 3.7526,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246434170639619e-05,
+      "loss": 3.7598,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245595575888567e-05,
+      "loss": 3.7448,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244756981137515e-05,
+      "loss": 3.7449,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243918386386463e-05,
+      "loss": 3.7417,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243081429515784e-05,
+      "loss": 3.7453,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242242834764732e-05,
+      "loss": 3.7513,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24140424001368e-05,
+      "loss": 3.747,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240565645262628e-05,
+      "loss": 3.7477,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239730326272322e-05,
+      "loss": 3.7397,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23889173152127e-05,
+      "loss": 3.7372,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238053136770218e-05,
+      "loss": 3.7388,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237214542019166e-05,
+      "loss": 3.7325,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236375947268114e-05,
+      "loss": 3.7467,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235537352517062e-05,
+      "loss": 3.7429,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23469875776601e-05,
+      "loss": 3.7441,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233860163014958e-05,
+      "loss": 3.7582,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233023206144279e-05,
+      "loss": 3.7553,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232184611393227e-05,
+      "loss": 3.7394,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231346016642175e-05,
+      "loss": 3.739,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2305074218911225e-05,
+      "loss": 3.7456,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229670465020444e-05,
+      "loss": 3.7391,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228831870269393e-05,
+      "loss": 3.7408,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227993275518341e-05,
+      "loss": 3.7432,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227154680767288e-05,
+      "loss": 3.7409,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2263177238966096e-05,
+      "loss": 3.7334,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254791291455576e-05,
+      "loss": 3.7295,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2246405343945056e-05,
+      "loss": 3.7313,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223801939643453e-05,
+      "loss": 3.7411,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2229649827727745e-05,
+      "loss": 3.7498,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2221263880217225e-05,
+      "loss": 3.7462,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22128779327067e-05,
+      "loss": 3.7283,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220449198519618e-05,
+      "loss": 3.7412,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2196122416489394e-05,
+      "loss": 3.7426,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218773646897888e-05,
+      "loss": 3.7248,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2179350521468354e-05,
+      "loss": 3.7338,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2170964573957834e-05,
+      "loss": 3.7332,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216259500525105e-05,
+      "loss": 3.7263,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215420905774053e-05,
+      "loss": 3.7444,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214582311023e-05,
+      "loss": 3.734,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213745354152322e-05,
+      "loss": 3.7381,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21290675940127e-05,
+      "loss": 3.7362,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212068164650217e-05,
+      "loss": 3.7342,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211229569899165e-05,
+      "loss": 3.7406,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210392613028487e-05,
+      "loss": 3.7245,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209554018277435e-05,
+      "loss": 3.702,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208715423526383e-05,
+      "loss": 3.7493,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207876828775331e-05,
+      "loss": 3.7304,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2070398719046524e-05,
+      "loss": 3.7372,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2062012771536004e-05,
+      "loss": 3.7292,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205362682402548e-05,
+      "loss": 3.7267,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204524087651496e-05,
+      "loss": 3.72,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203687130780817e-05,
+      "loss": 3.7313,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2028485360297646e-05,
+      "loss": 3.736,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2020099412787126e-05,
+      "loss": 3.7216,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011713465276606e-05,
+      "loss": 3.7446,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200334389656982e-05,
+      "loss": 3.7269,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19949579490593e-05,
+      "loss": 3.715,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198657200154878e-05,
+      "loss": 3.7336,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197818605403826e-05,
+      "loss": 3.7164,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196981648533148e-05,
+      "loss": 3.7096,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196143053782095e-05,
+      "loss": 3.7383,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195304459031043e-05,
+      "loss": 3.7342,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194465864279991e-05,
+      "loss": 3.7226,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193628907409312e-05,
+      "loss": 3.7265,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19279031265826e-05,
+      "loss": 3.7126,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191951717907208e-05,
+      "loss": 3.7298,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191113123156156e-05,
+      "loss": 3.7272,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902761662854775e-05,
+      "loss": 3.7294,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1894375715344255e-05,
+      "loss": 3.7279,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1885989767833735e-05,
+      "loss": 3.7297,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1877603820323215e-05,
+      "loss": 3.7415,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1869234251616424e-05,
+      "loss": 3.715,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860848304105904e-05,
+      "loss": 3.731,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1852462356595384e-05,
+      "loss": 3.7263,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1844076409084864e-05,
+      "loss": 3.7162,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183570684037807e-05,
+      "loss": 3.7362,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182732089286755e-05,
+      "loss": 3.739,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181893494535703e-05,
+      "loss": 3.7285,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181054899784651e-05,
+      "loss": 3.7424,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180217942913973e-05,
+      "loss": 3.7155,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179379348162921e-05,
+      "loss": 3.7151,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178540753411869e-05,
+      "loss": 3.7241,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177702158660817e-05,
+      "loss": 3.7345,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176865201790138e-05,
+      "loss": 3.7235,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176026607039086e-05,
+      "loss": 3.7243,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175188012288034e-05,
+      "loss": 3.7237,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174349417536982e-05,
+      "loss": 3.7324,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173512460666303e-05,
+      "loss": 3.7175,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172673865915251e-05,
+      "loss": 3.7195,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171835271164199e-05,
+      "loss": 3.7262,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170996676413147e-05,
+      "loss": 3.7373,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170159719542468e-05,
+      "loss": 3.7256,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169321124791416e-05,
+      "loss": 3.7082,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168482530040364e-05,
+      "loss": 3.7217,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167645573169685e-05,
+      "loss": 3.7216,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166806978418633e-05,
+      "loss": 3.7294,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165968383667581e-05,
+      "loss": 3.722,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165129788916529e-05,
+      "loss": 3.7218,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16429283204585e-05,
+      "loss": 3.7113,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163454237294798e-05,
+      "loss": 3.7118,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162615642543746e-05,
+      "loss": 3.7213,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161777047792694e-05,
+      "loss": 3.7169,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1609400909220156e-05,
+      "loss": 3.7055,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1601014961709636e-05,
+      "loss": 3.736,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1592645393002846e-05,
+      "loss": 3.7126,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1584259445492325e-05,
+      "loss": 3.7161,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1575873497981805e-05,
+      "loss": 3.7117,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1567487550471285e-05,
+      "loss": 3.7192,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1559101602960765e-05,
+      "loss": 3.7214,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1550715655450245e-05,
+      "loss": 3.7144,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1542329707939725e-05,
+      "loss": 3.7209,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1533943760429205e-05,
+      "loss": 3.7279,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1525574191722414e-05,
+      "loss": 3.7191,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1517188244211894e-05,
+      "loss": 3.7274,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1508802296701374e-05,
+      "loss": 3.7121,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1500416349190854e-05,
+      "loss": 3.7249,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149204678048407e-05,
+      "loss": 3.7109,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148366083297355e-05,
+      "loss": 3.7227,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147527488546303e-05,
+      "loss": 3.7119,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146688893795251e-05,
+      "loss": 3.7245,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145851936924572e-05,
+      "loss": 3.7208,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14501334217352e-05,
+      "loss": 3.7178,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144174747422468e-05,
+      "loss": 3.7062,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143336152671416e-05,
+      "loss": 3.7124,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142499195800737e-05,
+      "loss": 3.7108,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141660601049685e-05,
+      "loss": 3.7232,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140822006298633e-05,
+      "loss": 3.7118,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139983411547581e-05,
+      "loss": 3.7282,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1391464546769024e-05,
+      "loss": 3.7068,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1383078599258504e-05,
+      "loss": 3.7199,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1374692651747984e-05,
+      "loss": 3.7213,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1366306704237464e-05,
+      "loss": 3.7258,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135793713553067e-05,
+      "loss": 3.7179,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134955118802015e-05,
+      "loss": 3.7174,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134116524050963e-05,
+      "loss": 3.7095,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133277929299911e-05,
+      "loss": 3.7317,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132440972429232e-05,
+      "loss": 3.705,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13160237767818e-05,
+      "loss": 3.7208,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130763782927128e-05,
+      "loss": 3.7129,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129925188176076e-05,
+      "loss": 3.7216,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1290898691857707e-05,
+      "loss": 3.7162,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1282512744347186e-05,
+      "loss": 3.713,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1274126796836666e-05,
+      "loss": 3.7169,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1265740849326146e-05,
+      "loss": 3.7007,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1257354901815626e-05,
+      "loss": 3.7186,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.861588954925537,
+      "eval_runtime": 302.8093,
+      "eval_samples_per_second": 1260.169,
+      "eval_steps_per_second": 39.381,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1248968954305106e-05,
+      "loss": 3.7079,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1240583006794586e-05,
+      "loss": 3.7024,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1232197059284066e-05,
+      "loss": 3.7181,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1223827490577275e-05,
+      "loss": 3.7178,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1215457921870484e-05,
+      "loss": 3.724,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1207071974359964e-05,
+      "loss": 3.7131,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119868602684945e-05,
+      "loss": 3.7098,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119030007933893e-05,
+      "loss": 3.7072,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118191413182841e-05,
+      "loss": 3.7073,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117352818431789e-05,
+      "loss": 3.7169,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11651586156111e-05,
+      "loss": 3.7098,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115677266810058e-05,
+      "loss": 3.7135,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114840309939379e-05,
+      "loss": 3.7097,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114001715188327e-05,
+      "loss": 3.703,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113163120437275e-05,
+      "loss": 3.6982,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112324525686223e-05,
+      "loss": 3.6996,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111485930935171e-05,
+      "loss": 3.714,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110647336184119e-05,
+      "loss": 3.7101,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109808741433067e-05,
+      "loss": 3.7048,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108970146682015e-05,
+      "loss": 3.7254,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108131551930963e-05,
+      "loss": 3.7192,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107292957179911e-05,
+      "loss": 3.707,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106454362428859e-05,
+      "loss": 3.708,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105615767677807e-05,
+      "loss": 3.7047,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104778810807128e-05,
+      "loss": 3.7067,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103940216056076e-05,
+      "loss": 3.7054,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103101621305024e-05,
+      "loss": 3.7091,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102263026553972e-05,
+      "loss": 3.7081,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101426069683293e-05,
+      "loss": 3.6975,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.100587474932241e-05,
+      "loss": 3.6967,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0997488801811887e-05,
+      "loss": 3.6933,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098910285430137e-05,
+      "loss": 3.71,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098073328559458e-05,
+      "loss": 3.7154,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097234733808406e-05,
+      "loss": 3.7106,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096396139057354e-05,
+      "loss": 3.6993,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095557544306302e-05,
+      "loss": 3.6972,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094720587435623e-05,
+      "loss": 3.7108,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093883630564944e-05,
+      "loss": 3.6965,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093045035813892e-05,
+      "loss": 3.6948,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09220644106284e-05,
+      "loss": 3.6995,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091367846311788e-05,
+      "loss": 3.695,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090529251560736e-05,
+      "loss": 3.7059,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089690656809684e-05,
+      "loss": 3.7014,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088852062058633e-05,
+      "loss": 3.7058,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088013467307581e-05,
+      "loss": 3.6989,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0871765104369016e-05,
+      "loss": 3.7045,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0863379156858496e-05,
+      "loss": 3.7071,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0854993209347976e-05,
+      "loss": 3.6902,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0846607261837456e-05,
+      "loss": 3.6684,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0838237693130665e-05,
+      "loss": 3.7145,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0829851745620145e-05,
+      "loss": 3.6981,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0821465798109625e-05,
+      "loss": 3.7037,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0813079850599105e-05,
+      "loss": 3.6959,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0804710281892314e-05,
+      "loss": 3.6968,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0796324334381794e-05,
+      "loss": 3.6831,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0787938386871274e-05,
+      "loss": 3.6975,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077956881816449e-05,
+      "loss": 3.7055,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077118287065397e-05,
+      "loss": 3.685,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076279692314345e-05,
+      "loss": 3.7125,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075441097563293e-05,
+      "loss": 3.6954,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074604140692614e-05,
+      "loss": 3.6805,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073765545941562e-05,
+      "loss": 3.6958,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07292695119051e-05,
+      "loss": 3.6848,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072088356439458e-05,
+      "loss": 3.6803,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071251399568779e-05,
+      "loss": 3.7019,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070412804817727e-05,
+      "loss": 3.7003,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069574210066675e-05,
+      "loss": 3.6895,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068735615315623e-05,
+      "loss": 3.6976,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0678986584449443e-05,
+      "loss": 3.6751,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067060063693892e-05,
+      "loss": 3.7,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06622146894284e-05,
+      "loss": 3.6916,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065382874191788e-05,
+      "loss": 3.6942,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064545917321109e-05,
+      "loss": 3.7016,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.063707322570057e-05,
+      "loss": 3.6929,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062868727819005e-05,
+      "loss": 3.712,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062030133067953e-05,
+      "loss": 3.683,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061193176197274e-05,
+      "loss": 3.6993,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060354581446222e-05,
+      "loss": 3.6924,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05951598669517e-05,
+      "loss": 3.686,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058679029824492e-05,
+      "loss": 3.7028,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05784043507344e-05,
+      "loss": 3.7047,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057001840322388e-05,
+      "loss": 3.6983,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056163245571336e-05,
+      "loss": 3.7099,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055324650820284e-05,
+      "loss": 3.6844,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054486056069232e-05,
+      "loss": 3.6814,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05364746131818e-05,
+      "loss": 3.6915,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052808866567128e-05,
+      "loss": 3.7042,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0519719096964486e-05,
+      "loss": 3.6894,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0511333149453966e-05,
+      "loss": 3.6952,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0502947201943446e-05,
+      "loss": 3.6903,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0494577633236655e-05,
+      "loss": 3.6994,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0486191685726135e-05,
+      "loss": 3.6894,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047780573821562e-05,
+      "loss": 3.6865,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04694197907051e-05,
+      "loss": 3.6929,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0461033843194575e-05,
+      "loss": 3.7051,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0452647895684055e-05,
+      "loss": 3.6949,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0444261948173535e-05,
+      "loss": 3.6747,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043589237946675e-05,
+      "loss": 3.6895,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0427506431956224e-05,
+      "loss": 3.6891,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0419120484445704e-05,
+      "loss": 3.6983,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041075091573892e-05,
+      "loss": 3.6924,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04023649682284e-05,
+      "loss": 3.6872,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039397902071787e-05,
+      "loss": 3.682,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038559307320736e-05,
+      "loss": 3.6788,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037720712569684e-05,
+      "loss": 3.689,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036882117818632e-05,
+      "loss": 3.6852,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03604352306758e-05,
+      "loss": 3.6758,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035206566196901e-05,
+      "loss": 3.7013,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0343696093262224e-05,
+      "loss": 3.6856,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03353101457517e-05,
+      "loss": 3.6831,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032692419824118e-05,
+      "loss": 3.6838,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031853825073066e-05,
+      "loss": 3.6826,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031015230322014e-05,
+      "loss": 3.6914,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030176635570962e-05,
+      "loss": 3.682,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02933804081991e-05,
+      "loss": 3.6933,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028499446068858e-05,
+      "loss": 3.6967,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027662489198179e-05,
+      "loss": 3.687,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026823894447127e-05,
+      "loss": 3.6941,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025985299696075e-05,
+      "loss": 3.6828,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025146704945023e-05,
+      "loss": 3.6953,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024308110193971e-05,
+      "loss": 3.6811,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023469515442919e-05,
+      "loss": 3.6851,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022630920691867e-05,
+      "loss": 3.6836,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021793963821188e-05,
+      "loss": 3.6956,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020955369070136e-05,
+      "loss": 3.6944,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020116774319084e-05,
+      "loss": 3.6842,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019278179568032e-05,
+      "loss": 3.6706,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01843958481698e-05,
+      "loss": 3.6865,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017600990065928e-05,
+      "loss": 3.6791,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016762395314876e-05,
+      "loss": 3.6901,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015925438444198e-05,
+      "loss": 3.6832,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015086843693146e-05,
+      "loss": 3.6969,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014248248942094e-05,
+      "loss": 3.6763,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013409654191041e-05,
+      "loss": 3.6925,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012571059439989e-05,
+      "loss": 3.6886,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011732464688937e-05,
+      "loss": 3.6975,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010893869937885e-05,
+      "loss": 3.6873,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010055275186833e-05,
+      "loss": 3.6823,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009218318316154e-05,
+      "loss": 3.6836,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008379723565102e-05,
+      "loss": 3.6998,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00754112881405e-05,
+      "loss": 3.6732,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006702534062998e-05,
+      "loss": 3.6909,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0058655771923195e-05,
+      "loss": 3.6866,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0050269824412675e-05,
+      "loss": 3.687,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0041900255705884e-05,
+      "loss": 3.6842,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0033514308195364e-05,
+      "loss": 3.6866,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0025128360684844e-05,
+      "loss": 3.685,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0016742413174324e-05,
+      "loss": 3.674,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000837284446753e-05,
+      "loss": 3.6853,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.854919910430908,
+      "eval_runtime": 302.963,
+      "eval_samples_per_second": 1259.53,
+      "eval_steps_per_second": 39.361,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999998689695701e-05,
+      "loss": 3.675,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999160094944649e-05,
+      "loss": 3.675,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998321500193597e-05,
+      "loss": 3.6875,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997482905442545e-05,
+      "loss": 3.6875,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996645948571867e-05,
+      "loss": 3.6975,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995807353820815e-05,
+      "loss": 3.6781,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994968759069763e-05,
+      "loss": 3.6815,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994130164318711e-05,
+      "loss": 3.6797,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993291569567659e-05,
+      "loss": 3.676,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992452974816607e-05,
+      "loss": 3.6905,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991616017945928e-05,
+      "loss": 3.6788,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990777423194876e-05,
+      "loss": 3.682,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989940466324197e-05,
+      "loss": 3.6861,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989101871573145e-05,
+      "loss": 3.6671,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988263276822093e-05,
+      "loss": 3.67,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987424682071041e-05,
+      "loss": 3.6676,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986586087319989e-05,
+      "loss": 3.6856,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985747492568937e-05,
+      "loss": 3.6773,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984908897817885e-05,
+      "loss": 3.6768,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984070303066833e-05,
+      "loss": 3.6935,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983231708315781e-05,
+      "loss": 3.6894,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982393113564729e-05,
+      "loss": 3.6754,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981554518813677e-05,
+      "loss": 3.6788,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9807159240626246e-05,
+      "loss": 3.6769,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979878967191946e-05,
+      "loss": 3.6783,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979040372440894e-05,
+      "loss": 3.6732,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978201777689842e-05,
+      "loss": 3.6758,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9773631829387895e-05,
+      "loss": 3.6849,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976526226068111e-05,
+      "loss": 3.667,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975687631317059e-05,
+      "loss": 3.6652,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974849036566007e-05,
+      "loss": 3.6698,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974012079695329e-05,
+      "loss": 3.6757,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973173484944277e-05,
+      "loss": 3.6825,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972334890193225e-05,
+      "loss": 3.6893,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971496295442172e-05,
+      "loss": 3.6696,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.97065770069112e-05,
+      "loss": 3.6717,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969819105940068e-05,
+      "loss": 3.6834,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9689821490693896e-05,
+      "loss": 3.6658,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968143554318337e-05,
+      "loss": 3.6656,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967304959567285e-05,
+      "loss": 3.674,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966466364816233e-05,
+      "loss": 3.6642,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965627770065181e-05,
+      "loss": 3.6737,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964789175314129e-05,
+      "loss": 3.6748,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9639505805630776e-05,
+      "loss": 3.6779,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9631119858120256e-05,
+      "loss": 3.6663,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9622750289413465e-05,
+      "loss": 3.6795,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9614364341902945e-05,
+      "loss": 3.6762,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9605978394392425e-05,
+      "loss": 3.665,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9597592446881905e-05,
+      "loss": 3.643,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9589222878175114e-05,
+      "loss": 3.6819,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9580836930664594e-05,
+      "loss": 3.6675,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9572450983154074e-05,
+      "loss": 3.6754,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9564065035643553e-05,
+      "loss": 3.6687,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955569546693676e-05,
+      "loss": 3.6636,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954730951942624e-05,
+      "loss": 3.6604,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953892357191572e-05,
+      "loss": 3.6664,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953055400320894e-05,
+      "loss": 3.6808,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952218443450215e-05,
+      "loss": 3.6548,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951379848699163e-05,
+      "loss": 3.6878,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950541253948111e-05,
+      "loss": 3.667,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949702659197059e-05,
+      "loss": 3.6518,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948864064446007e-05,
+      "loss": 3.669,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948025469694955e-05,
+      "loss": 3.6551,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947186874943903e-05,
+      "loss": 3.6542,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946348280192851e-05,
+      "loss": 3.6704,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945512961202545e-05,
+      "loss": 3.6737,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944674366451493e-05,
+      "loss": 3.6571,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943835771700441e-05,
+      "loss": 3.6736,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942997176949389e-05,
+      "loss": 3.6462,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942158582198337e-05,
+      "loss": 3.6691,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941319987447285e-05,
+      "loss": 3.6647,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940481392696233e-05,
+      "loss": 3.6649,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939644435825554e-05,
+      "loss": 3.6748,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938805841074502e-05,
+      "loss": 3.6657,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93796724632345e-05,
+      "loss": 3.6848,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937128651572398e-05,
+      "loss": 3.6574,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936290056821346e-05,
+      "loss": 3.6706,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935451462070294e-05,
+      "loss": 3.6619,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934612867319242e-05,
+      "loss": 3.6593,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933775910448563e-05,
+      "loss": 3.6773,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9329373156975117e-05,
+      "loss": 3.6725,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9320987209464596e-05,
+      "loss": 3.6747,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9312601261954076e-05,
+      "loss": 3.6815,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9304215314443556e-05,
+      "loss": 3.6589,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9295845745736766e-05,
+      "loss": 3.6534,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9287459798226245e-05,
+      "loss": 3.6631,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9279073850715725e-05,
+      "loss": 3.6768,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92706879032052e-05,
+      "loss": 3.6607,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926230195569468e-05,
+      "loss": 3.6661,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9253932386987894e-05,
+      "loss": 3.6681,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9245546439477374e-05,
+      "loss": 3.667,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9237160491966854e-05,
+      "loss": 3.6649,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9228774544456334e-05,
+      "loss": 3.6571,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9220388596945814e-05,
+      "loss": 3.6662,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9212002649435294e-05,
+      "loss": 3.6768,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9203616701924774e-05,
+      "loss": 3.6694,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9195230754414254e-05,
+      "loss": 3.6506,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918686118570746e-05,
+      "loss": 3.6605,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917847523819694e-05,
+      "loss": 3.6612,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917008929068642e-05,
+      "loss": 3.6688,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916171972197963e-05,
+      "loss": 3.6688,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915333377446911e-05,
+      "loss": 3.6589,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914494782695859e-05,
+      "loss": 3.6574,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913656187944807e-05,
+      "loss": 3.6487,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912819231074129e-05,
+      "loss": 3.6615,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911980636323077e-05,
+      "loss": 3.6601,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911142041572025e-05,
+      "loss": 3.6518,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910303446820973e-05,
+      "loss": 3.6712,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909464852069921e-05,
+      "loss": 3.6551,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908627895199242e-05,
+      "loss": 3.6576,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90778930044819e-05,
+      "loss": 3.6533,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906950705697138e-05,
+      "loss": 3.6539,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906112110946086e-05,
+      "loss": 3.6675,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905273516195034e-05,
+      "loss": 3.6525,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904434921443982e-05,
+      "loss": 3.6659,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9035963266929297e-05,
+      "loss": 3.6707,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9027593698222506e-05,
+      "loss": 3.6613,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901920775071199e-05,
+      "loss": 3.6678,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901082180320147e-05,
+      "loss": 3.6548,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900243585569095e-05,
+      "loss": 3.6683,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899404990818043e-05,
+      "loss": 3.6582,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898566396066991e-05,
+      "loss": 3.6593,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897727801315939e-05,
+      "loss": 3.6585,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89689084444526e-05,
+      "loss": 3.6655,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896052249694208e-05,
+      "loss": 3.6697,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895213654943156e-05,
+      "loss": 3.6538,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8943750601921034e-05,
+      "loss": 3.6485,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8935364654410514e-05,
+      "loss": 3.6543,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8926978706899994e-05,
+      "loss": 3.6554,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8918592759389474e-05,
+      "loss": 3.6642,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891022319068269e-05,
+      "loss": 3.6566,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890183724317217e-05,
+      "loss": 3.6673,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889345129566165e-05,
+      "loss": 3.6498,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888506534815113e-05,
+      "loss": 3.6708,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887667940064061e-05,
+      "loss": 3.6582,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886829345313009e-05,
+      "loss": 3.6671,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885990750561957e-05,
+      "loss": 3.6595,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885152155810905e-05,
+      "loss": 3.6546,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884315198940226e-05,
+      "loss": 3.6571,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883476604189174e-05,
+      "loss": 3.6751,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882638009438122e-05,
+      "loss": 3.6503,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88179941468707e-05,
+      "loss": 3.6644,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880962457816391e-05,
+      "loss": 3.659,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8801238630653395e-05,
+      "loss": 3.6614,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8792869061946604e-05,
+      "loss": 3.6585,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8784483114436084e-05,
+      "loss": 3.6586,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8776097166925564e-05,
+      "loss": 3.6565,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8767711219415044e-05,
+      "loss": 3.6564,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875934165070825e-05,
+      "loss": 3.6566,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8491063117980957,
+      "eval_runtime": 304.7266,
+      "eval_samples_per_second": 1252.241,
+      "eval_steps_per_second": 39.133,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.875095570319773e-05,
+      "loss": 3.6477,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874256975568721e-05,
+      "loss": 3.6474,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.873418380817669e-05,
+      "loss": 3.6588,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.872579786066617e-05,
+      "loss": 3.6626,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871742829195938e-05,
+      "loss": 3.6722,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870904234444886e-05,
+      "loss": 3.6495,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870065639693835e-05,
+      "loss": 3.6571,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869227044942783e-05,
+      "loss": 3.6549,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.868388450191731e-05,
+      "loss": 3.6473,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.867549855440679e-05,
+      "loss": 3.6634,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86671289857e-05,
+      "loss": 3.6509,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865874303818948e-05,
+      "loss": 3.6575,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8650373469482686e-05,
+      "loss": 3.6618,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8641987521972166e-05,
+      "loss": 3.6429,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8633601574461646e-05,
+      "loss": 3.6418,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8625215626951126e-05,
+      "loss": 3.6434,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8616829679440606e-05,
+      "loss": 3.6571,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8608443731930086e-05,
+      "loss": 3.6588,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8600057784419566e-05,
+      "loss": 3.6432,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8591671836909046e-05,
+      "loss": 3.6664,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858330226820226e-05,
+      "loss": 3.663,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857491632069174e-05,
+      "loss": 3.6503,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.856653037318122e-05,
+      "loss": 3.6532,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8558144425670695e-05,
+      "loss": 3.652,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854977485696391e-05,
+      "loss": 3.6527,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854138890945339e-05,
+      "loss": 3.6492,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853300296194287e-05,
+      "loss": 3.6469,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8524617014432344e-05,
+      "loss": 3.6576,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8516231066921824e-05,
+      "loss": 3.6426,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8507845119411304e-05,
+      "loss": 3.6395,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849947555070452e-05,
+      "loss": 3.6445,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8491089603194e-05,
+      "loss": 3.6508,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848270365568348e-05,
+      "loss": 3.652,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847431770817296e-05,
+      "loss": 3.6639,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846593176066244e-05,
+      "loss": 3.6448,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.845754581315192e-05,
+      "loss": 3.6452,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84491598656414e-05,
+      "loss": 3.6617,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.844079029693461e-05,
+      "loss": 3.6381,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843240434942409e-05,
+      "loss": 3.6425,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.842401840191357e-05,
+      "loss": 3.6485,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.841563245440305e-05,
+      "loss": 3.6384,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840724650689253e-05,
+      "loss": 3.6444,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839886055938201e-05,
+      "loss": 3.6499,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839047461187149e-05,
+      "loss": 3.6531,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8382105043164704e-05,
+      "loss": 3.6417,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8373719095654184e-05,
+      "loss": 3.654,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8365333148143664e-05,
+      "loss": 3.648,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8356947200633144e-05,
+      "loss": 3.6415,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8348561253122624e-05,
+      "loss": 3.621,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8340175305612104e-05,
+      "loss": 3.6558,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833180573690531e-05,
+      "loss": 3.6448,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832341978939479e-05,
+      "loss": 3.6532,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831503384188427e-05,
+      "loss": 3.6372,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830664789437375e-05,
+      "loss": 3.6399,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.829826194686323e-05,
+      "loss": 3.634,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828989237815644e-05,
+      "loss": 3.64,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828150643064592e-05,
+      "loss": 3.6591,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.82731204831354e-05,
+      "loss": 3.6312,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826473453562489e-05,
+      "loss": 3.6587,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.825634858811436e-05,
+      "loss": 3.645,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824796264060384e-05,
+      "loss": 3.6262,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823957669309332e-05,
+      "loss": 3.6398,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823120712438653e-05,
+      "loss": 3.6286,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822282117687601e-05,
+      "loss": 3.633,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821443522936549e-05,
+      "loss": 3.6428,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820604928185497e-05,
+      "loss": 3.6492,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819767971314818e-05,
+      "loss": 3.6327,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818929376563766e-05,
+      "loss": 3.647,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818090781812714e-05,
+      "loss": 3.6248,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8172521870616626e-05,
+      "loss": 3.6446,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8164135923106106e-05,
+      "loss": 3.6416,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8155749975595586e-05,
+      "loss": 3.6435,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8147380406888795e-05,
+      "loss": 3.6439,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8138994459378275e-05,
+      "loss": 3.6422,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8130608511867755e-05,
+      "loss": 3.6588,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8122222564357235e-05,
+      "loss": 3.6321,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8113852995650444e-05,
+      "loss": 3.6481,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8105467048139924e-05,
+      "loss": 3.6382,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8097081100629404e-05,
+      "loss": 3.6315,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8088695153118884e-05,
+      "loss": 3.6486,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8080309205608364e-05,
+      "loss": 3.6491,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807193963690158e-05,
+      "loss": 3.6482,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806355368939106e-05,
+      "loss": 3.6569,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805516774188054e-05,
+      "loss": 3.6363,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804678179437002e-05,
+      "loss": 3.6274,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80383958468595e-05,
+      "loss": 3.6365,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803000989934898e-05,
+      "loss": 3.6491,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802162395183846e-05,
+      "loss": 3.6381,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.801325438313167e-05,
+      "loss": 3.6435,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.800486843562115e-05,
+      "loss": 3.6403,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.799648248811063e-05,
+      "loss": 3.644,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798809654060011e-05,
+      "loss": 3.6412,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797972697189332e-05,
+      "loss": 3.633,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79713410243828e-05,
+      "loss": 3.6414,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796295507687228e-05,
+      "loss": 3.6533,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7954569129361764e-05,
+      "loss": 3.644,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7946183181851244e-05,
+      "loss": 3.624,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7937813613144454e-05,
+      "loss": 3.6389,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7929427665633933e-05,
+      "loss": 3.6392,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7921041718123413e-05,
+      "loss": 3.6444,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7912655770612893e-05,
+      "loss": 3.6438,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7904269823102367e-05,
+      "loss": 3.633,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7895883875591847e-05,
+      "loss": 3.6341,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788751430688506e-05,
+      "loss": 3.6263,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787912835937454e-05,
+      "loss": 3.6385,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7870742411864016e-05,
+      "loss": 3.6336,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78623564643535e-05,
+      "loss": 3.6287,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785397051684298e-05,
+      "loss": 3.6408,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.784560094813619e-05,
+      "loss": 3.6374,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.783721500062567e-05,
+      "loss": 3.633,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782882905311515e-05,
+      "loss": 3.6317,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782044310560463e-05,
+      "loss": 3.6283,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781205715809411e-05,
+      "loss": 3.6446,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780367121058359e-05,
+      "loss": 3.625,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.779528526307307e-05,
+      "loss": 3.6439,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778691569436628e-05,
+      "loss": 3.6462,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777852974685576e-05,
+      "loss": 3.6378,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777014379934524e-05,
+      "loss": 3.6441,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776175785183472e-05,
+      "loss": 3.635,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7753388283127936e-05,
+      "loss": 3.6431,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7745002335617416e-05,
+      "loss": 3.6356,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7736616388106896e-05,
+      "loss": 3.6325,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7728230440596376e-05,
+      "loss": 3.6363,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7719844493085856e-05,
+      "loss": 3.6392,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7711474924379065e-05,
+      "loss": 3.6429,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7703088976868545e-05,
+      "loss": 3.6327,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7694703029358025e-05,
+      "loss": 3.6242,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7686317081847505e-05,
+      "loss": 3.6348,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7677931134336985e-05,
+      "loss": 3.6256,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7669545186826465e-05,
+      "loss": 3.6437,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7661175618119674e-05,
+      "loss": 3.6325,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7652789670609154e-05,
+      "loss": 3.6423,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7644403723098634e-05,
+      "loss": 3.6256,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763601777558812e-05,
+      "loss": 3.6464,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76276318280776e-05,
+      "loss": 3.6324,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761924588056708e-05,
+      "loss": 3.6481,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761087631186029e-05,
+      "loss": 3.6365,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760249036434977e-05,
+      "loss": 3.633,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759410441683925e-05,
+      "loss": 3.6297,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758571846932873e-05,
+      "loss": 3.6528,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75773325218182e-05,
+      "loss": 3.6216,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756894657430768e-05,
+      "loss": 3.6426,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75605770056009e-05,
+      "loss": 3.6338,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755219105809038e-05,
+      "loss": 3.6357,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754380511057986e-05,
+      "loss": 3.6376,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753541916306934e-05,
+      "loss": 3.631,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7527049594362554e-05,
+      "loss": 3.635,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751866364685203e-05,
+      "loss": 3.633,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751027769934151e-05,
+      "loss": 3.6299,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8466107845306396,
+      "eval_runtime": 302.6188,
+      "eval_samples_per_second": 1260.962,
+      "eval_steps_per_second": 39.406,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750189175183099e-05,
+      "loss": 3.6191,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749350580432047e-05,
+      "loss": 3.6252,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748511985680995e-05,
+      "loss": 3.6356,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747673390929943e-05,
+      "loss": 3.6413,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7468364340592636e-05,
+      "loss": 3.6482,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7459978393082116e-05,
+      "loss": 3.6279,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7451592445571596e-05,
+      "loss": 3.6296,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7443206498061076e-05,
+      "loss": 3.6359,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7434820550550556e-05,
+      "loss": 3.6244,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.742645098184377e-05,
+      "loss": 3.633,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.741806503433325e-05,
+      "loss": 3.635,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740967908682273e-05,
+      "loss": 3.6347,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740130951811594e-05,
+      "loss": 3.6377,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.739292357060542e-05,
+      "loss": 3.6199,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73845376230949e-05,
+      "loss": 3.6186,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737615167558438e-05,
+      "loss": 3.6172,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736776572807386e-05,
+      "loss": 3.6411,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735937978056334e-05,
+      "loss": 3.632,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735099383305282e-05,
+      "loss": 3.6213,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73426078855423e-05,
+      "loss": 3.6408,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733423831683551e-05,
+      "loss": 3.6436,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7325852369324996e-05,
+      "loss": 3.6299,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7317466421814476e-05,
+      "loss": 3.6294,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7309080474303956e-05,
+      "loss": 3.6276,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7300710905597165e-05,
+      "loss": 3.6366,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7292324958086645e-05,
+      "loss": 3.6261,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7283939010576125e-05,
+      "loss": 3.6213,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7275553063065605e-05,
+      "loss": 3.6377,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7267167115555085e-05,
+      "loss": 3.6187,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7258781168044565e-05,
+      "loss": 3.6147,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7250411599337774e-05,
+      "loss": 3.6271,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7242025651827254e-05,
+      "loss": 3.6253,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7233639704316734e-05,
+      "loss": 3.6324,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7225253756806214e-05,
+      "loss": 3.6408,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.721688418809943e-05,
+      "loss": 3.6197,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720849824058891e-05,
+      "loss": 3.6264,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720011229307839e-05,
+      "loss": 3.6372,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719172634556786e-05,
+      "loss": 3.6145,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.718335677686108e-05,
+      "loss": 3.6212,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717497082935056e-05,
+      "loss": 3.6257,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.716658488184004e-05,
+      "loss": 3.6165,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.715819893432951e-05,
+      "loss": 3.6237,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714981298681899e-05,
+      "loss": 3.6298,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714142703930847e-05,
+      "loss": 3.631,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.713305747060168e-05,
+      "loss": 3.6209,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.712467152309117e-05,
+      "loss": 3.6279,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711628557558065e-05,
+      "loss": 3.6267,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.710789962807013e-05,
+      "loss": 3.6227,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709951368055961e-05,
+      "loss": 3.598,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7091144111852817e-05,
+      "loss": 3.6307,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7082758164342297e-05,
+      "loss": 3.6252,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7074372216831777e-05,
+      "loss": 3.6286,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7065986269321256e-05,
+      "loss": 3.6197,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7057600321810736e-05,
+      "loss": 3.6143,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7049230753103946e-05,
+      "loss": 3.6148,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7040844805593425e-05,
+      "loss": 3.6187,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7032458858082905e-05,
+      "loss": 3.6341,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7024072910572385e-05,
+      "loss": 3.6106,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7015686963061865e-05,
+      "loss": 3.6332,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.700731739435508e-05,
+      "loss": 3.6271,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699893144684456e-05,
+      "loss": 3.6061,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699054549933404e-05,
+      "loss": 3.621,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698215955182352e-05,
+      "loss": 3.6049,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6973773604313e-05,
+      "loss": 3.6112,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696538765680248e-05,
+      "loss": 3.6166,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695701808809569e-05,
+      "loss": 3.6306,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694863214058517e-05,
+      "loss": 3.6117,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694024619307465e-05,
+      "loss": 3.6248,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693186024556413e-05,
+      "loss": 3.6024,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.692347429805361e-05,
+      "loss": 3.619,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691508835054309e-05,
+      "loss": 3.6187,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.690670240303257e-05,
+      "loss": 3.6241,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6898332834325786e-05,
+      "loss": 3.6204,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6889946886815266e-05,
+      "loss": 3.6186,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6881560939304746e-05,
+      "loss": 3.6368,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6873174991794226e-05,
+      "loss": 3.6106,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6864805423087435e-05,
+      "loss": 3.6232,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6856419475576915e-05,
+      "loss": 3.6142,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6848033528066395e-05,
+      "loss": 3.6154,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683964758055587e-05,
+      "loss": 3.6285,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6831278011849084e-05,
+      "loss": 3.6212,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6822892064338564e-05,
+      "loss": 3.6274,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6814506116828044e-05,
+      "loss": 3.6354,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6806120169317523e-05,
+      "loss": 3.623,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6797734221807003e-05,
+      "loss": 3.6026,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6789348274296483e-05,
+      "loss": 3.6135,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.67809787055897e-05,
+      "loss": 3.6272,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677259275807917e-05,
+      "loss": 3.6173,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676420681056865e-05,
+      "loss": 3.6206,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675582086305813e-05,
+      "loss": 3.6197,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.674745129435134e-05,
+      "loss": 3.621,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673906534684082e-05,
+      "loss": 3.6185,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.67306793993303e-05,
+      "loss": 3.6123,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672229345181978e-05,
+      "loss": 3.6228,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.671390750430926e-05,
+      "loss": 3.6292,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.670552155679874e-05,
+      "loss": 3.6235,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669715198809196e-05,
+      "loss": 3.6014,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668876604058144e-05,
+      "loss": 3.6176,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668038009307092e-05,
+      "loss": 3.62,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66719941455604e-05,
+      "loss": 3.6215,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6663624576853606e-05,
+      "loss": 3.6219,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6655238629343086e-05,
+      "loss": 3.6103,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6646852681832566e-05,
+      "loss": 3.6101,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6638466734322046e-05,
+      "loss": 3.6061,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6630097165615255e-05,
+      "loss": 3.6147,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6621711218104735e-05,
+      "loss": 3.6141,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6613325270594215e-05,
+      "loss": 3.6106,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6604939323083695e-05,
+      "loss": 3.6225,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.659656975437691e-05,
+      "loss": 3.6127,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.658818380686639e-05,
+      "loss": 3.6101,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657979785935587e-05,
+      "loss": 3.612,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657141191184535e-05,
+      "loss": 3.6049,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656302596433483e-05,
+      "loss": 3.6207,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655464001682431e-05,
+      "loss": 3.6097,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654625406931379e-05,
+      "loss": 3.6166,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6537884500607e-05,
+      "loss": 3.6237,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652949855309648e-05,
+      "loss": 3.6179,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652111260558596e-05,
+      "loss": 3.6177,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651272665807544e-05,
+      "loss": 3.6162,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650435708936865e-05,
+      "loss": 3.618,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649597114185813e-05,
+      "loss": 3.6206,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6487585194347615e-05,
+      "loss": 3.6112,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6479199246837095e-05,
+      "loss": 3.6138,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6470829678130304e-05,
+      "loss": 3.6185,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6462443730619784e-05,
+      "loss": 3.6222,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6454057783109264e-05,
+      "loss": 3.6092,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6445671835598744e-05,
+      "loss": 3.6043,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6437285888088224e-05,
+      "loss": 3.6093,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6428899940577704e-05,
+      "loss": 3.6051,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642051399306718e-05,
+      "loss": 3.6227,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641214442436039e-05,
+      "loss": 3.6126,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.640375847684987e-05,
+      "loss": 3.622,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.639537252933935e-05,
+      "loss": 3.6087,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.638698658182883e-05,
+      "loss": 3.6201,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637860063431831e-05,
+      "loss": 3.6112,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637021468680779e-05,
+      "loss": 3.6279,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636184511810101e-05,
+      "loss": 3.6141,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635345917059048e-05,
+      "loss": 3.612,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634507322307996e-05,
+      "loss": 3.6108,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633668727556944e-05,
+      "loss": 3.6287,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632830132805892e-05,
+      "loss": 3.6063,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63199153805484e-05,
+      "loss": 3.618,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631154581184161e-05,
+      "loss": 3.6152,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630315986433109e-05,
+      "loss": 3.6152,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629477391682057e-05,
+      "loss": 3.6165,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628638796931005e-05,
+      "loss": 3.6127,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627801840060327e-05,
+      "loss": 3.6132,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6269632453092747e-05,
+      "loss": 3.6146,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6261246505582227e-05,
+      "loss": 3.6088,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.844973087310791,
+      "eval_runtime": 304.7522,
+      "eval_samples_per_second": 1252.135,
+      "eval_steps_per_second": 39.13,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6252860558071707e-05,
+      "loss": 3.6041,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6244474610561186e-05,
+      "loss": 3.6047,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6236105041854396e-05,
+      "loss": 3.6133,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6227719094343876e-05,
+      "loss": 3.6221,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6219333146833355e-05,
+      "loss": 3.6271,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6210947199322835e-05,
+      "loss": 3.6071,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6202561251812315e-05,
+      "loss": 3.6126,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6194175304301795e-05,
+      "loss": 3.6156,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6185789356791275e-05,
+      "loss": 3.6049,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617741978808449e-05,
+      "loss": 3.6067,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616903384057397e-05,
+      "loss": 3.6197,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616064789306345e-05,
+      "loss": 3.6123,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615227832435666e-05,
+      "loss": 3.6169,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.614389237684614e-05,
+      "loss": 3.5966,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.613550642933562e-05,
+      "loss": 3.598,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61271204818251e-05,
+      "loss": 3.5987,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611873453431458e-05,
+      "loss": 3.6172,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611034858680406e-05,
+      "loss": 3.6134,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610196263929354e-05,
+      "loss": 3.6025,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609357669178301e-05,
+      "loss": 3.6186,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608520712307623e-05,
+      "loss": 3.6236,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607682117556571e-05,
+      "loss": 3.606,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606843522805519e-05,
+      "loss": 3.61,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606004928054467e-05,
+      "loss": 3.6075,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6051679711837885e-05,
+      "loss": 3.6167,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6043293764327365e-05,
+      "loss": 3.603,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603490781681684e-05,
+      "loss": 3.6006,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602652186930632e-05,
+      "loss": 3.6208,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60181359217958e-05,
+      "loss": 3.5958,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600974997428528e-05,
+      "loss": 3.5958,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600138040557849e-05,
+      "loss": 3.6032,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599299445806797e-05,
+      "loss": 3.6047,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598460851055745e-05,
+      "loss": 3.6143,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597622256304693e-05,
+      "loss": 3.6208,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596783661553641e-05,
+      "loss": 3.596,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595946704682962e-05,
+      "loss": 3.6079,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59510810993191e-05,
+      "loss": 3.6155,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.594269515180858e-05,
+      "loss": 3.5918,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.593430920429806e-05,
+      "loss": 3.5995,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.592593963559127e-05,
+      "loss": 3.6084,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591755368808075e-05,
+      "loss": 3.5916,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590916774057023e-05,
+      "loss": 3.6057,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590078179305971e-05,
+      "loss": 3.6108,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589239584554919e-05,
+      "loss": 3.6086,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58840262768424e-05,
+      "loss": 3.5987,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587564032933188e-05,
+      "loss": 3.6075,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586725438182136e-05,
+      "loss": 3.6044,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585886843431085e-05,
+      "loss": 3.6039,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585048248680033e-05,
+      "loss": 3.5793,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5842112918093536e-05,
+      "loss": 3.6088,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5833726970583016e-05,
+      "loss": 3.6068,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5825341023072496e-05,
+      "loss": 3.6061,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5816955075561976e-05,
+      "loss": 3.5986,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5808585506855185e-05,
+      "loss": 3.5956,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5800199559344665e-05,
+      "loss": 3.594,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5791813611834145e-05,
+      "loss": 3.5966,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5783427664323625e-05,
+      "loss": 3.6175,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5775041716813105e-05,
+      "loss": 3.591,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5766672148106314e-05,
+      "loss": 3.6066,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57582862005958e-05,
+      "loss": 3.6065,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574990025308528e-05,
+      "loss": 3.5887,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574151430557476e-05,
+      "loss": 3.5972,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573312835806424e-05,
+      "loss": 3.5889,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572475878935745e-05,
+      "loss": 3.5928,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571637284184693e-05,
+      "loss": 3.5919,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570798689433641e-05,
+      "loss": 3.6134,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569960094682589e-05,
+      "loss": 3.5944,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569121499931537e-05,
+      "loss": 3.5968,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568284543060858e-05,
+      "loss": 3.5879,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567445948309806e-05,
+      "loss": 3.5973,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566607353558754e-05,
+      "loss": 3.6003,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565768758807702e-05,
+      "loss": 3.6017,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56493016405665e-05,
+      "loss": 3.6048,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5640932071859714e-05,
+      "loss": 3.6008,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5632546124349194e-05,
+      "loss": 3.6122,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5624160176838674e-05,
+      "loss": 3.5919,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561577422932815e-05,
+      "loss": 3.6029,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.560740466062136e-05,
+      "loss": 3.5962,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559901871311084e-05,
+      "loss": 3.5941,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559063276560032e-05,
+      "loss": 3.6066,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5582246818089796e-05,
+      "loss": 3.6004,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557387724938301e-05,
+      "loss": 3.6102,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556549130187249e-05,
+      "loss": 3.6142,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555710535436197e-05,
+      "loss": 3.6069,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554871940685145e-05,
+      "loss": 3.5838,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554033345934093e-05,
+      "loss": 3.5961,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553196389063415e-05,
+      "loss": 3.6036,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.552357794312362e-05,
+      "loss": 3.6027,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55151919956131e-05,
+      "loss": 3.5959,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550680604810258e-05,
+      "loss": 3.6,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54984364793958e-05,
+      "loss": 3.6012,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549005053188527e-05,
+      "loss": 3.5967,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548166458437475e-05,
+      "loss": 3.5931,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.547327863686423e-05,
+      "loss": 3.6029,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5464909068157446e-05,
+      "loss": 3.6043,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5456523120646926e-05,
+      "loss": 3.6094,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5448137173136406e-05,
+      "loss": 3.5818,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5439751225625886e-05,
+      "loss": 3.5964,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5431365278115366e-05,
+      "loss": 3.5987,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5422995709408575e-05,
+      "loss": 3.6038,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5414609761898055e-05,
+      "loss": 3.6021,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5406223814387535e-05,
+      "loss": 3.5913,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5397837866877015e-05,
+      "loss": 3.5938,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5389468298170224e-05,
+      "loss": 3.5815,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5381082350659704e-05,
+      "loss": 3.601,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5372696403149184e-05,
+      "loss": 3.5916,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5364310455638664e-05,
+      "loss": 3.5888,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5355924508128144e-05,
+      "loss": 3.6017,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534755493942136e-05,
+      "loss": 3.596,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533916899191084e-05,
+      "loss": 3.5887,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533078304440032e-05,
+      "loss": 3.5955,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53223970968898e-05,
+      "loss": 3.5836,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531402752818301e-05,
+      "loss": 3.6051,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.530564158067249e-05,
+      "loss": 3.5876,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.529725563316197e-05,
+      "loss": 3.6012,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528886968565145e-05,
+      "loss": 3.5962,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528050011694466e-05,
+      "loss": 3.6026,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527211416943414e-05,
+      "loss": 3.5985,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526372822192362e-05,
+      "loss": 3.5944,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52553422744131e-05,
+      "loss": 3.6036,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524695632690258e-05,
+      "loss": 3.5981,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523858675819579e-05,
+      "loss": 3.5906,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523020081068527e-05,
+      "loss": 3.5937,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522181486317475e-05,
+      "loss": 3.5977,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521342891566423e-05,
+      "loss": 3.6024,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520504296815371e-05,
+      "loss": 3.5916,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519667339944692e-05,
+      "loss": 3.5877,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51882874519364e-05,
+      "loss": 3.589,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517990150442588e-05,
+      "loss": 3.5872,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517151555691536e-05,
+      "loss": 3.599,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516314598820857e-05,
+      "loss": 3.5928,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515476004069805e-05,
+      "loss": 3.606,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.514637409318753e-05,
+      "loss": 3.5859,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.513798814567702e-05,
+      "loss": 3.6033,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512961857697023e-05,
+      "loss": 3.5954,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5121232629459707e-05,
+      "loss": 3.603,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5112846681949187e-05,
+      "loss": 3.5953,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5104460734438666e-05,
+      "loss": 3.5929,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5096074786928146e-05,
+      "loss": 3.5934,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5087705218221356e-05,
+      "loss": 3.606,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5079319270710836e-05,
+      "loss": 3.5894,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5070933323200315e-05,
+      "loss": 3.5938,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5062547375689795e-05,
+      "loss": 3.5996,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5054177806983005e-05,
+      "loss": 3.5949,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5045791859472484e-05,
+      "loss": 3.5944,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.503740591196197e-05,
+      "loss": 3.6014,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502901996445145e-05,
+      "loss": 3.5906,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502063401694093e-05,
+      "loss": 3.5926,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501226444823414e-05,
+      "loss": 3.5909,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8433103561401367,
+      "eval_runtime": 304.381,
+      "eval_samples_per_second": 1253.662,
+      "eval_steps_per_second": 39.178,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.500387850072362e-05,
+      "loss": 3.5985,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.49954925532131e-05,
+      "loss": 3.5792,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498710660570257e-05,
+      "loss": 3.5954,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497872065819205e-05,
+      "loss": 3.6008,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497033471068153e-05,
+      "loss": 3.6024,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496196514197475e-05,
+      "loss": 3.5917,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495357919446422e-05,
+      "loss": 3.5908,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494519324695371e-05,
+      "loss": 3.5945,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493680729944319e-05,
+      "loss": 3.5864,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492842135193267e-05,
+      "loss": 3.5898,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492005178322588e-05,
+      "loss": 3.595,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491166583571536e-05,
+      "loss": 3.5955,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4903296267008574e-05,
+      "loss": 3.603,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489491031949805e-05,
+      "loss": 3.5791,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.488652437198753e-05,
+      "loss": 3.5764,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487813842447701e-05,
+      "loss": 3.5746,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486975247696649e-05,
+      "loss": 3.6024,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486136652945597e-05,
+      "loss": 3.5904,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.485298058194545e-05,
+      "loss": 3.5851,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484459463443493e-05,
+      "loss": 3.5984,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483622506572814e-05,
+      "loss": 3.6027,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482783911821762e-05,
+      "loss": 3.5893,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48194531707071e-05,
+      "loss": 3.5895,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481106722319658e-05,
+      "loss": 3.588,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480269765448979e-05,
+      "loss": 3.6007,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479431170697927e-05,
+      "loss": 3.5834,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478592575946875e-05,
+      "loss": 3.5803,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477753981195823e-05,
+      "loss": 3.6024,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476915386444771e-05,
+      "loss": 3.573,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476078429574092e-05,
+      "loss": 3.5795,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.47523983482304e-05,
+      "loss": 3.5813,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.474401240071988e-05,
+      "loss": 3.5863,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.473562645320936e-05,
+      "loss": 3.5949,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4727256884502576e-05,
+      "loss": 3.6003,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4718870936992056e-05,
+      "loss": 3.5809,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4710484989481536e-05,
+      "loss": 3.5858,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4702099041971016e-05,
+      "loss": 3.597,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4693713094460496e-05,
+      "loss": 3.5747,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4685343525753705e-05,
+      "loss": 3.5861,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4676957578243185e-05,
+      "loss": 3.5868,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4668571630732665e-05,
+      "loss": 3.5744,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4660185683222145e-05,
+      "loss": 3.5829,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4651799735711625e-05,
+      "loss": 3.5965,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4643430167004834e-05,
+      "loss": 3.5859,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4635044219494314e-05,
+      "loss": 3.5814,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4626658271983794e-05,
+      "loss": 3.5938,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461827232447328e-05,
+      "loss": 3.5839,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460988637696276e-05,
+      "loss": 3.5869,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460150042945224e-05,
+      "loss": 3.5561,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459313086074545e-05,
+      "loss": 3.5842,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.458474491323493e-05,
+      "loss": 3.5934,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.457635896572441e-05,
+      "loss": 3.5848,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.456797301821388e-05,
+      "loss": 3.5784,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455958707070336e-05,
+      "loss": 3.5792,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455120112319284e-05,
+      "loss": 3.5729,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.454283155448606e-05,
+      "loss": 3.574,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.453444560697553e-05,
+      "loss": 3.6038,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.452605965946502e-05,
+      "loss": 3.5748,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45176737119545e-05,
+      "loss": 3.585,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4509304143247714e-05,
+      "loss": 3.5888,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450091819573719e-05,
+      "loss": 3.5706,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449253224822667e-05,
+      "loss": 3.5791,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448414630071615e-05,
+      "loss": 3.571,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4475776732009357e-05,
+      "loss": 3.5723,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4467390784498836e-05,
+      "loss": 3.5747,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4459004836988316e-05,
+      "loss": 3.5904,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4450618889477796e-05,
+      "loss": 3.5822,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4442232941967276e-05,
+      "loss": 3.5753,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4433863373260485e-05,
+      "loss": 3.5676,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442547742574997e-05,
+      "loss": 3.5791,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.441709147823945e-05,
+      "loss": 3.5801,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440870553072893e-05,
+      "loss": 3.5859,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440031958321841e-05,
+      "loss": 3.5833,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439195001451162e-05,
+      "loss": 3.5816,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.43835640670011e-05,
+      "loss": 3.596,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.437517811949058e-05,
+      "loss": 3.5736,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436679217198006e-05,
+      "loss": 3.5805,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435840622446954e-05,
+      "loss": 3.5831,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435003665576275e-05,
+      "loss": 3.578,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434165070825223e-05,
+      "loss": 3.5868,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433326476074171e-05,
+      "loss": 3.5862,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432487881323119e-05,
+      "loss": 3.5883,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4316509244524406e-05,
+      "loss": 3.597,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4308123297013886e-05,
+      "loss": 3.5855,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4299737349503366e-05,
+      "loss": 3.5648,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4291351401992846e-05,
+      "loss": 3.5752,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4282981833286055e-05,
+      "loss": 3.5855,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4274595885775535e-05,
+      "loss": 3.5871,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4266209938265015e-05,
+      "loss": 3.5777,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4257823990754495e-05,
+      "loss": 3.5863,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4249454422047704e-05,
+      "loss": 3.5865,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4241068474537184e-05,
+      "loss": 3.5768,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4232682527026664e-05,
+      "loss": 3.575,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4224296579516144e-05,
+      "loss": 3.5847,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421592701080936e-05,
+      "loss": 3.5817,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420754106329884e-05,
+      "loss": 3.5909,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419915511578832e-05,
+      "loss": 3.5713,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41907691682778e-05,
+      "loss": 3.5747,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418238322076728e-05,
+      "loss": 3.5838,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417401365206049e-05,
+      "loss": 3.5837,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.416562770454997e-05,
+      "loss": 3.5839,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.415724175703945e-05,
+      "loss": 3.572,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414885580952893e-05,
+      "loss": 3.5735,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414048624082214e-05,
+      "loss": 3.5671,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413210029331162e-05,
+      "loss": 3.5832,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41237143458011e-05,
+      "loss": 3.5709,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.411532839829058e-05,
+      "loss": 3.5717,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4106942450780064e-05,
+      "loss": 3.5823,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409857288207327e-05,
+      "loss": 3.5766,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409018693456275e-05,
+      "loss": 3.5755,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408180098705223e-05,
+      "loss": 3.5784,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407341503954171e-05,
+      "loss": 3.5633,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406504547083492e-05,
+      "loss": 3.5841,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40566595233244e-05,
+      "loss": 3.5758,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404827357581388e-05,
+      "loss": 3.5783,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403988762830336e-05,
+      "loss": 3.5817,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403151805959657e-05,
+      "loss": 3.5861,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402313211208605e-05,
+      "loss": 3.5789,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401474616457553e-05,
+      "loss": 3.5793,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400636021706501e-05,
+      "loss": 3.5851,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399797426955449e-05,
+      "loss": 3.577,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398960470084771e-05,
+      "loss": 3.5743,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398121875333719e-05,
+      "loss": 3.5803,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3972832805826667e-05,
+      "loss": 3.5767,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.396444685831614e-05,
+      "loss": 3.5853,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3956077289609356e-05,
+      "loss": 3.5683,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3947691342098836e-05,
+      "loss": 3.5744,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3939305394588316e-05,
+      "loss": 3.5679,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393091944707779e-05,
+      "loss": 3.5725,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392253349956727e-05,
+      "loss": 3.5829,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3914163930860485e-05,
+      "loss": 3.5739,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3905777983349965e-05,
+      "loss": 3.5869,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3897392035839444e-05,
+      "loss": 3.5729,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3889006088328924e-05,
+      "loss": 3.5822,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.388063651962214e-05,
+      "loss": 3.5785,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3872250572111613e-05,
+      "loss": 3.5831,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3863864624601093e-05,
+      "loss": 3.5754,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.385547867709057e-05,
+      "loss": 3.5765,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.384710910838379e-05,
+      "loss": 3.578,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383872316087326e-05,
+      "loss": 3.5899,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383033721336274e-05,
+      "loss": 3.5733,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382195126585222e-05,
+      "loss": 3.5745,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.381358169714544e-05,
+      "loss": 3.584,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380519574963492e-05,
+      "loss": 3.5763,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3796826180928134e-05,
+      "loss": 3.5736,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3788440233417614e-05,
+      "loss": 3.5837,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378005428590709e-05,
+      "loss": 3.5713,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377166833839657e-05,
+      "loss": 3.5768,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.376328239088605e-05,
+      "loss": 3.5721,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8440442085266113,
+      "eval_runtime": 302.1485,
+      "eval_samples_per_second": 1262.925,
+      "eval_steps_per_second": 39.467,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.375489644337553e-05,
+      "loss": 3.571,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.374651049586501e-05,
+      "loss": 3.5644,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.373812454835449e-05,
+      "loss": 3.5762,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372973860084397e-05,
+      "loss": 3.5888,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3721369032137176e-05,
+      "loss": 3.581,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3712983084626656e-05,
+      "loss": 3.5783,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.370459713711614e-05,
+      "loss": 3.5682,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.369621118960562e-05,
+      "loss": 3.5789,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.368784162089883e-05,
+      "loss": 3.5673,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367945567338831e-05,
+      "loss": 3.5767,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367106972587779e-05,
+      "loss": 3.5768,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366268377836727e-05,
+      "loss": 3.5782,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365433058846421e-05,
+      "loss": 3.5839,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.364594464095369e-05,
+      "loss": 3.5617,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363755869344317e-05,
+      "loss": 3.5606,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362917274593265e-05,
+      "loss": 3.5614,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362078679842213e-05,
+      "loss": 3.5851,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361240085091161e-05,
+      "loss": 3.5706,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3604014903401096e-05,
+      "loss": 3.5682,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3595628955890576e-05,
+      "loss": 3.5798,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3587259387183785e-05,
+      "loss": 3.588,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3578873439673265e-05,
+      "loss": 3.5736,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3570487492162745e-05,
+      "loss": 3.57,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3562101544652225e-05,
+      "loss": 3.5723,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3553731975945434e-05,
+      "loss": 3.584,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3545346028434914e-05,
+      "loss": 3.5686,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3536960080924394e-05,
+      "loss": 3.5628,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3528574133413874e-05,
+      "loss": 3.5857,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3520188185903354e-05,
+      "loss": 3.5552,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351181861719656e-05,
+      "loss": 3.5631,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.350343266968605e-05,
+      "loss": 3.5631,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.349504672217553e-05,
+      "loss": 3.5676,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.348666077466501e-05,
+      "loss": 3.5749,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.347829120595822e-05,
+      "loss": 3.5832,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34699052584477e-05,
+      "loss": 3.567,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346151931093718e-05,
+      "loss": 3.5681,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.345313336342666e-05,
+      "loss": 3.5784,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.344476379471987e-05,
+      "loss": 3.5606,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.343637784720935e-05,
+      "loss": 3.5676,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.342799189969883e-05,
+      "loss": 3.5677,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341960595218831e-05,
+      "loss": 3.5639,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341123638348152e-05,
+      "loss": 3.5611,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3402850435971004e-05,
+      "loss": 3.5813,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3394464488460484e-05,
+      "loss": 3.5657,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3386078540949964e-05,
+      "loss": 3.569,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.337770897224317e-05,
+      "loss": 3.574,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336932302473265e-05,
+      "loss": 3.5688,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336093707722213e-05,
+      "loss": 3.5683,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335255112971161e-05,
+      "loss": 3.5437,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334416518220109e-05,
+      "loss": 3.5634,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.33357956134943e-05,
+      "loss": 3.5748,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.332740966598378e-05,
+      "loss": 3.5722,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331902371847326e-05,
+      "loss": 3.558,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331063777096274e-05,
+      "loss": 3.5598,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.330226820225596e-05,
+      "loss": 3.5626,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.329388225474544e-05,
+      "loss": 3.5546,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.328549630723492e-05,
+      "loss": 3.5875,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32771103597244e-05,
+      "loss": 3.5536,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.326872441221387e-05,
+      "loss": 3.5722,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.326033846470335e-05,
+      "loss": 3.5717,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3251968895996566e-05,
+      "loss": 3.5521,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.324358294848604e-05,
+      "loss": 3.5626,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.323519700097552e-05,
+      "loss": 3.5524,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3226811053465e-05,
+      "loss": 3.5554,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3218441484758215e-05,
+      "loss": 3.5554,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3210055537247695e-05,
+      "loss": 3.5739,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3201669589737175e-05,
+      "loss": 3.5655,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3193283642226655e-05,
+      "loss": 3.5615,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3184897694716135e-05,
+      "loss": 3.5506,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3176528126009344e-05,
+      "loss": 3.5612,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3168142178498824e-05,
+      "loss": 3.565,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3159756230988304e-05,
+      "loss": 3.5683,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3151370283477784e-05,
+      "loss": 3.5605,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.314300071477099e-05,
+      "loss": 3.5668,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.313461476726047e-05,
+      "loss": 3.5763,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.312622881974995e-05,
+      "loss": 3.5597,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.311784287223943e-05,
+      "loss": 3.5627,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310947330353264e-05,
+      "loss": 3.5613,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310108735602213e-05,
+      "loss": 3.5657,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309270140851161e-05,
+      "loss": 3.5707,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.308431546100109e-05,
+      "loss": 3.5676,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30759458922943e-05,
+      "loss": 3.5722,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.306755994478378e-05,
+      "loss": 3.5764,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.305917399727326e-05,
+      "loss": 3.5684,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.305078804976274e-05,
+      "loss": 3.5509,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304240210225222e-05,
+      "loss": 3.5567,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.303403253354543e-05,
+      "loss": 3.5717,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.302564658603491e-05,
+      "loss": 3.5679,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.301726063852439e-05,
+      "loss": 3.5643,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3008874691013867e-05,
+      "loss": 3.5644,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.300050512230708e-05,
+      "loss": 3.5739,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299211917479656e-05,
+      "loss": 3.559,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.298373322728604e-05,
+      "loss": 3.558,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.297534727977552e-05,
+      "loss": 3.571,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.296697771106873e-05,
+      "loss": 3.5614,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295859176355821e-05,
+      "loss": 3.5788,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295020581604769e-05,
+      "loss": 3.5503,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294181986853717e-05,
+      "loss": 3.5572,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.293345029983038e-05,
+      "loss": 3.5667,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.292506435231986e-05,
+      "loss": 3.567,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.291667840480934e-05,
+      "loss": 3.5682,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.290829245729882e-05,
+      "loss": 3.5591,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.28999065097883e-05,
+      "loss": 3.5558,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2891536941081516e-05,
+      "loss": 3.5486,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2883150993570996e-05,
+      "loss": 3.5662,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2874765046060476e-05,
+      "loss": 3.5538,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2866379098549956e-05,
+      "loss": 3.5552,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2858009529843165e-05,
+      "loss": 3.5676,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2849623582332645e-05,
+      "loss": 3.5602,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2841237634822125e-05,
+      "loss": 3.5531,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2832851687311605e-05,
+      "loss": 3.5651,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2824482118604814e-05,
+      "loss": 3.5477,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2816096171094294e-05,
+      "loss": 3.5648,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2807710223583774e-05,
+      "loss": 3.5588,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2799324276073254e-05,
+      "loss": 3.561,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279095470736647e-05,
+      "loss": 3.5668,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278256875985595e-05,
+      "loss": 3.5663,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.277418281234543e-05,
+      "loss": 3.5604,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.276579686483491e-05,
+      "loss": 3.5625,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.275742729612812e-05,
+      "loss": 3.5701,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.27490413486176e-05,
+      "loss": 3.556,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.274065540110708e-05,
+      "loss": 3.5565,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273226945359656e-05,
+      "loss": 3.5616,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.272389988488977e-05,
+      "loss": 3.5666,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.271551393737925e-05,
+      "loss": 3.5668,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.270712798986873e-05,
+      "loss": 3.5546,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269874204235821e-05,
+      "loss": 3.5544,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269035609484769e-05,
+      "loss": 3.5561,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26819865261409e-05,
+      "loss": 3.554,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.267360057863038e-05,
+      "loss": 3.5619,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.266521463111986e-05,
+      "loss": 3.5636,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.265682868360934e-05,
+      "loss": 3.5664,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264845911490255e-05,
+      "loss": 3.5548,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264007316739203e-05,
+      "loss": 3.5667,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263168721988151e-05,
+      "loss": 3.5624,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.262330127237099e-05,
+      "loss": 3.5711,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26149317036642e-05,
+      "loss": 3.557,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.260654575615368e-05,
+      "loss": 3.5611,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.259815980864316e-05,
+      "loss": 3.5639,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258977386113264e-05,
+      "loss": 3.5707,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258140429242586e-05,
+      "loss": 3.5567,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.257301834491534e-05,
+      "loss": 3.5562,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.256463239740482e-05,
+      "loss": 3.5664,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25562464498943e-05,
+      "loss": 3.5597,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2547876881187506e-05,
+      "loss": 3.5573,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2539490933676986e-05,
+      "loss": 3.5709,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2531104986166466e-05,
+      "loss": 3.5482,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252271903865594e-05,
+      "loss": 3.5649,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2514349469949155e-05,
+      "loss": 3.5517,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8438479900360107,
+      "eval_runtime": 302.1612,
+      "eval_samples_per_second": 1262.872,
+      "eval_steps_per_second": 39.466,
+      "step": 1068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2505963522438635e-05,
+      "loss": 3.5569,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2497577574928115e-05,
+      "loss": 3.5489,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2489191627417595e-05,
+      "loss": 3.5581,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2480805679907075e-05,
+      "loss": 3.5726,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2472419732396555e-05,
+      "loss": 3.5659,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2464033784886035e-05,
+      "loss": 3.5622,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2455664216179244e-05,
+      "loss": 3.5493,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2447278268668724e-05,
+      "loss": 3.563,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2438892321158204e-05,
+      "loss": 3.5547,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2430506373647684e-05,
+      "loss": 3.5568,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242213680494089e-05,
+      "loss": 3.5602,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.241375085743037e-05,
+      "loss": 3.563,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.240538128872359e-05,
+      "loss": 3.5675,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.239699534121307e-05,
+      "loss": 3.5459,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238860939370255e-05,
+      "loss": 3.5405,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238022344619203e-05,
+      "loss": 3.5472,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.237183749868151e-05,
+      "loss": 3.5647,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.236345155117099e-05,
+      "loss": 3.5568,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.235506560366047e-05,
+      "loss": 3.5533,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.234667965614995e-05,
+      "loss": 3.5606,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.233831008744316e-05,
+      "loss": 3.5734,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232992413993264e-05,
+      "loss": 3.5551,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232153819242212e-05,
+      "loss": 3.554,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.23131522449116e-05,
+      "loss": 3.5576,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.230476629740108e-05,
+      "loss": 3.5683,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2296396728694286e-05,
+      "loss": 3.5491,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2288010781183766e-05,
+      "loss": 3.5497,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227962483367325e-05,
+      "loss": 3.5678,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227123888616273e-05,
+      "loss": 3.5444,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.226286931745594e-05,
+      "loss": 3.5436,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.225448336994542e-05,
+      "loss": 3.5451,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.22460974224349e-05,
+      "loss": 3.5571,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223771147492438e-05,
+      "loss": 3.5578,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222934190621759e-05,
+      "loss": 3.5656,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222095595870707e-05,
+      "loss": 3.5536,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221257001119655e-05,
+      "loss": 3.5501,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220418406368603e-05,
+      "loss": 3.5632,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.219579811617551e-05,
+      "loss": 3.5435,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.218742854746872e-05,
+      "loss": 3.5503,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.217904259995821e-05,
+      "loss": 3.5499,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2170656652447687e-05,
+      "loss": 3.551,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2162270704937167e-05,
+      "loss": 3.5419,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2153901136230376e-05,
+      "loss": 3.5656,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2145515188719856e-05,
+      "loss": 3.5481,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2137129241209336e-05,
+      "loss": 3.5556,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2128743293698816e-05,
+      "loss": 3.5573,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2120373724992025e-05,
+      "loss": 3.5502,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2111987777481505e-05,
+      "loss": 3.5532,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2103601829970985e-05,
+      "loss": 3.5313,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2095215882460464e-05,
+      "loss": 3.5471,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2086846313753674e-05,
+      "loss": 3.558,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207846036624316e-05,
+      "loss": 3.5517,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207007441873264e-05,
+      "loss": 3.543,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206168847122212e-05,
+      "loss": 3.5421,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.205331890251533e-05,
+      "loss": 3.549,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.204493295500481e-05,
+      "loss": 3.5364,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.203654700749429e-05,
+      "loss": 3.5709,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.202816105998377e-05,
+      "loss": 3.5427,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201977511247325e-05,
+      "loss": 3.5516,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201140554376646e-05,
+      "loss": 3.5568,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.200301959625594e-05,
+      "loss": 3.5384,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.199463364874542e-05,
+      "loss": 3.5437,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.19862477012349e-05,
+      "loss": 3.5414,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1977878132528114e-05,
+      "loss": 3.5363,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1969492185017594e-05,
+      "loss": 3.5351,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1961106237507074e-05,
+      "loss": 3.5592,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1952720289996554e-05,
+      "loss": 3.5486,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.194435072128976e-05,
+      "loss": 3.5482,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.193596477377924e-05,
+      "loss": 3.5343,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.192757882626872e-05,
+      "loss": 3.5427,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1919192878758196e-05,
+      "loss": 3.5499,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1910806931247676e-05,
+      "loss": 3.553,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190243736254089e-05,
+      "loss": 3.5461,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.189405141503037e-05,
+      "loss": 3.5527,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.188566546751985e-05,
+      "loss": 3.5576,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.187727952000933e-05,
+      "loss": 3.5482,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.186890995130255e-05,
+      "loss": 3.5435,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.186052400379202e-05,
+      "loss": 3.5496,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.18521380562815e-05,
+      "loss": 3.5494,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.184375210877098e-05,
+      "loss": 3.5505,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.183536616126046e-05,
+      "loss": 3.5535,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.182699659255367e-05,
+      "loss": 3.5549,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181861064504315e-05,
+      "loss": 3.5612,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181022469753263e-05,
+      "loss": 3.5501,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180183875002211e-05,
+      "loss": 3.537,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.179345280251159e-05,
+      "loss": 3.543,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1785083233804805e-05,
+      "loss": 3.5575,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1776697286294285e-05,
+      "loss": 3.5525,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1768311338783765e-05,
+      "loss": 3.5493,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1759925391273245e-05,
+      "loss": 3.5465,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1751555822566454e-05,
+      "loss": 3.5558,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1743169875055934e-05,
+      "loss": 3.5438,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1734783927545414e-05,
+      "loss": 3.5461,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1726397980034894e-05,
+      "loss": 3.5471,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.17180284113281e-05,
+      "loss": 3.5483,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170964246381758e-05,
+      "loss": 3.556,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170125651630706e-05,
+      "loss": 3.5473,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169287056879654e-05,
+      "loss": 3.536,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.168448462128602e-05,
+      "loss": 3.5524,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.167611505257924e-05,
+      "loss": 3.5503,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.166772910506872e-05,
+      "loss": 3.5519,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16593431575582e-05,
+      "loss": 3.5467,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.165095721004768e-05,
+      "loss": 3.5389,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164258764134089e-05,
+      "loss": 3.5371,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.163420169383037e-05,
+      "loss": 3.5478,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.162581574631985e-05,
+      "loss": 3.5394,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.161742979880933e-05,
+      "loss": 3.5385,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160906023010254e-05,
+      "loss": 3.5533,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160067428259202e-05,
+      "loss": 3.547,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.15922883350815e-05,
+      "loss": 3.539,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.158390238757098e-05,
+      "loss": 3.5479,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157553281886419e-05,
+      "loss": 3.5348,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.156714687135367e-05,
+      "loss": 3.548,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.155876092384315e-05,
+      "loss": 3.543,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.155037497633263e-05,
+      "loss": 3.5461,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154200540762584e-05,
+      "loss": 3.5481,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.153361946011532e-05,
+      "loss": 3.555,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.15252335126048e-05,
+      "loss": 3.5449,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.151684756509428e-05,
+      "loss": 3.5472,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150847799638749e-05,
+      "loss": 3.5527,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150009204887697e-05,
+      "loss": 3.5433,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.149170610136645e-05,
+      "loss": 3.5369,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.148332015385593e-05,
+      "loss": 3.5499,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1474950585149146e-05,
+      "loss": 3.5487,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1466564637638626e-05,
+      "loss": 3.553,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1458178690128106e-05,
+      "loss": 3.5375,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1449792742617586e-05,
+      "loss": 3.5395,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1441406795107066e-05,
+      "loss": 3.5418,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1433020847596546e-05,
+      "loss": 3.5435,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1424634900086026e-05,
+      "loss": 3.5382,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1416265331379235e-05,
+      "loss": 3.5499,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1407879383868715e-05,
+      "loss": 3.5516,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1399493436358195e-05,
+      "loss": 3.5404,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1391107488847675e-05,
+      "loss": 3.5487,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1382737920140884e-05,
+      "loss": 3.549,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1374351972630364e-05,
+      "loss": 3.5549,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1365966025119844e-05,
+      "loss": 3.5388,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.135758007760933e-05,
+      "loss": 3.5473,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1349194130098804e-05,
+      "loss": 3.5516,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.134082456139202e-05,
+      "loss": 3.5521,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.13324386138815e-05,
+      "loss": 3.5421,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.132405266637098e-05,
+      "loss": 3.5423,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.131566671886045e-05,
+      "loss": 3.5502,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.130729715015367e-05,
+      "loss": 3.5423,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.129891120264315e-05,
+      "loss": 3.5418,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.129054163393636e-05,
+      "loss": 3.5556,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128215568642584e-05,
+      "loss": 3.5363,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.127376973891532e-05,
+      "loss": 3.5473,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.12653837914048e-05,
+      "loss": 3.5396,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.844848155975342,
+      "eval_runtime": 301.655,
+      "eval_samples_per_second": 1264.991,
+      "eval_steps_per_second": 39.532,
+      "step": 1144800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.125699784389428e-05,
+      "loss": 3.5383,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.124861189638376e-05,
+      "loss": 3.535,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.124022594887324e-05,
+      "loss": 3.5414,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123184000136272e-05,
+      "loss": 3.5559,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.122347043265593e-05,
+      "loss": 3.5487,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.121508448514541e-05,
+      "loss": 3.5482,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1206698537634887e-05,
+      "loss": 3.5364,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1198312590124367e-05,
+      "loss": 3.5461,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1189926642613846e-05,
+      "loss": 3.5373,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1181557073907056e-05,
+      "loss": 3.5444,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1173171126396536e-05,
+      "loss": 3.5428,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.116478517888602e-05,
+      "loss": 3.5489,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.115641561017923e-05,
+      "loss": 3.5542,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.114804604147245e-05,
+      "loss": 3.5357,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.113966009396193e-05,
+      "loss": 3.5266,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.11312741464514e-05,
+      "loss": 3.5295,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.112288819894088e-05,
+      "loss": 3.5454,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.111450225143036e-05,
+      "loss": 3.5432,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.110611630391984e-05,
+      "loss": 3.5377,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.109773035640932e-05,
+      "loss": 3.5397,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.108936078770253e-05,
+      "loss": 3.5631,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.108097484019201e-05,
+      "loss": 3.5386,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107258889268149e-05,
+      "loss": 3.5364,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1064202945170976e-05,
+      "loss": 3.5426,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1055833376464185e-05,
+      "loss": 3.554,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1047447428953665e-05,
+      "loss": 3.5302,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1039061481443145e-05,
+      "loss": 3.5368,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1030675533932625e-05,
+      "loss": 3.5486,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1022289586422105e-05,
+      "loss": 3.5347,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1013920017715314e-05,
+      "loss": 3.5279,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1005534070204794e-05,
+      "loss": 3.5252,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0997148122694274e-05,
+      "loss": 3.5412,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0988762175183754e-05,
+      "loss": 3.5481,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.098039260647696e-05,
+      "loss": 3.5468,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097200665896644e-05,
+      "loss": 3.5412,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.096362071145593e-05,
+      "loss": 3.5334,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.095523476394541e-05,
+      "loss": 3.5462,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.094686519523862e-05,
+      "loss": 3.5298,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09384792477281e-05,
+      "loss": 3.537,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.093009330021758e-05,
+      "loss": 3.5333,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092170735270706e-05,
+      "loss": 3.5354,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.091332140519654e-05,
+      "loss": 3.5267,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.090495183648975e-05,
+      "loss": 3.5517,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.089656588897923e-05,
+      "loss": 3.5334,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.088817994146871e-05,
+      "loss": 3.5397,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087979399395819e-05,
+      "loss": 3.5435,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0871424425251397e-05,
+      "loss": 3.5349,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0863038477740877e-05,
+      "loss": 3.5375,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.085465253023036e-05,
+      "loss": 3.5176,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.084626658271984e-05,
+      "loss": 3.5314,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.083789701401305e-05,
+      "loss": 3.5429,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.082951106650253e-05,
+      "loss": 3.5382,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.082112511899201e-05,
+      "loss": 3.528,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.081273917148149e-05,
+      "loss": 3.526,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.08043696027747e-05,
+      "loss": 3.5343,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.079598365526418e-05,
+      "loss": 3.5255,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.078759770775366e-05,
+      "loss": 3.5509,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.077921176024314e-05,
+      "loss": 3.5286,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.077082581273262e-05,
+      "loss": 3.535,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076245624402583e-05,
+      "loss": 3.5477,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.075407029651532e-05,
+      "loss": 3.5198,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.07456843490048e-05,
+      "loss": 3.5251,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.073729840149428e-05,
+      "loss": 3.5284,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0728928832787486e-05,
+      "loss": 3.5218,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0720542885276966e-05,
+      "loss": 3.5251,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0712156937766446e-05,
+      "loss": 3.5402,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0703770990255926e-05,
+      "loss": 3.5379,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0695401421549135e-05,
+      "loss": 3.5319,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0687015474038615e-05,
+      "loss": 3.5191,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0678629526528095e-05,
+      "loss": 3.528,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0670243579017575e-05,
+      "loss": 3.5325,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0661857631507055e-05,
+      "loss": 3.5353,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.065348806280027e-05,
+      "loss": 3.5344,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.064510211528975e-05,
+      "loss": 3.5364,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.063671616777923e-05,
+      "loss": 3.5427,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0628330220268704e-05,
+      "loss": 3.5352,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.061996065156192e-05,
+      "loss": 3.5241,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.06115747040514e-05,
+      "loss": 3.5373,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.060318875654088e-05,
+      "loss": 3.5324,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.059480280903035e-05,
+      "loss": 3.533,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.058641686151983e-05,
+      "loss": 3.5437,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.057804729281305e-05,
+      "loss": 3.5381,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056966134530253e-05,
+      "loss": 3.5483,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056127539779201e-05,
+      "loss": 3.5364,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055288945028149e-05,
+      "loss": 3.5224,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0544519881574704e-05,
+      "loss": 3.5279,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.053613393406418e-05,
+      "loss": 3.534,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.052774798655366e-05,
+      "loss": 3.5416,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.051936203904314e-05,
+      "loss": 3.5342,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0510992470336353e-05,
+      "loss": 3.5312,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0502606522825826e-05,
+      "loss": 3.5412,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0494220575315306e-05,
+      "loss": 3.5298,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0485834627804786e-05,
+      "loss": 3.5322,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0477465059098002e-05,
+      "loss": 3.5316,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.046907911158748e-05,
+      "loss": 3.5359,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.046069316407696e-05,
+      "loss": 3.5444,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.045230721656644e-05,
+      "loss": 3.5295,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.044393764785965e-05,
+      "loss": 3.5214,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.043555170034913e-05,
+      "loss": 3.5353,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.042716575283861e-05,
+      "loss": 3.5345,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.041877980532809e-05,
+      "loss": 3.5397,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.04104102366213e-05,
+      "loss": 3.5295,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.040202428911078e-05,
+      "loss": 3.5268,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.039363834160026e-05,
+      "loss": 3.5196,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.038525239408974e-05,
+      "loss": 3.5372,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0376866446579223e-05,
+      "loss": 3.5222,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0368496877872432e-05,
+      "loss": 3.5278,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0360110930361912e-05,
+      "loss": 3.5333,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0351724982851392e-05,
+      "loss": 3.5292,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0343339035340872e-05,
+      "loss": 3.5247,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0334969466634085e-05,
+      "loss": 3.5302,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0326583519123565e-05,
+      "loss": 3.5231,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0318197571613045e-05,
+      "loss": 3.5332,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0309811624102525e-05,
+      "loss": 3.5279,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0301442055395734e-05,
+      "loss": 3.5322,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0293056107885214e-05,
+      "loss": 3.5309,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0284670160374694e-05,
+      "loss": 3.5411,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0276284212864177e-05,
+      "loss": 3.531,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0267914644157386e-05,
+      "loss": 3.5316,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0259528696646866e-05,
+      "loss": 3.5369,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0251142749136346e-05,
+      "loss": 3.5339,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0242756801625826e-05,
+      "loss": 3.5168,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0234370854115306e-05,
+      "loss": 3.5359,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.022600128540852e-05,
+      "loss": 3.5351,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0217615337897998e-05,
+      "loss": 3.5391,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0209229390387478e-05,
+      "loss": 3.5221,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0200843442876958e-05,
+      "loss": 3.5264,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0192473874170167e-05,
+      "loss": 3.5246,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0184087926659647e-05,
+      "loss": 3.5268,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.017570197914913e-05,
+      "loss": 3.5242,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.016731603163861e-05,
+      "loss": 3.5368,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.015894646293182e-05,
+      "loss": 3.5395,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.01505605154213e-05,
+      "loss": 3.5287,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.014217456791078e-05,
+      "loss": 3.5291,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.013378862040026e-05,
+      "loss": 3.5351,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0125419051693472e-05,
+      "loss": 3.5372,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0117033104182952e-05,
+      "loss": 3.5325,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0108647156672432e-05,
+      "loss": 3.5311,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0100261209161912e-05,
+      "loss": 3.5369,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.009189164045512e-05,
+      "loss": 3.5337,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.00835056929446e-05,
+      "loss": 3.5288,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0075119745434084e-05,
+      "loss": 3.5261,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0066733797923564e-05,
+      "loss": 3.5364,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0058364229216773e-05,
+      "loss": 3.5293,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0049978281706253e-05,
+      "loss": 3.5229,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0041608712999462e-05,
+      "loss": 3.5434,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0033222765488946e-05,
+      "loss": 3.5239,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0024836817978426e-05,
+      "loss": 3.5291,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0016450870467906e-05,
+      "loss": 3.5247,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.0008064922957386e-05,
+      "loss": 3.5253,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8455498218536377,
+      "eval_runtime": 301.971,
+      "eval_samples_per_second": 1263.668,
+      "eval_steps_per_second": 39.491,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9999678975446866e-05,
+      "loss": 3.5168,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9991293027936345e-05,
+      "loss": 3.527,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9982907080425825e-05,
+      "loss": 3.5394,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9974521132915305e-05,
+      "loss": 3.5363,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9966151564208518e-05,
+      "loss": 3.5365,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9957765616697998e-05,
+      "loss": 3.5201,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9949379669187478e-05,
+      "loss": 3.53,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9940993721676958e-05,
+      "loss": 3.5217,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9932624152970167e-05,
+      "loss": 3.5331,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9924238205459647e-05,
+      "loss": 3.5303,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.991585225794913e-05,
+      "loss": 3.5351,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.990746631043861e-05,
+      "loss": 3.5333,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.989909674173182e-05,
+      "loss": 3.5241,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.98907107942213e-05,
+      "loss": 3.513,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.988232484671078e-05,
+      "loss": 3.519,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9873938899200256e-05,
+      "loss": 3.5225,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9865552951689736e-05,
+      "loss": 3.532,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9857167004179216e-05,
+      "loss": 3.5238,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9848781056668696e-05,
+      "loss": 3.5282,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9840395109158176e-05,
+      "loss": 3.5501,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9832025540451385e-05,
+      "loss": 3.5266,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9823639592940868e-05,
+      "loss": 3.523,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9815253645430348e-05,
+      "loss": 3.5269,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9806867697919828e-05,
+      "loss": 3.5368,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9798498129213037e-05,
+      "loss": 3.5183,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9790112181702517e-05,
+      "loss": 3.5197,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9781726234191997e-05,
+      "loss": 3.536,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9773340286681477e-05,
+      "loss": 3.5215,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.976497071797469e-05,
+      "loss": 3.5104,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.975658477046417e-05,
+      "loss": 3.5126,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.974819882295365e-05,
+      "loss": 3.5286,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.973981287544313e-05,
+      "loss": 3.5298,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9731443306736338e-05,
+      "loss": 3.5363,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9723057359225818e-05,
+      "loss": 3.5243,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.97146714117153e-05,
+      "loss": 3.5177,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.970628546420478e-05,
+      "loss": 3.5327,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.969791589549799e-05,
+      "loss": 3.5158,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.968952994798747e-05,
+      "loss": 3.524,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.968114400047695e-05,
+      "loss": 3.5162,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.967275805296643e-05,
+      "loss": 3.5238,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9664388484259643e-05,
+      "loss": 3.5105,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9656002536749123e-05,
+      "loss": 3.537,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9647616589238603e-05,
+      "loss": 3.5182,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9639230641728083e-05,
+      "loss": 3.5228,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9630861073021292e-05,
+      "loss": 3.5299,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9622475125510772e-05,
+      "loss": 3.5263,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9614089178000255e-05,
+      "loss": 3.5217,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9605703230489735e-05,
+      "loss": 3.508,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9597333661782944e-05,
+      "loss": 3.511,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9588947714272424e-05,
+      "loss": 3.5294,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9580561766761904e-05,
+      "loss": 3.524,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9572175819251384e-05,
+      "loss": 3.5129,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9563789871740864e-05,
+      "loss": 3.5147,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9555420303034077e-05,
+      "loss": 3.5208,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9547034355523557e-05,
+      "loss": 3.5108,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9538648408013037e-05,
+      "loss": 3.5303,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9530262460502516e-05,
+      "loss": 3.5154,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9521892891795726e-05,
+      "loss": 3.521,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.951350694428521e-05,
+      "loss": 3.5288,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.950512099677469e-05,
+      "loss": 3.5165,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.949673504926417e-05,
+      "loss": 3.5074,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9488365480557378e-05,
+      "loss": 3.5153,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9479979533046858e-05,
+      "loss": 3.5082,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9471593585536338e-05,
+      "loss": 3.5101,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9463207638025818e-05,
+      "loss": 3.5241,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.945483806931903e-05,
+      "loss": 3.527,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.944645212180851e-05,
+      "loss": 3.5194,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.943806617429799e-05,
+      "loss": 3.5099,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.942968022678747e-05,
+      "loss": 3.5104,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.942129427927695e-05,
+      "loss": 3.5194,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9412924710570163e-05,
+      "loss": 3.5208,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9404538763059643e-05,
+      "loss": 3.5218,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9396152815549122e-05,
+      "loss": 3.523,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9387766868038602e-05,
+      "loss": 3.5266,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.937939729933181e-05,
+      "loss": 3.5268,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.937101135182129e-05,
+      "loss": 3.508,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.936262540431077e-05,
+      "loss": 3.5216,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9354239456800255e-05,
+      "loss": 3.522,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9345869888093464e-05,
+      "loss": 3.5153,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9337483940582944e-05,
+      "loss": 3.5298,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9329097993072424e-05,
+      "loss": 3.5251,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9320712045561904e-05,
+      "loss": 3.5332,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9312342476855116e-05,
+      "loss": 3.5224,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9303956529344596e-05,
+      "loss": 3.5056,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9295570581834076e-05,
+      "loss": 3.5142,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9287184634323556e-05,
+      "loss": 3.5206,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9278798686813036e-05,
+      "loss": 3.5263,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9270429118106245e-05,
+      "loss": 3.5219,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9262043170595725e-05,
+      "loss": 3.5191,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.925365722308521e-05,
+      "loss": 3.5238,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.924527127557469e-05,
+      "loss": 3.5174,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9236901706867898e-05,
+      "loss": 3.5139,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9228515759357377e-05,
+      "loss": 3.5194,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9220129811846857e-05,
+      "loss": 3.523,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9211743864336334e-05,
+      "loss": 3.5344,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.920337429562955e-05,
+      "loss": 3.5141,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.919498834811903e-05,
+      "loss": 3.5096,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.918660240060851e-05,
+      "loss": 3.5178,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9178216453097983e-05,
+      "loss": 3.5223,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.91698468843912e-05,
+      "loss": 3.5254,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.916146093688068e-05,
+      "loss": 3.514,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9153074989370162e-05,
+      "loss": 3.5191,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9144689041859635e-05,
+      "loss": 3.5013,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9136303094349115e-05,
+      "loss": 3.5214,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.912793352564233e-05,
+      "loss": 3.513,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9119547578131808e-05,
+      "loss": 3.5117,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9111161630621288e-05,
+      "loss": 3.5188,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9102775683110768e-05,
+      "loss": 3.5195,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9094406114403983e-05,
+      "loss": 3.511,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9086020166893457e-05,
+      "loss": 3.5193,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9077634219382937e-05,
+      "loss": 3.5116,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9069248271872417e-05,
+      "loss": 3.5164,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9060878703165632e-05,
+      "loss": 3.5132,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.905249275565511e-05,
+      "loss": 3.5184,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.904410680814459e-05,
+      "loss": 3.518,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9035737239437805e-05,
+      "loss": 3.5243,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9027351291927278e-05,
+      "loss": 3.5185,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.901896534441676e-05,
+      "loss": 3.5193,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.901057939690624e-05,
+      "loss": 3.5205,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9002209828199457e-05,
+      "loss": 3.5172,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.899382388068893e-05,
+      "loss": 3.5049,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.898543793317841e-05,
+      "loss": 3.5249,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.897705198566789e-05,
+      "loss": 3.5176,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8968682416961106e-05,
+      "loss": 3.5245,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8960296469450583e-05,
+      "loss": 3.5114,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8951910521940063e-05,
+      "loss": 3.5123,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8943524574429543e-05,
+      "loss": 3.507,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8935155005722752e-05,
+      "loss": 3.514,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.892676905821223e-05,
+      "loss": 3.511,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8918383110701715e-05,
+      "loss": 3.5248,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8909997163191195e-05,
+      "loss": 3.527,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8901627594484404e-05,
+      "loss": 3.5138,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8893241646973884e-05,
+      "loss": 3.5131,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8884855699463364e-05,
+      "loss": 3.5185,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8876469751952844e-05,
+      "loss": 3.5221,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8868100183246056e-05,
+      "loss": 3.5228,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8859714235735536e-05,
+      "loss": 3.5195,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8851328288225016e-05,
+      "loss": 3.5172,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8842942340714496e-05,
+      "loss": 3.5194,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8834556393203976e-05,
+      "loss": 3.5176,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8826186824497185e-05,
+      "loss": 3.514,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8817800876986665e-05,
+      "loss": 3.5249,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.880941492947615e-05,
+      "loss": 3.5165,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.880102898196563e-05,
+      "loss": 3.5065,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8792675792062574e-05,
+      "loss": 3.534,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8784289844552054e-05,
+      "loss": 3.5117,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8775903897041527e-05,
+      "loss": 3.5141,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.876751794953101e-05,
+      "loss": 3.5108,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.875913200202049e-05,
+      "loss": 3.5095,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.846127510070801,
+      "eval_runtime": 307.9881,
+      "eval_samples_per_second": 1238.98,
+      "eval_steps_per_second": 38.719,
+      "step": 1297440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.875074605450997e-05,
+      "loss": 3.5033,
+      "step": 1297920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.874236010699945e-05,
+      "loss": 3.5174,
+      "step": 1298432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.873397415948893e-05,
+      "loss": 3.5296,
+      "step": 1298944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.872558821197841e-05,
+      "loss": 3.51,
+      "step": 1299456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.871721864327162e-05,
+      "loss": 3.5313,
+      "step": 1299968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8708832695761102e-05,
+      "loss": 3.5048,
+      "step": 1300480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8700446748250582e-05,
+      "loss": 3.5172,
+      "step": 1300992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8692060800740062e-05,
+      "loss": 3.5087,
+      "step": 1301504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.868369123203327e-05,
+      "loss": 3.5197,
+      "step": 1302016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.867530528452275e-05,
+      "loss": 3.517,
+      "step": 1302528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.866691933701223e-05,
+      "loss": 3.5185,
+      "step": 1303040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.865853338950171e-05,
+      "loss": 3.5179,
+      "step": 1303552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8650180199598653e-05,
+      "loss": 3.5068,
+      "step": 1304064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8641794252088133e-05,
+      "loss": 3.5015,
+      "step": 1304576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8633408304577613e-05,
+      "loss": 3.5054,
+      "step": 1305088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8625022357067093e-05,
+      "loss": 3.5101,
+      "step": 1305600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8616636409556573e-05,
+      "loss": 3.5166,
+      "step": 1306112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8608250462046056e-05,
+      "loss": 3.51,
+      "step": 1306624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8599864514535536e-05,
+      "loss": 3.5159,
+      "step": 1307136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8591478567025016e-05,
+      "loss": 3.5363,
+      "step": 1307648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8583108998318225e-05,
+      "loss": 3.5135,
+      "step": 1308160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8574723050807705e-05,
+      "loss": 3.5076,
+      "step": 1308672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8566337103297185e-05,
+      "loss": 3.5136,
+      "step": 1309184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8557951155786665e-05,
+      "loss": 3.5217,
+      "step": 1309696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8549581587079877e-05,
+      "loss": 3.5062,
+      "step": 1310208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8541195639569357e-05,
+      "loss": 3.5112,
+      "step": 1310720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8532809692058837e-05,
+      "loss": 3.5189,
+      "step": 1311232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8524423744548317e-05,
+      "loss": 3.5099,
+      "step": 1311744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8516054175841526e-05,
+      "loss": 3.4931,
+      "step": 1312256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.850766822833101e-05,
+      "loss": 3.5043,
+      "step": 1312768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.849928228082049e-05,
+      "loss": 3.5119,
+      "step": 1313280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.849089633330997e-05,
+      "loss": 3.5181,
+      "step": 1313792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.848252676460318e-05,
+      "loss": 3.5248,
+      "step": 1314304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.847414081709266e-05,
+      "loss": 3.51,
+      "step": 1314816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.846575486958214e-05,
+      "loss": 3.5036,
+      "step": 1315328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.845736892207162e-05,
+      "loss": 3.5205,
+      "step": 1315840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.844899935336483e-05,
+      "loss": 3.5019,
+      "step": 1316352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.844061340585431e-05,
+      "loss": 3.512,
+      "step": 1316864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.843222745834379e-05,
+      "loss": 3.5012,
+      "step": 1317376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.842384151083327e-05,
+      "loss": 3.5102,
+      "step": 1317888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.841545556332275e-05,
+      "loss": 3.5003,
+      "step": 1318400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8407085994615963e-05,
+      "loss": 3.5256,
+      "step": 1318912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8398700047105443e-05,
+      "loss": 3.5081,
+      "step": 1319424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8390314099594923e-05,
+      "loss": 3.5091,
+      "step": 1319936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8381928152084403e-05,
+      "loss": 3.5133,
+      "step": 1320448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8373558583377612e-05,
+      "loss": 3.5147,
+      "step": 1320960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8365172635867092e-05,
+      "loss": 3.5069,
+      "step": 1321472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8356786688356572e-05,
+      "loss": 3.4972,
+      "step": 1321984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8348400740846056e-05,
+      "loss": 3.4956,
+      "step": 1322496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8340031172139265e-05,
+      "loss": 3.5164,
+      "step": 1323008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8331645224628745e-05,
+      "loss": 3.5066,
+      "step": 1323520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8323259277118225e-05,
+      "loss": 3.499,
+      "step": 1324032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8314873329607704e-05,
+      "loss": 3.5017,
+      "step": 1324544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8306503760900917e-05,
+      "loss": 3.5092,
+      "step": 1325056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8298117813390397e-05,
+      "loss": 3.4977,
+      "step": 1325568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8289731865879877e-05,
+      "loss": 3.5182,
+      "step": 1326080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8281345918369357e-05,
+      "loss": 3.5033,
+      "step": 1326592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8272976349662566e-05,
+      "loss": 3.5059,
+      "step": 1327104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8264590402152046e-05,
+      "loss": 3.5179,
+      "step": 1327616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8256204454641526e-05,
+      "loss": 3.4972,
+      "step": 1328128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.824781850713101e-05,
+      "loss": 3.4971,
+      "step": 1328640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8239448938424218e-05,
+      "loss": 3.5051,
+      "step": 1329152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8231062990913698e-05,
+      "loss": 3.4928,
+      "step": 1329664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8222677043403178e-05,
+      "loss": 3.4973,
+      "step": 1330176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8214291095892658e-05,
+      "loss": 3.5105,
+      "step": 1330688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.820592152718587e-05,
+      "loss": 3.5139,
+      "step": 1331200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.819753557967535e-05,
+      "loss": 3.5033,
+      "step": 1331712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.818914963216483e-05,
+      "loss": 3.4993,
+      "step": 1332224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8180763684654304e-05,
+      "loss": 3.4915,
+      "step": 1332736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8172377737143784e-05,
+      "loss": 3.5058,
+      "step": 1333248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8164008168437e-05,
+      "loss": 3.5072,
+      "step": 1333760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.815562222092648e-05,
+      "loss": 3.5072,
+      "step": 1334272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8147236273415956e-05,
+      "loss": 3.5116,
+      "step": 1334784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8138850325905436e-05,
+      "loss": 3.5117,
+      "step": 1335296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8130480757198652e-05,
+      "loss": 3.5148,
+      "step": 1335808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8122094809688132e-05,
+      "loss": 3.4952,
+      "step": 1336320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.811370886217761e-05,
+      "loss": 3.507,
+      "step": 1336832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.810532291466709e-05,
+      "loss": 3.5099,
+      "step": 1337344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8096953345960304e-05,
+      "loss": 3.4974,
+      "step": 1337856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8088567398449777e-05,
+      "loss": 3.5219,
+      "step": 1338368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8080181450939257e-05,
+      "loss": 3.5109,
+      "step": 1338880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8071795503428737e-05,
+      "loss": 3.5183,
+      "step": 1339392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8063425934721953e-05,
+      "loss": 3.5123,
+      "step": 1339904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.805503998721143e-05,
+      "loss": 3.4945,
+      "step": 1340416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.804665403970091e-05,
+      "loss": 3.5054,
+      "step": 1340928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.803826809219039e-05,
+      "loss": 3.5067,
+      "step": 1341440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.80298985234836e-05,
+      "loss": 3.5112,
+      "step": 1341952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.802151257597308e-05,
+      "loss": 3.5098,
+      "step": 1342464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8013126628462562e-05,
+      "loss": 3.5032,
+      "step": 1342976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8004740680952042e-05,
+      "loss": 3.5125,
+      "step": 1343488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.799637111224525e-05,
+      "loss": 3.5074,
+      "step": 1344000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.798798516473473e-05,
+      "loss": 3.4959,
+      "step": 1344512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.797959921722421e-05,
+      "loss": 3.504,
+      "step": 1345024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.797121326971369e-05,
+      "loss": 3.5115,
+      "step": 1345536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7962843701006904e-05,
+      "loss": 3.5203,
+      "step": 1346048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7954457753496383e-05,
+      "loss": 3.5038,
+      "step": 1346560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7946071805985863e-05,
+      "loss": 3.4919,
+      "step": 1347072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7937685858475343e-05,
+      "loss": 3.5083,
+      "step": 1347584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7929316289768552e-05,
+      "loss": 3.5095,
+      "step": 1348096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7920930342258032e-05,
+      "loss": 3.5098,
+      "step": 1348608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7912544394747512e-05,
+      "loss": 3.5033,
+      "step": 1349120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7904158447236996e-05,
+      "loss": 3.5009,
+      "step": 1349632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7895788878530205e-05,
+      "loss": 3.491,
+      "step": 1350144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7887402931019685e-05,
+      "loss": 3.5056,
+      "step": 1350656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7879016983509165e-05,
+      "loss": 3.4973,
+      "step": 1351168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7870631035998645e-05,
+      "loss": 3.5014,
+      "step": 1351680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7862261467291857e-05,
+      "loss": 3.5032,
+      "step": 1352192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7853875519781337e-05,
+      "loss": 3.5072,
+      "step": 1352704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7845489572270817e-05,
+      "loss": 3.4972,
+      "step": 1353216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7837103624760297e-05,
+      "loss": 3.5079,
+      "step": 1353728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7828734056053506e-05,
+      "loss": 3.5018,
+      "step": 1354240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7820348108542986e-05,
+      "loss": 3.4999,
+      "step": 1354752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7811962161032466e-05,
+      "loss": 3.5003,
+      "step": 1355264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.780357621352195e-05,
+      "loss": 3.5015,
+      "step": 1355776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.779520664481516e-05,
+      "loss": 3.5098,
+      "step": 1356288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.778682069730464e-05,
+      "loss": 3.5122,
+      "step": 1356800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.777843474979412e-05,
+      "loss": 3.5056,
+      "step": 1357312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.77700488022836e-05,
+      "loss": 3.5036,
+      "step": 1357824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.776167923357681e-05,
+      "loss": 3.5074,
+      "step": 1358336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.775329328606629e-05,
+      "loss": 3.5057,
+      "step": 1358848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.774490733855577e-05,
+      "loss": 3.4926,
+      "step": 1359360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.773652139104525e-05,
+      "loss": 3.511,
+      "step": 1359872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.772815182233846e-05,
+      "loss": 3.5063,
+      "step": 1360384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.771976587482794e-05,
+      "loss": 3.5118,
+      "step": 1360896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.771137992731742e-05,
+      "loss": 3.496,
+      "step": 1361408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7702993979806903e-05,
+      "loss": 3.5002,
+      "step": 1361920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7694624411100112e-05,
+      "loss": 3.4968,
+      "step": 1362432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7686238463589592e-05,
+      "loss": 3.4977,
+      "step": 1362944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7677852516079072e-05,
+      "loss": 3.4969,
+      "step": 1363456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7669466568568552e-05,
+      "loss": 3.5108,
+      "step": 1363968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7661096999861765e-05,
+      "loss": 3.516,
+      "step": 1364480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7652711052351244e-05,
+      "loss": 3.5034,
+      "step": 1364992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7644325104840724e-05,
+      "loss": 3.4952,
+      "step": 1365504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7635939157330204e-05,
+      "loss": 3.5064,
+      "step": 1366016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7627569588623413e-05,
+      "loss": 3.5082,
+      "step": 1366528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7619183641112893e-05,
+      "loss": 3.5107,
+      "step": 1367040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7610797693602373e-05,
+      "loss": 3.5045,
+      "step": 1367552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7602411746091857e-05,
+      "loss": 3.5067,
+      "step": 1368064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7594042177385066e-05,
+      "loss": 3.4984,
+      "step": 1368576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7585656229874546e-05,
+      "loss": 3.509,
+      "step": 1369088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7577270282364026e-05,
+      "loss": 3.5014,
+      "step": 1369600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7568884334853506e-05,
+      "loss": 3.5064,
+      "step": 1370112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7560514766146718e-05,
+      "loss": 3.5079,
+      "step": 1370624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7552128818636198e-05,
+      "loss": 3.4982,
+      "step": 1371136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7543759249929407e-05,
+      "loss": 3.5183,
+      "step": 1371648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7535373302418887e-05,
+      "loss": 3.5011,
+      "step": 1372160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7526987354908367e-05,
+      "loss": 3.4997,
+      "step": 1372672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7518601407397847e-05,
+      "loss": 3.4982,
+      "step": 1373184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7510215459887327e-05,
+      "loss": 3.4965,
+      "step": 1373696
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.848931074142456,
+      "eval_runtime": 304.4942,
+      "eval_samples_per_second": 1253.196,
+      "eval_steps_per_second": 39.163,
+      "step": 1373760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.750182951237681e-05,
+      "loss": 3.4913,
+      "step": 1374208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.749344356486629e-05,
+      "loss": 3.5047,
+      "step": 1374720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.748505761735577e-05,
+      "loss": 3.5151,
+      "step": 1375232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.747668804864898e-05,
+      "loss": 3.5009,
+      "step": 1375744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.746830210113846e-05,
+      "loss": 3.5188,
+      "step": 1376256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.745991615362794e-05,
+      "loss": 3.4949,
+      "step": 1376768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.745153020611742e-05,
+      "loss": 3.5066,
+      "step": 1377280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7443160637410632e-05,
+      "loss": 3.4898,
+      "step": 1377792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7434774689900112e-05,
+      "loss": 3.5081,
+      "step": 1378304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.742638874238959e-05,
+      "loss": 3.5033,
+      "step": 1378816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.741800279487907e-05,
+      "loss": 3.5022,
+      "step": 1379328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.740963322617228e-05,
+      "loss": 3.5122,
+      "step": 1379840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7401263657465493e-05,
+      "loss": 3.4935,
+      "step": 1380352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7392877709954973e-05,
+      "loss": 3.4917,
+      "step": 1380864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7384491762444453e-05,
+      "loss": 3.4902,
+      "step": 1381376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7376105814933933e-05,
+      "loss": 3.4997,
+      "step": 1381888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7367719867423413e-05,
+      "loss": 3.5055,
+      "step": 1382400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7359333919912893e-05,
+      "loss": 3.4972,
+      "step": 1382912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7350947972402373e-05,
+      "loss": 3.5004,
+      "step": 1383424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7342562024891853e-05,
+      "loss": 3.5229,
+      "step": 1383936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7334192456185065e-05,
+      "loss": 3.4997,
+      "step": 1384448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7325806508674545e-05,
+      "loss": 3.4949,
+      "step": 1384960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7317420561164025e-05,
+      "loss": 3.5028,
+      "step": 1385472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7309034613653505e-05,
+      "loss": 3.5071,
+      "step": 1385984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7300665044946718e-05,
+      "loss": 3.4939,
+      "step": 1386496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7292279097436198e-05,
+      "loss": 3.4998,
+      "step": 1387008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7283893149925678e-05,
+      "loss": 3.5028,
+      "step": 1387520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.727550720241515e-05,
+      "loss": 3.4966,
+      "step": 1388032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7267137633708367e-05,
+      "loss": 3.4825,
+      "step": 1388544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7258751686197847e-05,
+      "loss": 3.4913,
+      "step": 1389056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7250365738687327e-05,
+      "loss": 3.501,
+      "step": 1389568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7241979791176803e-05,
+      "loss": 3.5005,
+      "step": 1390080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.723361022247002e-05,
+      "loss": 3.5135,
+      "step": 1390592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.72252242749595e-05,
+      "loss": 3.4949,
+      "step": 1391104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.721683832744898e-05,
+      "loss": 3.4913,
+      "step": 1391616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7208452379938456e-05,
+      "loss": 3.5068,
+      "step": 1392128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7200082811231668e-05,
+      "loss": 3.4936,
+      "step": 1392640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.719169686372115e-05,
+      "loss": 3.4939,
+      "step": 1393152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7183310916210625e-05,
+      "loss": 3.4889,
+      "step": 1393664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7174924968700104e-05,
+      "loss": 3.4981,
+      "step": 1394176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.716655539999332e-05,
+      "loss": 3.4858,
+      "step": 1394688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.71581694524828e-05,
+      "loss": 3.5139,
+      "step": 1395200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7149783504972277e-05,
+      "loss": 3.4903,
+      "step": 1395712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7141397557461757e-05,
+      "loss": 3.5003,
+      "step": 1396224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7133027988754973e-05,
+      "loss": 3.5033,
+      "step": 1396736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7124642041244453e-05,
+      "loss": 3.497,
+      "step": 1397248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7116256093733926e-05,
+      "loss": 3.4964,
+      "step": 1397760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7107870146223406e-05,
+      "loss": 3.4836,
+      "step": 1398272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.709950057751662e-05,
+      "loss": 3.4835,
+      "step": 1398784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7091114630006098e-05,
+      "loss": 3.5031,
+      "step": 1399296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7082728682495578e-05,
+      "loss": 3.4964,
+      "step": 1399808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7074342734985058e-05,
+      "loss": 3.4857,
+      "step": 1400320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7065973166278274e-05,
+      "loss": 3.4924,
+      "step": 1400832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.705758721876775e-05,
+      "loss": 3.4968,
+      "step": 1401344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.704920127125723e-05,
+      "loss": 3.4795,
+      "step": 1401856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.704081532374671e-05,
+      "loss": 3.5027,
+      "step": 1402368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7032445755039926e-05,
+      "loss": 3.4911,
+      "step": 1402880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.70240598075294e-05,
+      "loss": 3.4918,
+      "step": 1403392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.701567386001888e-05,
+      "loss": 3.5077,
+      "step": 1403904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.700728791250836e-05,
+      "loss": 3.4859,
+      "step": 1404416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6998918343801572e-05,
+      "loss": 3.4842,
+      "step": 1404928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6990532396291052e-05,
+      "loss": 3.4933,
+      "step": 1405440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6982146448780532e-05,
+      "loss": 3.4811,
+      "step": 1405952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6973760501270012e-05,
+      "loss": 3.4844,
+      "step": 1406464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.696539093256322e-05,
+      "loss": 3.4947,
+      "step": 1406976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6957004985052704e-05,
+      "loss": 3.5001,
+      "step": 1407488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6948619037542184e-05,
+      "loss": 3.4886,
+      "step": 1408000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6940233090031664e-05,
+      "loss": 3.4883,
+      "step": 1408512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6931863521324873e-05,
+      "loss": 3.4789,
+      "step": 1409024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6923477573814353e-05,
+      "loss": 3.4942,
+      "step": 1409536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6915091626303833e-05,
+      "loss": 3.4922,
+      "step": 1410048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6906705678793313e-05,
+      "loss": 3.4976,
+      "step": 1410560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6898336110086526e-05,
+      "loss": 3.4967,
+      "step": 1411072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6889950162576006e-05,
+      "loss": 3.4976,
+      "step": 1411584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6881564215065486e-05,
+      "loss": 3.5068,
+      "step": 1412096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6873178267554965e-05,
+      "loss": 3.4801,
+      "step": 1412608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6864808698848175e-05,
+      "loss": 3.4924,
+      "step": 1413120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6856422751337658e-05,
+      "loss": 3.4966,
+      "step": 1413632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6848036803827138e-05,
+      "loss": 3.4845,
+      "step": 1414144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6839650856316618e-05,
+      "loss": 3.5066,
+      "step": 1414656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6831281287609827e-05,
+      "loss": 3.4994,
+      "step": 1415168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6822895340099307e-05,
+      "loss": 3.503,
+      "step": 1415680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6814509392588787e-05,
+      "loss": 3.501,
+      "step": 1416192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6806123445078267e-05,
+      "loss": 3.4805,
+      "step": 1416704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.679773749756775e-05,
+      "loss": 3.4951,
+      "step": 1417216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.678936792886096e-05,
+      "loss": 3.4919,
+      "step": 1417728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.678098198135044e-05,
+      "loss": 3.5008,
+      "step": 1418240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.677259603383992e-05,
+      "loss": 3.4948,
+      "step": 1418752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.67642100863294e-05,
+      "loss": 3.4916,
+      "step": 1419264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.675584051762261e-05,
+      "loss": 3.4987,
+      "step": 1419776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.674745457011209e-05,
+      "loss": 3.4973,
+      "step": 1420288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.673906862260157e-05,
+      "loss": 3.4838,
+      "step": 1420800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.673068267509105e-05,
+      "loss": 3.4905,
+      "step": 1421312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.672231310638426e-05,
+      "loss": 3.4981,
+      "step": 1421824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.671392715887374e-05,
+      "loss": 3.5071,
+      "step": 1422336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.670554121136322e-05,
+      "loss": 3.4938,
+      "step": 1422848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6697155263852704e-05,
+      "loss": 3.4789,
+      "step": 1423360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6688785695145913e-05,
+      "loss": 3.4945,
+      "step": 1423872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6680399747635393e-05,
+      "loss": 3.4973,
+      "step": 1424384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6672013800124873e-05,
+      "loss": 3.4955,
+      "step": 1424896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6663627852614353e-05,
+      "loss": 3.4928,
+      "step": 1425408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6655258283907565e-05,
+      "loss": 3.4912,
+      "step": 1425920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6646872336397045e-05,
+      "loss": 3.4783,
+      "step": 1426432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6638486388886525e-05,
+      "loss": 3.4974,
+      "step": 1426944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6630100441376005e-05,
+      "loss": 3.4801,
+      "step": 1427456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6621730872669214e-05,
+      "loss": 3.4927,
+      "step": 1427968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6613344925158694e-05,
+      "loss": 3.4848,
+      "step": 1428480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6604958977648174e-05,
+      "loss": 3.5001,
+      "step": 1428992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6596573030137657e-05,
+      "loss": 3.4823,
+      "step": 1429504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6588203461430867e-05,
+      "loss": 3.4985,
+      "step": 1430016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6579817513920347e-05,
+      "loss": 3.4887,
+      "step": 1430528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6571431566409826e-05,
+      "loss": 3.4842,
+      "step": 1431040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6563045618899306e-05,
+      "loss": 3.4938,
+      "step": 1431552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.655467605019252e-05,
+      "loss": 3.4863,
+      "step": 1432064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6546290102682e-05,
+      "loss": 3.4999,
+      "step": 1432576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.653790415517148e-05,
+      "loss": 3.4985,
+      "step": 1433088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.652951820766096e-05,
+      "loss": 3.494,
+      "step": 1433600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6521148638954168e-05,
+      "loss": 3.492,
+      "step": 1434112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6512762691443648e-05,
+      "loss": 3.4925,
+      "step": 1434624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6504376743933128e-05,
+      "loss": 3.4964,
+      "step": 1435136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.649599079642261e-05,
+      "loss": 3.478,
+      "step": 1435648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.648762122771582e-05,
+      "loss": 3.4982,
+      "step": 1436160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.64792352802053e-05,
+      "loss": 3.4946,
+      "step": 1436672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.647084933269478e-05,
+      "loss": 3.4971,
+      "step": 1437184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.646246338518426e-05,
+      "loss": 3.4862,
+      "step": 1437696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6454093816477473e-05,
+      "loss": 3.4916,
+      "step": 1438208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6445707868966953e-05,
+      "loss": 3.4798,
+      "step": 1438720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6437321921456433e-05,
+      "loss": 3.4858,
+      "step": 1439232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6428935973945912e-05,
+      "loss": 3.4853,
+      "step": 1439744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.642056640523912e-05,
+      "loss": 3.4966,
+      "step": 1440256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.64121804577286e-05,
+      "loss": 3.5043,
+      "step": 1440768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.640379451021808e-05,
+      "loss": 3.4915,
+      "step": 1441280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.639540856270756e-05,
+      "loss": 3.4852,
+      "step": 1441792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6387038994000774e-05,
+      "loss": 3.492,
+      "step": 1442304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6378653046490254e-05,
+      "loss": 3.494,
+      "step": 1442816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6370267098979734e-05,
+      "loss": 3.4972,
+      "step": 1443328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6361881151469214e-05,
+      "loss": 3.4942,
+      "step": 1443840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6353511582762426e-05,
+      "loss": 3.4901,
+      "step": 1444352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6345125635251906e-05,
+      "loss": 3.4922,
+      "step": 1444864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6336739687741386e-05,
+      "loss": 3.4986,
+      "step": 1445376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6328353740230866e-05,
+      "loss": 3.4863,
+      "step": 1445888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6319984171524075e-05,
+      "loss": 3.495,
+      "step": 1446400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6311598224013555e-05,
+      "loss": 3.4948,
+      "step": 1446912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6303212276503035e-05,
+      "loss": 3.4836,
+      "step": 1447424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6294842707796248e-05,
+      "loss": 3.5036,
+      "step": 1447936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6286456760285728e-05,
+      "loss": 3.493,
+      "step": 1448448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6278070812775208e-05,
+      "loss": 3.4859,
+      "step": 1448960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6269684865264687e-05,
+      "loss": 3.4846,
+      "step": 1449472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6261298917754167e-05,
+      "loss": 3.4797,
+      "step": 1449984
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.849992036819458,
+      "eval_runtime": 304.149,
+      "eval_samples_per_second": 1254.619,
+      "eval_steps_per_second": 39.208,
+      "step": 1450080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6252912970243647e-05,
+      "loss": 3.4753,
+      "step": 1450496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6244527022733124e-05,
+      "loss": 3.4886,
+      "step": 1451008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6236141075222604e-05,
+      "loss": 3.5009,
+      "step": 1451520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6227755127712084e-05,
+      "loss": 3.4898,
+      "step": 1452032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.62193855590053e-05,
+      "loss": 3.5088,
+      "step": 1452544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6210999611494773e-05,
+      "loss": 3.4793,
+      "step": 1453056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6202613663984253e-05,
+      "loss": 3.4949,
+      "step": 1453568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6194227716473736e-05,
+      "loss": 3.4752,
+      "step": 1454080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6185858147766952e-05,
+      "loss": 3.495,
+      "step": 1454592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6177472200256425e-05,
+      "loss": 3.492,
+      "step": 1455104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6169086252745905e-05,
+      "loss": 3.4888,
+      "step": 1455616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6160700305235385e-05,
+      "loss": 3.4971,
+      "step": 1456128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.615234711533233e-05,
+      "loss": 3.4825,
+      "step": 1456640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6143961167821814e-05,
+      "loss": 3.4809,
+      "step": 1457152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6135575220311294e-05,
+      "loss": 3.4792,
+      "step": 1457664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6127189272800773e-05,
+      "loss": 3.4861,
+      "step": 1458176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6118803325290247e-05,
+      "loss": 3.4921,
+      "step": 1458688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6110417377779727e-05,
+      "loss": 3.4846,
+      "step": 1459200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6102031430269207e-05,
+      "loss": 3.4893,
+      "step": 1459712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.609364548275869e-05,
+      "loss": 3.5075,
+      "step": 1460224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.60852759140519e-05,
+      "loss": 3.4906,
+      "step": 1460736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.607688996654138e-05,
+      "loss": 3.4883,
+      "step": 1461248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.606850401903086e-05,
+      "loss": 3.4893,
+      "step": 1461760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.606011807152034e-05,
+      "loss": 3.4928,
+      "step": 1462272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.605174850281355e-05,
+      "loss": 3.4802,
+      "step": 1462784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.604336255530303e-05,
+      "loss": 3.4897,
+      "step": 1463296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.603497660779251e-05,
+      "loss": 3.4908,
+      "step": 1463808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.602659066028199e-05,
+      "loss": 3.4816,
+      "step": 1464320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.60182210915752e-05,
+      "loss": 3.4753,
+      "step": 1464832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.600983514406468e-05,
+      "loss": 3.4768,
+      "step": 1465344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.600144919655416e-05,
+      "loss": 3.4902,
+      "step": 1465856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5993079627847373e-05,
+      "loss": 3.4875,
+      "step": 1466368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5984693680336853e-05,
+      "loss": 3.501,
+      "step": 1466880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5976307732826333e-05,
+      "loss": 3.4892,
+      "step": 1467392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5967921785315813e-05,
+      "loss": 3.4748,
+      "step": 1467904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.595955221660902e-05,
+      "loss": 3.4903,
+      "step": 1468416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5951166269098505e-05,
+      "loss": 3.4863,
+      "step": 1468928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5942780321587985e-05,
+      "loss": 3.4788,
+      "step": 1469440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5934394374077465e-05,
+      "loss": 3.4784,
+      "step": 1469952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5926008426566945e-05,
+      "loss": 3.482,
+      "step": 1470464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5917638857860154e-05,
+      "loss": 3.4753,
+      "step": 1470976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5909252910349634e-05,
+      "loss": 3.5,
+      "step": 1471488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5900866962839114e-05,
+      "loss": 3.4825,
+      "step": 1472000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5892481015328597e-05,
+      "loss": 3.4866,
+      "step": 1472512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5884111446621806e-05,
+      "loss": 3.4913,
+      "step": 1473024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5875725499111286e-05,
+      "loss": 3.4814,
+      "step": 1473536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5867339551600766e-05,
+      "loss": 3.4878,
+      "step": 1474048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5858953604090246e-05,
+      "loss": 3.471,
+      "step": 1474560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.585058403538346e-05,
+      "loss": 3.4652,
+      "step": 1475072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.584219808787294e-05,
+      "loss": 3.4944,
+      "step": 1475584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.583381214036242e-05,
+      "loss": 3.4837,
+      "step": 1476096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.58254261928519e-05,
+      "loss": 3.478,
+      "step": 1476608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5817056624145108e-05,
+      "loss": 3.4775,
+      "step": 1477120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5808670676634588e-05,
+      "loss": 3.482,
+      "step": 1477632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5800284729124068e-05,
+      "loss": 3.4673,
+      "step": 1478144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.579189878161355e-05,
+      "loss": 3.4888,
+      "step": 1478656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.578352921290676e-05,
+      "loss": 3.485,
+      "step": 1479168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.577514326539624e-05,
+      "loss": 3.4798,
+      "step": 1479680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.576675731788572e-05,
+      "loss": 3.4921,
+      "step": 1480192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.57583713703752e-05,
+      "loss": 3.4736,
+      "step": 1480704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5750001801668412e-05,
+      "loss": 3.4732,
+      "step": 1481216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5741615854157892e-05,
+      "loss": 3.4869,
+      "step": 1481728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5733229906647372e-05,
+      "loss": 3.4647,
+      "step": 1482240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5724843959136852e-05,
+      "loss": 3.4696,
+      "step": 1482752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.571647439043006e-05,
+      "loss": 3.4876,
+      "step": 1483264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.570808844291954e-05,
+      "loss": 3.4865,
+      "step": 1483776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.569970249540902e-05,
+      "loss": 3.4724,
+      "step": 1484288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5691316547898505e-05,
+      "loss": 3.4821,
+      "step": 1484800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5682946979191714e-05,
+      "loss": 3.4639,
+      "step": 1485312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5674561031681194e-05,
+      "loss": 3.4826,
+      "step": 1485824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5666175084170674e-05,
+      "loss": 3.4799,
+      "step": 1486336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5657789136660154e-05,
+      "loss": 3.4854,
+      "step": 1486848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5649419567953366e-05,
+      "loss": 3.4831,
+      "step": 1487360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5641033620442846e-05,
+      "loss": 3.487,
+      "step": 1487872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5632647672932326e-05,
+      "loss": 3.4965,
+      "step": 1488384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5624261725421806e-05,
+      "loss": 3.4678,
+      "step": 1488896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5615892156715015e-05,
+      "loss": 3.4856,
+      "step": 1489408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5607506209204495e-05,
+      "loss": 3.4868,
+      "step": 1489920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5599120261693975e-05,
+      "loss": 3.4672,
+      "step": 1490432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5590734314183458e-05,
+      "loss": 3.4972,
+      "step": 1490944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5582364745476667e-05,
+      "loss": 3.4865,
+      "step": 1491456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5573978797966147e-05,
+      "loss": 3.4879,
+      "step": 1491968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5565592850455627e-05,
+      "loss": 3.4905,
+      "step": 1492480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5557206902945107e-05,
+      "loss": 3.4726,
+      "step": 1492992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.554883733423832e-05,
+      "loss": 3.4749,
+      "step": 1493504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.55404513867278e-05,
+      "loss": 3.4818,
+      "step": 1494016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.553206543921728e-05,
+      "loss": 3.489,
+      "step": 1494528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.552367949170676e-05,
+      "loss": 3.4829,
+      "step": 1495040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.551530992299997e-05,
+      "loss": 3.4817,
+      "step": 1495552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.550692397548945e-05,
+      "loss": 3.4827,
+      "step": 1496064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.549853802797893e-05,
+      "loss": 3.4903,
+      "step": 1496576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.549015208046841e-05,
+      "loss": 3.4697,
+      "step": 1497088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.548178251176162e-05,
+      "loss": 3.4788,
+      "step": 1497600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.54733965642511e-05,
+      "loss": 3.4859,
+      "step": 1498112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.546501061674058e-05,
+      "loss": 3.4938,
+      "step": 1498624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.545662466923006e-05,
+      "loss": 3.4824,
+      "step": 1499136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5448255100523273e-05,
+      "loss": 3.4689,
+      "step": 1499648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5439869153012753e-05,
+      "loss": 3.4801,
+      "step": 1500160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5431483205502233e-05,
+      "loss": 3.483,
+      "step": 1500672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5423097257991713e-05,
+      "loss": 3.484,
+      "step": 1501184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5414727689284922e-05,
+      "loss": 3.4822,
+      "step": 1501696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5406341741774402e-05,
+      "loss": 3.4772,
+      "step": 1502208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5397955794263882e-05,
+      "loss": 3.4675,
+      "step": 1502720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5389569846753362e-05,
+      "loss": 3.4811,
+      "step": 1503232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5381200278046575e-05,
+      "loss": 3.4679,
+      "step": 1503744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5372814330536055e-05,
+      "loss": 3.4782,
+      "step": 1504256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5364428383025535e-05,
+      "loss": 3.4708,
+      "step": 1504768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5356042435515015e-05,
+      "loss": 3.4868,
+      "step": 1505280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5347672866808224e-05,
+      "loss": 3.4691,
+      "step": 1505792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5339286919297707e-05,
+      "loss": 3.4886,
+      "step": 1506304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5330900971787187e-05,
+      "loss": 3.4723,
+      "step": 1506816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5322515024276667e-05,
+      "loss": 3.4769,
+      "step": 1507328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5314145455569876e-05,
+      "loss": 3.4788,
+      "step": 1507840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5305759508059356e-05,
+      "loss": 3.4777,
+      "step": 1508352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5297373560548836e-05,
+      "loss": 3.4818,
+      "step": 1508864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5288987613038316e-05,
+      "loss": 3.4901,
+      "step": 1509376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.52806016655278e-05,
+      "loss": 3.4818,
+      "step": 1509888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5272232096821008e-05,
+      "loss": 3.4778,
+      "step": 1510400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5263846149310488e-05,
+      "loss": 3.4807,
+      "step": 1510912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5255460201799968e-05,
+      "loss": 3.484,
+      "step": 1511424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5247074254289445e-05,
+      "loss": 3.4665,
+      "step": 1511936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.523870468558266e-05,
+      "loss": 3.486,
+      "step": 1512448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.523031873807214e-05,
+      "loss": 3.4782,
+      "step": 1512960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.522193279056162e-05,
+      "loss": 3.4876,
+      "step": 1513472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5213546843051094e-05,
+      "loss": 3.475,
+      "step": 1513984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.520517727434431e-05,
+      "loss": 3.4802,
+      "step": 1514496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.519679132683379e-05,
+      "loss": 3.467,
+      "step": 1515008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.518840537932327e-05,
+      "loss": 3.4774,
+      "step": 1515520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5180019431812746e-05,
+      "loss": 3.4756,
+      "step": 1516032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5171649863105962e-05,
+      "loss": 3.4803,
+      "step": 1516544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5163263915595442e-05,
+      "loss": 3.4881,
+      "step": 1517056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5154877968084915e-05,
+      "loss": 3.4824,
+      "step": 1517568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.51464920205744e-05,
+      "loss": 3.4724,
+      "step": 1518080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5138122451867614e-05,
+      "loss": 3.4805,
+      "step": 1518592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5129736504357094e-05,
+      "loss": 3.4856,
+      "step": 1519104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5121350556846567e-05,
+      "loss": 3.4846,
+      "step": 1519616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5112964609336047e-05,
+      "loss": 3.4824,
+      "step": 1520128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5104595040629263e-05,
+      "loss": 3.478,
+      "step": 1520640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5096209093118743e-05,
+      "loss": 3.4781,
+      "step": 1521152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.508782314560822e-05,
+      "loss": 3.4895,
+      "step": 1521664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.50794371980977e-05,
+      "loss": 3.4729,
+      "step": 1522176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.507105125058718e-05,
+      "loss": 3.4853,
+      "step": 1522688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.506268168188039e-05,
+      "loss": 3.4781,
+      "step": 1523200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.505429573436987e-05,
+      "loss": 3.4777,
+      "step": 1523712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5045926165663085e-05,
+      "loss": 3.4924,
+      "step": 1524224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5037540218152568e-05,
+      "loss": 3.4805,
+      "step": 1524736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.502915427064204e-05,
+      "loss": 3.4726,
+      "step": 1525248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.502076832313152e-05,
+      "loss": 3.4712,
+      "step": 1525760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5012382375621e-05,
+      "loss": 3.4741,
+      "step": 1526272
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8496294021606445,
+      "eval_runtime": 302.8437,
+      "eval_samples_per_second": 1260.026,
+      "eval_steps_per_second": 39.377,
+      "step": 1526400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.500399642811048e-05,
+      "loss": 3.4686,
+      "step": 1526784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.499561048059996e-05,
+      "loss": 3.4777,
+      "step": 1527296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4987224533089444e-05,
+      "loss": 3.4832,
+      "step": 1527808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4978854964382657e-05,
+      "loss": 3.4792,
+      "step": 1528320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4970469016872133e-05,
+      "loss": 3.4957,
+      "step": 1528832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4962083069361613e-05,
+      "loss": 3.4746,
+      "step": 1529344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4953697121851093e-05,
+      "loss": 3.4784,
+      "step": 1529856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4945327553144306e-05,
+      "loss": 3.4678,
+      "step": 1530368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4936941605633786e-05,
+      "loss": 3.4792,
+      "step": 1530880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4928555658123266e-05,
+      "loss": 3.4795,
+      "step": 1531392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4920169710612746e-05,
+      "loss": 3.4792,
+      "step": 1531904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4911800141905955e-05,
+      "loss": 3.4858,
+      "step": 1532416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4903430573199167e-05,
+      "loss": 3.467,
+      "step": 1532928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4895044625688647e-05,
+      "loss": 3.4692,
+      "step": 1533440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.488665867817813e-05,
+      "loss": 3.4695,
+      "step": 1533952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4878272730667607e-05,
+      "loss": 3.4724,
+      "step": 1534464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4869886783157087e-05,
+      "loss": 3.4791,
+      "step": 1534976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4861500835646567e-05,
+      "loss": 3.4737,
+      "step": 1535488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4853114888136047e-05,
+      "loss": 3.4791,
+      "step": 1536000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4844728940625527e-05,
+      "loss": 3.4975,
+      "step": 1536512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.483635937191874e-05,
+      "loss": 3.4789,
+      "step": 1537024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.482797342440822e-05,
+      "loss": 3.4741,
+      "step": 1537536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.48195874768977e-05,
+      "loss": 3.4769,
+      "step": 1538048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.481120152938718e-05,
+      "loss": 3.4785,
+      "step": 1538560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.480283196068039e-05,
+      "loss": 3.469,
+      "step": 1539072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.479444601316987e-05,
+      "loss": 3.4787,
+      "step": 1539584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.478606006565935e-05,
+      "loss": 3.4756,
+      "step": 1540096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.477767411814883e-05,
+      "loss": 3.4734,
+      "step": 1540608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.476930454944204e-05,
+      "loss": 3.4635,
+      "step": 1541120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.476091860193152e-05,
+      "loss": 3.4672,
+      "step": 1541632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4752532654421e-05,
+      "loss": 3.4748,
+      "step": 1542144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.474414670691048e-05,
+      "loss": 3.4768,
+      "step": 1542656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4735777138203693e-05,
+      "loss": 3.4871,
+      "step": 1543168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4727391190693173e-05,
+      "loss": 3.4761,
+      "step": 1543680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4719005243182653e-05,
+      "loss": 3.4625,
+      "step": 1544192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4710619295672133e-05,
+      "loss": 3.4803,
+      "step": 1544704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4702249726965342e-05,
+      "loss": 3.4785,
+      "step": 1545216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4693863779454822e-05,
+      "loss": 3.4646,
+      "step": 1545728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4685477831944305e-05,
+      "loss": 3.4647,
+      "step": 1546240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4677091884433785e-05,
+      "loss": 3.4733,
+      "step": 1546752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4668722315726994e-05,
+      "loss": 3.462,
+      "step": 1547264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4660336368216474e-05,
+      "loss": 3.4881,
+      "step": 1547776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4651950420705954e-05,
+      "loss": 3.4707,
+      "step": 1548288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.464356447319543e-05,
+      "loss": 3.4718,
+      "step": 1548800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4635194904488647e-05,
+      "loss": 3.4804,
+      "step": 1549312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4626808956978127e-05,
+      "loss": 3.4748,
+      "step": 1549824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4618423009467607e-05,
+      "loss": 3.4752,
+      "step": 1550336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4610037061957083e-05,
+      "loss": 3.4615,
+      "step": 1550848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4601667493250296e-05,
+      "loss": 3.4488,
+      "step": 1551360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4593281545739776e-05,
+      "loss": 3.4888,
+      "step": 1551872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4584895598229256e-05,
+      "loss": 3.4684,
+      "step": 1552384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4576509650718736e-05,
+      "loss": 3.4705,
+      "step": 1552896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4568140082011948e-05,
+      "loss": 3.4683,
+      "step": 1553408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4559754134501428e-05,
+      "loss": 3.4659,
+      "step": 1553920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4551368186990905e-05,
+      "loss": 3.4589,
+      "step": 1554432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4542982239480385e-05,
+      "loss": 3.4745,
+      "step": 1554944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.45346126707736e-05,
+      "loss": 3.4712,
+      "step": 1555456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.452622672326308e-05,
+      "loss": 3.4676,
+      "step": 1555968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4517840775752557e-05,
+      "loss": 3.4852,
+      "step": 1556480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4509454828242037e-05,
+      "loss": 3.459,
+      "step": 1556992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.450108525953525e-05,
+      "loss": 3.4576,
+      "step": 1557504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.449269931202473e-05,
+      "loss": 3.4797,
+      "step": 1558016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.448431336451421e-05,
+      "loss": 3.4494,
+      "step": 1558528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4475943795807422e-05,
+      "loss": 3.4546,
+      "step": 1559040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4467557848296902e-05,
+      "loss": 3.4781,
+      "step": 1559552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4459171900786378e-05,
+      "loss": 3.4754,
+      "step": 1560064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4450785953275858e-05,
+      "loss": 3.4616,
+      "step": 1560576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4442400005765338e-05,
+      "loss": 3.4704,
+      "step": 1561088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4434030437058554e-05,
+      "loss": 3.4534,
+      "step": 1561600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.442564448954803e-05,
+      "loss": 3.4696,
+      "step": 1562112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.441725854203751e-05,
+      "loss": 3.4666,
+      "step": 1562624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.440887259452699e-05,
+      "loss": 3.4716,
+      "step": 1563136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4400503025820203e-05,
+      "loss": 3.4694,
+      "step": 1563648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4392117078309683e-05,
+      "loss": 3.4755,
+      "step": 1564160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4383731130799163e-05,
+      "loss": 3.4827,
+      "step": 1564672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4375345183288643e-05,
+      "loss": 3.4609,
+      "step": 1565184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4366975614581852e-05,
+      "loss": 3.4693,
+      "step": 1565696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4358589667071332e-05,
+      "loss": 3.4747,
+      "step": 1566208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4350203719560812e-05,
+      "loss": 3.4537,
+      "step": 1566720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4341817772050292e-05,
+      "loss": 3.4834,
+      "step": 1567232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4333448203343504e-05,
+      "loss": 3.4796,
+      "step": 1567744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4325062255832984e-05,
+      "loss": 3.4762,
+      "step": 1568256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4316676308322464e-05,
+      "loss": 3.4795,
+      "step": 1568768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4308290360811944e-05,
+      "loss": 3.4629,
+      "step": 1569280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4299920792105153e-05,
+      "loss": 3.4615,
+      "step": 1569792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4291534844594637e-05,
+      "loss": 3.47,
+      "step": 1570304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4283148897084117e-05,
+      "loss": 3.4772,
+      "step": 1570816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4274762949573597e-05,
+      "loss": 3.4723,
+      "step": 1571328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4266393380866806e-05,
+      "loss": 3.4701,
+      "step": 1571840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4258007433356286e-05,
+      "loss": 3.4689,
+      "step": 1572352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4249621485845766e-05,
+      "loss": 3.479,
+      "step": 1572864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4241235538335246e-05,
+      "loss": 3.463,
+      "step": 1573376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4232865969628458e-05,
+      "loss": 3.466,
+      "step": 1573888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4224480022117938e-05,
+      "loss": 3.4697,
+      "step": 1574400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4216094074607418e-05,
+      "loss": 3.483,
+      "step": 1574912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4207708127096898e-05,
+      "loss": 3.4707,
+      "step": 1575424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4199338558390107e-05,
+      "loss": 3.4593,
+      "step": 1575936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.419095261087959e-05,
+      "loss": 3.4669,
+      "step": 1576448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.418256666336907e-05,
+      "loss": 3.4689,
+      "step": 1576960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.417418071585855e-05,
+      "loss": 3.475,
+      "step": 1577472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.416581114715176e-05,
+      "loss": 3.4679,
+      "step": 1577984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.415742519964124e-05,
+      "loss": 3.4691,
+      "step": 1578496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.414903925213072e-05,
+      "loss": 3.4567,
+      "step": 1579008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.41406533046202e-05,
+      "loss": 3.4665,
+      "step": 1579520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.413228373591341e-05,
+      "loss": 3.4592,
+      "step": 1580032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.412389778840289e-05,
+      "loss": 3.4705,
+      "step": 1580544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.411551184089237e-05,
+      "loss": 3.4531,
+      "step": 1581056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.410712589338185e-05,
+      "loss": 3.4823,
+      "step": 1581568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.409875632467506e-05,
+      "loss": 3.4581,
+      "step": 1582080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.409037037716454e-05,
+      "loss": 3.4707,
+      "step": 1582592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4081984429654024e-05,
+      "loss": 3.4634,
+      "step": 1583104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4073598482143504e-05,
+      "loss": 3.4639,
+      "step": 1583616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4065228913436713e-05,
+      "loss": 3.4668,
+      "step": 1584128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4056842965926193e-05,
+      "loss": 3.4671,
+      "step": 1584640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4048457018415673e-05,
+      "loss": 3.4711,
+      "step": 1585152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4040071070905153e-05,
+      "loss": 3.4781,
+      "step": 1585664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4031701502198365e-05,
+      "loss": 3.4691,
+      "step": 1586176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4023315554687845e-05,
+      "loss": 3.4667,
+      "step": 1586688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4014929607177325e-05,
+      "loss": 3.4668,
+      "step": 1587200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4006543659666802e-05,
+      "loss": 3.4754,
+      "step": 1587712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3998174090960014e-05,
+      "loss": 3.4511,
+      "step": 1588224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3989788143449494e-05,
+      "loss": 3.4763,
+      "step": 1588736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3981402195938978e-05,
+      "loss": 3.466,
+      "step": 1589248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3973016248428454e-05,
+      "loss": 3.4763,
+      "step": 1589760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3964646679721667e-05,
+      "loss": 3.4628,
+      "step": 1590272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3956260732211147e-05,
+      "loss": 3.4704,
+      "step": 1590784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3947874784700627e-05,
+      "loss": 3.4597,
+      "step": 1591296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3939488837190107e-05,
+      "loss": 3.4605,
+      "step": 1591808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.393111926848332e-05,
+      "loss": 3.4669,
+      "step": 1592320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.39227333209728e-05,
+      "loss": 3.4648,
+      "step": 1592832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3914347373462276e-05,
+      "loss": 3.4717,
+      "step": 1593344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3905961425951755e-05,
+      "loss": 3.4767,
+      "step": 1593856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3897591857244968e-05,
+      "loss": 3.4592,
+      "step": 1594368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3889205909734448e-05,
+      "loss": 3.4697,
+      "step": 1594880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3880819962223928e-05,
+      "loss": 3.4733,
+      "step": 1595392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3872434014713408e-05,
+      "loss": 3.4699,
+      "step": 1595904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.386406444600662e-05,
+      "loss": 3.4679,
+      "step": 1596416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.38556784984961e-05,
+      "loss": 3.471,
+      "step": 1596928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3847292550985577e-05,
+      "loss": 3.4627,
+      "step": 1597440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.383890660347506e-05,
+      "loss": 3.4786,
+      "step": 1597952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3830537034768273e-05,
+      "loss": 3.4598,
+      "step": 1598464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.382215108725775e-05,
+      "loss": 3.4734,
+      "step": 1598976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.381376513974723e-05,
+      "loss": 3.4666,
+      "step": 1599488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.380537919223671e-05,
+      "loss": 3.4641,
+      "step": 1600000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.379700962352992e-05,
+      "loss": 3.4803,
+      "step": 1600512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.37886236760194e-05,
+      "loss": 3.4683,
+      "step": 1601024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.378023772850888e-05,
+      "loss": 3.4637,
+      "step": 1601536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.377185178099836e-05,
+      "loss": 3.4546,
+      "step": 1602048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.376346583348784e-05,
+      "loss": 3.463,
+      "step": 1602560
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8523378372192383,
+      "eval_runtime": 303.0089,
+      "eval_samples_per_second": 1259.339,
+      "eval_steps_per_second": 39.355,
+      "step": 1602720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.375507988597732e-05,
+      "loss": 3.4553,
+      "step": 1603072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.37466939384668e-05,
+      "loss": 3.4625,
+      "step": 1603584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.373830799095628e-05,
+      "loss": 3.4752,
+      "step": 1604096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.372992204344576e-05,
+      "loss": 3.4697,
+      "step": 1604608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3721552474738974e-05,
+      "loss": 3.4824,
+      "step": 1605120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3713166527228454e-05,
+      "loss": 3.4641,
+      "step": 1605632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.370478057971793e-05,
+      "loss": 3.462,
+      "step": 1606144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.369639463220741e-05,
+      "loss": 3.4606,
+      "step": 1606656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3688025063500623e-05,
+      "loss": 3.4627,
+      "step": 1607168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3679639115990103e-05,
+      "loss": 3.4692,
+      "step": 1607680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3671253168479583e-05,
+      "loss": 3.4692,
+      "step": 1608192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3662867220969063e-05,
+      "loss": 3.4702,
+      "step": 1608704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3654514031066004e-05,
+      "loss": 3.4628,
+      "step": 1609216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3646128083555484e-05,
+      "loss": 3.4576,
+      "step": 1609728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3637742136044964e-05,
+      "loss": 3.4571,
+      "step": 1610240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3629356188534447e-05,
+      "loss": 3.4579,
+      "step": 1610752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3620970241023927e-05,
+      "loss": 3.4661,
+      "step": 1611264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3612584293513404e-05,
+      "loss": 3.4629,
+      "step": 1611776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3604198346002884e-05,
+      "loss": 3.4662,
+      "step": 1612288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3595812398492364e-05,
+      "loss": 3.4829,
+      "step": 1612800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3587442829785576e-05,
+      "loss": 3.4716,
+      "step": 1613312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3579056882275056e-05,
+      "loss": 3.4574,
+      "step": 1613824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3570670934764536e-05,
+      "loss": 3.4654,
+      "step": 1614336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3562284987254016e-05,
+      "loss": 3.4641,
+      "step": 1614848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.355391541854723e-05,
+      "loss": 3.4584,
+      "step": 1615360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3545529471036705e-05,
+      "loss": 3.468,
+      "step": 1615872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3537143523526185e-05,
+      "loss": 3.4625,
+      "step": 1616384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.352875757601567e-05,
+      "loss": 3.4613,
+      "step": 1616896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3520388007308878e-05,
+      "loss": 3.4575,
+      "step": 1617408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3512002059798358e-05,
+      "loss": 3.4519,
+      "step": 1617920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3503616112287838e-05,
+      "loss": 3.4649,
+      "step": 1618432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3495230164777318e-05,
+      "loss": 3.4654,
+      "step": 1618944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.348686059607053e-05,
+      "loss": 3.4735,
+      "step": 1619456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.347847464856001e-05,
+      "loss": 3.466,
+      "step": 1619968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.347008870104949e-05,
+      "loss": 3.4517,
+      "step": 1620480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.346170275353897e-05,
+      "loss": 3.4698,
+      "step": 1620992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.345333318483218e-05,
+      "loss": 3.4675,
+      "step": 1621504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.344494723732166e-05,
+      "loss": 3.4502,
+      "step": 1622016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.343656128981114e-05,
+      "loss": 3.4592,
+      "step": 1622528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3428175342300622e-05,
+      "loss": 3.4572,
+      "step": 1623040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.341980577359383e-05,
+      "loss": 3.4544,
+      "step": 1623552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.341141982608331e-05,
+      "loss": 3.4722,
+      "step": 1624064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.340303387857279e-05,
+      "loss": 3.4606,
+      "step": 1624576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.339464793106227e-05,
+      "loss": 3.4585,
+      "step": 1625088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3386278362355484e-05,
+      "loss": 3.466,
+      "step": 1625600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3377892414844964e-05,
+      "loss": 3.4631,
+      "step": 1626112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3369506467334444e-05,
+      "loss": 3.4665,
+      "step": 1626624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3361120519823924e-05,
+      "loss": 3.4482,
+      "step": 1627136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3352750951117133e-05,
+      "loss": 3.4327,
+      "step": 1627648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3344365003606613e-05,
+      "loss": 3.4768,
+      "step": 1628160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3335979056096093e-05,
+      "loss": 3.4569,
+      "step": 1628672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3327593108585573e-05,
+      "loss": 3.4654,
+      "step": 1629184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3319223539878785e-05,
+      "loss": 3.4525,
+      "step": 1629696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3310837592368265e-05,
+      "loss": 3.4543,
+      "step": 1630208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3302451644857745e-05,
+      "loss": 3.4483,
+      "step": 1630720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3294065697347225e-05,
+      "loss": 3.4572,
+      "step": 1631232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3285696128640434e-05,
+      "loss": 3.4643,
+      "step": 1631744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3277310181129917e-05,
+      "loss": 3.452,
+      "step": 1632256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3268924233619397e-05,
+      "loss": 3.4774,
+      "step": 1632768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3260538286108877e-05,
+      "loss": 3.4526,
+      "step": 1633280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3252168717402086e-05,
+      "loss": 3.4424,
+      "step": 1633792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3243782769891566e-05,
+      "loss": 3.4657,
+      "step": 1634304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3235396822381046e-05,
+      "loss": 3.4406,
+      "step": 1634816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3227010874870526e-05,
+      "loss": 3.4413,
+      "step": 1635328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.321864130616374e-05,
+      "loss": 3.467,
+      "step": 1635840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.321025535865322e-05,
+      "loss": 3.4618,
+      "step": 1636352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.32018694111427e-05,
+      "loss": 3.4545,
+      "step": 1636864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.319348346363218e-05,
+      "loss": 3.4596,
+      "step": 1637376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3185113894925388e-05,
+      "loss": 3.4411,
+      "step": 1637888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.317672794741487e-05,
+      "loss": 3.4594,
+      "step": 1638400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.316834199990435e-05,
+      "loss": 3.4536,
+      "step": 1638912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3159956052393828e-05,
+      "loss": 3.464,
+      "step": 1639424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.315158648368704e-05,
+      "loss": 3.4599,
+      "step": 1639936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.314320053617652e-05,
+      "loss": 3.463,
+      "step": 1640448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3134814588666e-05,
+      "loss": 3.4688,
+      "step": 1640960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.312642864115548e-05,
+      "loss": 3.4473,
+      "step": 1641472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3118059072448692e-05,
+      "loss": 3.4608,
+      "step": 1641984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3109673124938172e-05,
+      "loss": 3.461,
+      "step": 1642496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3101287177427652e-05,
+      "loss": 3.4447,
+      "step": 1643008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.309290122991713e-05,
+      "loss": 3.471,
+      "step": 1643520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.308453166121034e-05,
+      "loss": 3.4684,
+      "step": 1644032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3076145713699825e-05,
+      "loss": 3.4616,
+      "step": 1644544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.30677597661893e-05,
+      "loss": 3.4718,
+      "step": 1645056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.305937381867878e-05,
+      "loss": 3.4505,
+      "step": 1645568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3051004249971994e-05,
+      "loss": 3.4504,
+      "step": 1646080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3042618302461474e-05,
+      "loss": 3.4605,
+      "step": 1646592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3034232354950954e-05,
+      "loss": 3.4627,
+      "step": 1647104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3025846407440434e-05,
+      "loss": 3.4665,
+      "step": 1647616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3017476838733646e-05,
+      "loss": 3.4566,
+      "step": 1648128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3009090891223126e-05,
+      "loss": 3.4559,
+      "step": 1648640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3000704943712603e-05,
+      "loss": 3.4669,
+      "step": 1649152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2992318996202083e-05,
+      "loss": 3.4533,
+      "step": 1649664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2983949427495295e-05,
+      "loss": 3.4536,
+      "step": 1650176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2975563479984775e-05,
+      "loss": 3.4589,
+      "step": 1650688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2967177532474255e-05,
+      "loss": 3.4704,
+      "step": 1651200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2958791584963735e-05,
+      "loss": 3.4607,
+      "step": 1651712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2950422016256947e-05,
+      "loss": 3.4466,
+      "step": 1652224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2942036068746424e-05,
+      "loss": 3.4569,
+      "step": 1652736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2933650121235907e-05,
+      "loss": 3.4608,
+      "step": 1653248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2925264173725387e-05,
+      "loss": 3.4633,
+      "step": 1653760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.29168946050186e-05,
+      "loss": 3.458,
+      "step": 1654272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2908508657508076e-05,
+      "loss": 3.4563,
+      "step": 1654784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2900122709997556e-05,
+      "loss": 3.4433,
+      "step": 1655296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2891736762487036e-05,
+      "loss": 3.4526,
+      "step": 1655808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.288336719378025e-05,
+      "loss": 3.4536,
+      "step": 1656320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.287498124626973e-05,
+      "loss": 3.4578,
+      "step": 1656832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.286659529875921e-05,
+      "loss": 3.4401,
+      "step": 1657344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.285820935124869e-05,
+      "loss": 3.4718,
+      "step": 1657856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2849839782541898e-05,
+      "loss": 3.4484,
+      "step": 1658368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2841453835031378e-05,
+      "loss": 3.4581,
+      "step": 1658880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2833067887520858e-05,
+      "loss": 3.4479,
+      "step": 1659392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.282468194001034e-05,
+      "loss": 3.4577,
+      "step": 1659904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.281631237130355e-05,
+      "loss": 3.4567,
+      "step": 1660416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.280792642379303e-05,
+      "loss": 3.4523,
+      "step": 1660928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.279954047628251e-05,
+      "loss": 3.4592,
+      "step": 1661440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.279115452877199e-05,
+      "loss": 3.4699,
+      "step": 1661952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2782784960065202e-05,
+      "loss": 3.4541,
+      "step": 1662464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2774399012554682e-05,
+      "loss": 3.4623,
+      "step": 1662976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2766013065044162e-05,
+      "loss": 3.4512,
+      "step": 1663488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2757627117533642e-05,
+      "loss": 3.4646,
+      "step": 1664000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.274925754882685e-05,
+      "loss": 3.4451,
+      "step": 1664512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.274087160131633e-05,
+      "loss": 3.4626,
+      "step": 1665024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.273248565380581e-05,
+      "loss": 3.4531,
+      "step": 1665536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2724099706295295e-05,
+      "loss": 3.4627,
+      "step": 1666048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2715730137588504e-05,
+      "loss": 3.4548,
+      "step": 1666560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2707344190077984e-05,
+      "loss": 3.4583,
+      "step": 1667072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2698958242567464e-05,
+      "loss": 3.4445,
+      "step": 1667584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2690572295056944e-05,
+      "loss": 3.4523,
+      "step": 1668096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2682202726350156e-05,
+      "loss": 3.4548,
+      "step": 1668608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2673816778839636e-05,
+      "loss": 3.4581,
+      "step": 1669120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2665430831329116e-05,
+      "loss": 3.4547,
+      "step": 1669632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2657044883818596e-05,
+      "loss": 3.4653,
+      "step": 1670144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2648675315111805e-05,
+      "loss": 3.4499,
+      "step": 1670656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2640289367601285e-05,
+      "loss": 3.4585,
+      "step": 1671168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2631903420090765e-05,
+      "loss": 3.4583,
+      "step": 1671680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2623517472580248e-05,
+      "loss": 3.4645,
+      "step": 1672192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2615147903873457e-05,
+      "loss": 3.4539,
+      "step": 1672704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2606761956362937e-05,
+      "loss": 3.4599,
+      "step": 1673216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2598376008852417e-05,
+      "loss": 3.4503,
+      "step": 1673728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2589990061341897e-05,
+      "loss": 3.4716,
+      "step": 1674240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.258162049263511e-05,
+      "loss": 3.4484,
+      "step": 1674752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.257323454512459e-05,
+      "loss": 3.4617,
+      "step": 1675264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.256484859761407e-05,
+      "loss": 3.454,
+      "step": 1675776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.255646265010355e-05,
+      "loss": 3.4557,
+      "step": 1676288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.254809308139676e-05,
+      "loss": 3.4641,
+      "step": 1676800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.253970713388624e-05,
+      "loss": 3.4569,
+      "step": 1677312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.253132118637572e-05,
+      "loss": 3.4569,
+      "step": 1677824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.25229352388652e-05,
+      "loss": 3.4393,
+      "step": 1678336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.251454929135468e-05,
+      "loss": 3.456,
+      "step": 1678848
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.85417103767395,
+      "eval_runtime": 304.0282,
+      "eval_samples_per_second": 1255.117,
+      "eval_steps_per_second": 39.223,
+      "step": 1679040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.250616334384416e-05,
+      "loss": 3.4485,
+      "step": 1679360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.249777739633364e-05,
+      "loss": 3.4485,
+      "step": 1679872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.248939144882312e-05,
+      "loss": 3.4597,
+      "step": 1680384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.248102188011633e-05,
+      "loss": 3.4604,
+      "step": 1680896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.247263593260581e-05,
+      "loss": 3.4688,
+      "step": 1681408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.246424998509529e-05,
+      "loss": 3.4523,
+      "step": 1681920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.245586403758477e-05,
+      "loss": 3.4501,
+      "step": 1682432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.244749446887798e-05,
+      "loss": 3.4491,
+      "step": 1682944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.243910852136746e-05,
+      "loss": 3.4529,
+      "step": 1683456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.243072257385694e-05,
+      "loss": 3.4596,
+      "step": 1683968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.242233662634642e-05,
+      "loss": 3.4551,
+      "step": 1684480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2413967057639632e-05,
+      "loss": 3.458,
+      "step": 1684992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2405597488932845e-05,
+      "loss": 3.451,
+      "step": 1685504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.239721154142232e-05,
+      "loss": 3.4462,
+      "step": 1686016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.23888255939118e-05,
+      "loss": 3.4493,
+      "step": 1686528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.238043964640128e-05,
+      "loss": 3.4445,
+      "step": 1687040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2372053698890764e-05,
+      "loss": 3.4584,
+      "step": 1687552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2363667751380244e-05,
+      "loss": 3.4544,
+      "step": 1688064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2355281803869724e-05,
+      "loss": 3.454,
+      "step": 1688576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2346895856359204e-05,
+      "loss": 3.4688,
+      "step": 1689088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2338526287652413e-05,
+      "loss": 3.4661,
+      "step": 1689600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2330140340141893e-05,
+      "loss": 3.4491,
+      "step": 1690112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2321754392631373e-05,
+      "loss": 3.4511,
+      "step": 1690624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2313368445120853e-05,
+      "loss": 3.4536,
+      "step": 1691136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2304998876414066e-05,
+      "loss": 3.4527,
+      "step": 1691648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2296612928903546e-05,
+      "loss": 3.4541,
+      "step": 1692160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2288226981393026e-05,
+      "loss": 3.4531,
+      "step": 1692672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2279841033882502e-05,
+      "loss": 3.4548,
+      "step": 1693184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2271471465175718e-05,
+      "loss": 3.4426,
+      "step": 1693696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2263085517665198e-05,
+      "loss": 3.4444,
+      "step": 1694208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2254699570154678e-05,
+      "loss": 3.4462,
+      "step": 1694720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2246313622644155e-05,
+      "loss": 3.4536,
+      "step": 1695232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2237944053937367e-05,
+      "loss": 3.4626,
+      "step": 1695744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2229558106426847e-05,
+      "loss": 3.4573,
+      "step": 1696256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2221172158916327e-05,
+      "loss": 3.4422,
+      "step": 1696768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2212786211405807e-05,
+      "loss": 3.4542,
+      "step": 1697280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.220441664269902e-05,
+      "loss": 3.4588,
+      "step": 1697792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.21960306951885e-05,
+      "loss": 3.4412,
+      "step": 1698304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2187644747677976e-05,
+      "loss": 3.4461,
+      "step": 1698816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2179258800167456e-05,
+      "loss": 3.4466,
+      "step": 1699328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2170889231460672e-05,
+      "loss": 3.4402,
+      "step": 1699840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.216250328395015e-05,
+      "loss": 3.4589,
+      "step": 1700352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2154117336439628e-05,
+      "loss": 3.4498,
+      "step": 1700864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.214574776773284e-05,
+      "loss": 3.4519,
+      "step": 1701376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.213736182022232e-05,
+      "loss": 3.4525,
+      "step": 1701888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.21289758727118e-05,
+      "loss": 3.4507,
+      "step": 1702400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.212058992520128e-05,
+      "loss": 3.457,
+      "step": 1702912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2112220356494493e-05,
+      "loss": 3.4416,
+      "step": 1703424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2103834408983973e-05,
+      "loss": 3.4188,
+      "step": 1703936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.209544846147345e-05,
+      "loss": 3.4631,
+      "step": 1704448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.208706251396293e-05,
+      "loss": 3.4478,
+      "step": 1704960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2078692945256142e-05,
+      "loss": 3.4552,
+      "step": 1705472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2070306997745622e-05,
+      "loss": 3.4433,
+      "step": 1705984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2061921050235102e-05,
+      "loss": 3.4425,
+      "step": 1706496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2053535102724582e-05,
+      "loss": 3.4353,
+      "step": 1707008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2045165534017794e-05,
+      "loss": 3.4493,
+      "step": 1707520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.203677958650727e-05,
+      "loss": 3.454,
+      "step": 1708032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2028393638996754e-05,
+      "loss": 3.4387,
+      "step": 1708544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2020007691486234e-05,
+      "loss": 3.4643,
+      "step": 1709056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2011638122779447e-05,
+      "loss": 3.4431,
+      "step": 1709568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2003252175268923e-05,
+      "loss": 3.4349,
+      "step": 1710080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1994866227758403e-05,
+      "loss": 3.4508,
+      "step": 1710592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1986480280247883e-05,
+      "loss": 3.4357,
+      "step": 1711104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1978110711541096e-05,
+      "loss": 3.4287,
+      "step": 1711616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1969724764030576e-05,
+      "loss": 3.4564,
+      "step": 1712128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1961338816520056e-05,
+      "loss": 3.4515,
+      "step": 1712640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1952952869009536e-05,
+      "loss": 3.4441,
+      "step": 1713152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1944583300302745e-05,
+      "loss": 3.446,
+      "step": 1713664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1936197352792225e-05,
+      "loss": 3.4325,
+      "step": 1714176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1927811405281705e-05,
+      "loss": 3.447,
+      "step": 1714688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1919425457771188e-05,
+      "loss": 3.4455,
+      "step": 1715200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1911039510260668e-05,
+      "loss": 3.4474,
+      "step": 1715712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1902669941553877e-05,
+      "loss": 3.4497,
+      "step": 1716224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1894283994043357e-05,
+      "loss": 3.45,
+      "step": 1716736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1885898046532837e-05,
+      "loss": 3.4612,
+      "step": 1717248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1877512099022317e-05,
+      "loss": 3.4365,
+      "step": 1717760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.186914253031553e-05,
+      "loss": 3.4542,
+      "step": 1718272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.186075658280501e-05,
+      "loss": 3.4451,
+      "step": 1718784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.185237063529449e-05,
+      "loss": 3.4358,
+      "step": 1719296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.184398468778397e-05,
+      "loss": 3.4579,
+      "step": 1719808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.183561511907718e-05,
+      "loss": 3.4606,
+      "step": 1720320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.182722917156666e-05,
+      "loss": 3.4493,
+      "step": 1720832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.181884322405614e-05,
+      "loss": 3.4614,
+      "step": 1721344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.181047365534935e-05,
+      "loss": 3.4367,
+      "step": 1721856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.180208770783883e-05,
+      "loss": 3.439,
+      "step": 1722368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.179370176032831e-05,
+      "loss": 3.4486,
+      "step": 1722880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.178531581281779e-05,
+      "loss": 3.455,
+      "step": 1723392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1776946244111003e-05,
+      "loss": 3.4477,
+      "step": 1723904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1768560296600483e-05,
+      "loss": 3.4493,
+      "step": 1724416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1760174349089963e-05,
+      "loss": 3.4483,
+      "step": 1724928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1751788401579443e-05,
+      "loss": 3.4542,
+      "step": 1725440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1743418832872652e-05,
+      "loss": 3.4436,
+      "step": 1725952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1735032885362132e-05,
+      "loss": 3.4433,
+      "step": 1726464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1726646937851612e-05,
+      "loss": 3.4505,
+      "step": 1726976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1718260990341095e-05,
+      "loss": 3.46,
+      "step": 1727488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1709891421634304e-05,
+      "loss": 3.4496,
+      "step": 1728000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1701505474123784e-05,
+      "loss": 3.4338,
+      "step": 1728512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1693119526613264e-05,
+      "loss": 3.4481,
+      "step": 1729024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1684733579102744e-05,
+      "loss": 3.4465,
+      "step": 1729536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1676364010395957e-05,
+      "loss": 3.4526,
+      "step": 1730048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1667978062885437e-05,
+      "loss": 3.4446,
+      "step": 1730560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1659592115374917e-05,
+      "loss": 3.4454,
+      "step": 1731072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1651206167864397e-05,
+      "loss": 3.4379,
+      "step": 1731584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1642836599157606e-05,
+      "loss": 3.4364,
+      "step": 1732096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1634450651647086e-05,
+      "loss": 3.4478,
+      "step": 1732608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1626064704136566e-05,
+      "loss": 3.4455,
+      "step": 1733120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.161767875662605e-05,
+      "loss": 3.4297,
+      "step": 1733632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1609309187919258e-05,
+      "loss": 3.4609,
+      "step": 1734144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1600923240408738e-05,
+      "loss": 3.4375,
+      "step": 1734656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1592537292898218e-05,
+      "loss": 3.444,
+      "step": 1735168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1584151345387695e-05,
+      "loss": 3.4391,
+      "step": 1735680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.157578177668091e-05,
+      "loss": 3.4446,
+      "step": 1736192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.156739582917039e-05,
+      "loss": 3.4475,
+      "step": 1736704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.155900988165987e-05,
+      "loss": 3.4406,
+      "step": 1737216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1550623934149347e-05,
+      "loss": 3.4473,
+      "step": 1737728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.154225436544256e-05,
+      "loss": 3.4539,
+      "step": 1738240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.153386841793204e-05,
+      "loss": 3.4468,
+      "step": 1738752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.152548247042152e-05,
+      "loss": 3.4512,
+      "step": 1739264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1517096522911e-05,
+      "loss": 3.4401,
+      "step": 1739776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1508726954204212e-05,
+      "loss": 3.4524,
+      "step": 1740288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1500341006693692e-05,
+      "loss": 3.4363,
+      "step": 1740800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1491955059183168e-05,
+      "loss": 3.4514,
+      "step": 1741312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1483569111672648e-05,
+      "loss": 3.4428,
+      "step": 1741824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1475199542965864e-05,
+      "loss": 3.4531,
+      "step": 1742336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1466813595455344e-05,
+      "loss": 3.447,
+      "step": 1742848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.145842764794482e-05,
+      "loss": 3.4443,
+      "step": 1743360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.14500417004343e-05,
+      "loss": 3.4359,
+      "step": 1743872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1441672131727513e-05,
+      "loss": 3.4397,
+      "step": 1744384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.143328618421699e-05,
+      "loss": 3.4412,
+      "step": 1744896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1424900236706473e-05,
+      "loss": 3.4501,
+      "step": 1745408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1416514289195953e-05,
+      "loss": 3.4413,
+      "step": 1745920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1408144720489165e-05,
+      "loss": 3.4559,
+      "step": 1746432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1399758772978642e-05,
+      "loss": 3.4363,
+      "step": 1746944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1391372825468122e-05,
+      "loss": 3.4495,
+      "step": 1747456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1382986877957602e-05,
+      "loss": 3.4473,
+      "step": 1747968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1374617309250818e-05,
+      "loss": 3.4523,
+      "step": 1748480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1366231361740294e-05,
+      "loss": 3.4451,
+      "step": 1748992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1357845414229774e-05,
+      "loss": 3.4472,
+      "step": 1749504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1349459466719254e-05,
+      "loss": 3.4411,
+      "step": 1750016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1341089898012463e-05,
+      "loss": 3.4605,
+      "step": 1750528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1332703950501943e-05,
+      "loss": 3.4353,
+      "step": 1751040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1324318002991427e-05,
+      "loss": 3.4513,
+      "step": 1751552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1315932055480907e-05,
+      "loss": 3.4448,
+      "step": 1752064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1307562486774116e-05,
+      "loss": 3.4499,
+      "step": 1752576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1299192918067328e-05,
+      "loss": 3.4498,
+      "step": 1753088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1290806970556808e-05,
+      "loss": 3.4464,
+      "step": 1753600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1282421023046288e-05,
+      "loss": 3.4462,
+      "step": 1754112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1274035075535768e-05,
+      "loss": 3.432,
+      "step": 1754624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1265649128025248e-05,
+      "loss": 3.4458,
+      "step": 1755136
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8550140857696533,
+      "eval_runtime": 303.8103,
+      "eval_samples_per_second": 1256.017,
+      "eval_steps_per_second": 39.251,
+      "step": 1755360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1257263180514728e-05,
+      "loss": 3.4393,
+      "step": 1755648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1248877233004208e-05,
+      "loss": 3.4363,
+      "step": 1756160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1240491285493688e-05,
+      "loss": 3.4472,
+      "step": 1756672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1232105337983168e-05,
+      "loss": 3.4513,
+      "step": 1757184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.122373576927638e-05,
+      "loss": 3.4569,
+      "step": 1757696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.121534982176586e-05,
+      "loss": 3.4451,
+      "step": 1758208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.120696387425534e-05,
+      "loss": 3.4393,
+      "step": 1758720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.119857792674482e-05,
+      "loss": 3.441,
+      "step": 1759232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.119020835803803e-05,
+      "loss": 3.4421,
+      "step": 1759744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.118182241052751e-05,
+      "loss": 3.4476,
+      "step": 1760256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.117343646301699e-05,
+      "loss": 3.4421,
+      "step": 1760768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1165050515506473e-05,
+      "loss": 3.4454,
+      "step": 1761280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.115669732560341e-05,
+      "loss": 3.4452,
+      "step": 1761792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.114831137809289e-05,
+      "loss": 3.4348,
+      "step": 1762304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.113992543058237e-05,
+      "loss": 3.4338,
+      "step": 1762816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.113153948307185e-05,
+      "loss": 3.4325,
+      "step": 1763328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1123153535561334e-05,
+      "loss": 3.4481,
+      "step": 1763840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1114767588050814e-05,
+      "loss": 3.4457,
+      "step": 1764352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1106381640540294e-05,
+      "loss": 3.4385,
+      "step": 1764864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.109799569302977e-05,
+      "loss": 3.4588,
+      "step": 1765376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1089626124322983e-05,
+      "loss": 3.453,
+      "step": 1765888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1081240176812463e-05,
+      "loss": 3.4418,
+      "step": 1766400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1072854229301943e-05,
+      "loss": 3.4425,
+      "step": 1766912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1064468281791423e-05,
+      "loss": 3.437,
+      "step": 1767424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1056098713084635e-05,
+      "loss": 3.4438,
+      "step": 1767936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1047712765574115e-05,
+      "loss": 3.4422,
+      "step": 1768448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1039326818063592e-05,
+      "loss": 3.4421,
+      "step": 1768960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1030940870553072e-05,
+      "loss": 3.444,
+      "step": 1769472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1022571301846288e-05,
+      "loss": 3.4321,
+      "step": 1769984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1014185354335768e-05,
+      "loss": 3.4318,
+      "step": 1770496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1005799406825244e-05,
+      "loss": 3.4312,
+      "step": 1771008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0997413459314724e-05,
+      "loss": 3.4456,
+      "step": 1771520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0989043890607937e-05,
+      "loss": 3.4503,
+      "step": 1772032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0980657943097417e-05,
+      "loss": 3.447,
+      "step": 1772544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0972271995586897e-05,
+      "loss": 3.437,
+      "step": 1773056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0963886048076376e-05,
+      "loss": 3.4354,
+      "step": 1773568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.095551647936959e-05,
+      "loss": 3.449,
+      "step": 1774080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0947130531859066e-05,
+      "loss": 3.4372,
+      "step": 1774592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0938744584348545e-05,
+      "loss": 3.4316,
+      "step": 1775104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0930358636838025e-05,
+      "loss": 3.4352,
+      "step": 1775616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.092198906813124e-05,
+      "loss": 3.4335,
+      "step": 1776128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0913603120620718e-05,
+      "loss": 3.4442,
+      "step": 1776640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0905217173110198e-05,
+      "loss": 3.4427,
+      "step": 1777152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0896831225599678e-05,
+      "loss": 3.4408,
+      "step": 1777664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0888461656892887e-05,
+      "loss": 3.4388,
+      "step": 1778176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0880075709382367e-05,
+      "loss": 3.4433,
+      "step": 1778688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.087168976187185e-05,
+      "loss": 3.445,
+      "step": 1779200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.086330381436133e-05,
+      "loss": 3.4308,
+      "step": 1779712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.085493424565454e-05,
+      "loss": 3.4073,
+      "step": 1780224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.084654829814402e-05,
+      "loss": 3.4516,
+      "step": 1780736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.08381623506335e-05,
+      "loss": 3.4363,
+      "step": 1781248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.082977640312298e-05,
+      "loss": 3.4433,
+      "step": 1781760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.082140683441619e-05,
+      "loss": 3.4343,
+      "step": 1782272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.081302088690567e-05,
+      "loss": 3.4351,
+      "step": 1782784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.080463493939515e-05,
+      "loss": 3.4196,
+      "step": 1783296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.079624899188463e-05,
+      "loss": 3.4366,
+      "step": 1783808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.078787942317784e-05,
+      "loss": 3.4472,
+      "step": 1784320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.077949347566732e-05,
+      "loss": 3.4264,
+      "step": 1784832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0771107528156804e-05,
+      "loss": 3.4535,
+      "step": 1785344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0762721580646284e-05,
+      "loss": 3.4349,
+      "step": 1785856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0754352011939493e-05,
+      "loss": 3.4228,
+      "step": 1786368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0745966064428973e-05,
+      "loss": 3.4366,
+      "step": 1786880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0737580116918453e-05,
+      "loss": 3.4264,
+      "step": 1787392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0729194169407933e-05,
+      "loss": 3.4217,
+      "step": 1787904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0720824600701145e-05,
+      "loss": 3.4425,
+      "step": 1788416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0712438653190625e-05,
+      "loss": 3.4402,
+      "step": 1788928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0704052705680105e-05,
+      "loss": 3.4317,
+      "step": 1789440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0695666758169585e-05,
+      "loss": 3.4391,
+      "step": 1789952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0687297189462794e-05,
+      "loss": 3.4183,
+      "step": 1790464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0678911241952274e-05,
+      "loss": 3.4386,
+      "step": 1790976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0670525294441758e-05,
+      "loss": 3.4295,
+      "step": 1791488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0662139346931237e-05,
+      "loss": 3.4355,
+      "step": 1792000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0653769778224447e-05,
+      "loss": 3.4443,
+      "step": 1792512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0645383830713927e-05,
+      "loss": 3.4349,
+      "step": 1793024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0636997883203406e-05,
+      "loss": 3.4542,
+      "step": 1793536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0628611935692886e-05,
+      "loss": 3.4254,
+      "step": 1794048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.06202423669861e-05,
+      "loss": 3.4413,
+      "step": 1794560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.061185641947558e-05,
+      "loss": 3.4346,
+      "step": 1795072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.060347047196506e-05,
+      "loss": 3.4279,
+      "step": 1795584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.059508452445454e-05,
+      "loss": 3.4447,
+      "step": 1796096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0586714955747748e-05,
+      "loss": 3.4481,
+      "step": 1796608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0578329008237228e-05,
+      "loss": 3.4408,
+      "step": 1797120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.056994306072671e-05,
+      "loss": 3.4513,
+      "step": 1797632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.056155711321619e-05,
+      "loss": 3.4272,
+      "step": 1798144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.05531875445094e-05,
+      "loss": 3.4258,
+      "step": 1798656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.054480159699888e-05,
+      "loss": 3.4364,
+      "step": 1799168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.053641564948836e-05,
+      "loss": 3.4476,
+      "step": 1799680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.052802970197784e-05,
+      "loss": 3.435,
+      "step": 1800192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0519660133271053e-05,
+      "loss": 3.4408,
+      "step": 1800704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0511274185760533e-05,
+      "loss": 3.4348,
+      "step": 1801216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0502888238250013e-05,
+      "loss": 3.441,
+      "step": 1801728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.049450229073949e-05,
+      "loss": 3.4352,
+      "step": 1802240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.04861327220327e-05,
+      "loss": 3.4316,
+      "step": 1802752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.047774677452218e-05,
+      "loss": 3.4372,
+      "step": 1803264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0469360827011665e-05,
+      "loss": 3.4504,
+      "step": 1803776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.046097487950114e-05,
+      "loss": 3.4401,
+      "step": 1804288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0452605310794354e-05,
+      "loss": 3.42,
+      "step": 1804800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0444219363283834e-05,
+      "loss": 3.4342,
+      "step": 1805312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0435833415773314e-05,
+      "loss": 3.4352,
+      "step": 1805824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.042744746826279e-05,
+      "loss": 3.4433,
+      "step": 1806336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0419077899556006e-05,
+      "loss": 3.4374,
+      "step": 1806848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0410691952045486e-05,
+      "loss": 3.4316,
+      "step": 1807360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0402306004534963e-05,
+      "loss": 3.4286,
+      "step": 1807872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0393920057024443e-05,
+      "loss": 3.4247,
+      "step": 1808384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0385550488317655e-05,
+      "loss": 3.4369,
+      "step": 1808896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0377164540807135e-05,
+      "loss": 3.4335,
+      "step": 1809408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0368778593296615e-05,
+      "loss": 3.4215,
+      "step": 1809920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0360392645786095e-05,
+      "loss": 3.447,
+      "step": 1810432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0352023077079308e-05,
+      "loss": 3.4319,
+      "step": 1810944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0343637129568788e-05,
+      "loss": 3.4325,
+      "step": 1811456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0335251182058264e-05,
+      "loss": 3.4329,
+      "step": 1811968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0326865234547744e-05,
+      "loss": 3.4281,
+      "step": 1812480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.031849566584096e-05,
+      "loss": 3.4377,
+      "step": 1812992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0310109718330436e-05,
+      "loss": 3.4316,
+      "step": 1813504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0301723770819916e-05,
+      "loss": 3.4402,
+      "step": 1814016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0293337823309396e-05,
+      "loss": 3.4448,
+      "step": 1814528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.028496825460261e-05,
+      "loss": 3.4349,
+      "step": 1815040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.027658230709209e-05,
+      "loss": 3.439,
+      "step": 1815552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.026819635958157e-05,
+      "loss": 3.4296,
+      "step": 1816064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.025981041207105e-05,
+      "loss": 3.4427,
+      "step": 1816576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0251440843364258e-05,
+      "loss": 3.4289,
+      "step": 1817088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0243054895853738e-05,
+      "loss": 3.4324,
+      "step": 1817600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0234668948343218e-05,
+      "loss": 3.433,
+      "step": 1818112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0226283000832698e-05,
+      "loss": 3.4422,
+      "step": 1818624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.021791343212591e-05,
+      "loss": 3.4385,
+      "step": 1819136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.020952748461539e-05,
+      "loss": 3.4322,
+      "step": 1819648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.020114153710487e-05,
+      "loss": 3.4213,
+      "step": 1820160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.019275558959435e-05,
+      "loss": 3.4348,
+      "step": 1820672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.018438602088756e-05,
+      "loss": 3.4304,
+      "step": 1821184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0176000073377043e-05,
+      "loss": 3.4365,
+      "step": 1821696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0167614125866522e-05,
+      "loss": 3.4328,
+      "step": 1822208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0159228178356002e-05,
+      "loss": 3.4445,
+      "step": 1822720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.015085860964921e-05,
+      "loss": 3.4247,
+      "step": 1823232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.014247266213869e-05,
+      "loss": 3.4441,
+      "step": 1823744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.013408671462817e-05,
+      "loss": 3.4348,
+      "step": 1824256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.012570076711765e-05,
+      "loss": 3.4414,
+      "step": 1824768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0117331198410864e-05,
+      "loss": 3.4361,
+      "step": 1825280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0108945250900344e-05,
+      "loss": 3.4325,
+      "step": 1825792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0100559303389824e-05,
+      "loss": 3.432,
+      "step": 1826304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0092173355879304e-05,
+      "loss": 3.4495,
+      "step": 1826816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0083803787172513e-05,
+      "loss": 3.4247,
+      "step": 1827328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0075417839661996e-05,
+      "loss": 3.4387,
+      "step": 1827840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0067031892151476e-05,
+      "loss": 3.4361,
+      "step": 1828352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0058645944640956e-05,
+      "loss": 3.4379,
+      "step": 1828864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0050292754737898e-05,
+      "loss": 3.4359,
+      "step": 1829376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0041906807227378e-05,
+      "loss": 3.4364,
+      "step": 1829888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0033520859716858e-05,
+      "loss": 3.4346,
+      "step": 1830400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0025134912206338e-05,
+      "loss": 3.4234,
+      "step": 1830912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0016748964695818e-05,
+      "loss": 3.433,
+      "step": 1831424
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8559415340423584,
+      "eval_runtime": 304.312,
+      "eval_samples_per_second": 1253.947,
+      "eval_steps_per_second": 39.187,
+      "step": 1831680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0008363017185297e-05,
+      "loss": 3.4246,
+      "step": 1831936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9999977069674777e-05,
+      "loss": 3.4272,
+      "step": 1832448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9991591122164257e-05,
+      "loss": 3.4363,
+      "step": 1832960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9983205174653737e-05,
+      "loss": 3.4383,
+      "step": 1833472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.997483560594695e-05,
+      "loss": 3.4484,
+      "step": 1833984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.996644965843643e-05,
+      "loss": 3.4297,
+      "step": 1834496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.995806371092591e-05,
+      "loss": 3.4328,
+      "step": 1835008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9949677763415386e-05,
+      "loss": 3.4303,
+      "step": 1835520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.99413081947086e-05,
+      "loss": 3.4298,
+      "step": 1836032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.993292224719808e-05,
+      "loss": 3.4396,
+      "step": 1836544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.992453629968756e-05,
+      "loss": 3.4314,
+      "step": 1837056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.991615035217704e-05,
+      "loss": 3.4339,
+      "step": 1837568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.990779716227398e-05,
+      "loss": 3.4411,
+      "step": 1838080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.989941121476346e-05,
+      "loss": 3.4177,
+      "step": 1838592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.989102526725294e-05,
+      "loss": 3.4231,
+      "step": 1839104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.988263931974242e-05,
+      "loss": 3.4211,
+      "step": 1839616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9874253372231904e-05,
+      "loss": 3.439,
+      "step": 1840128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9865867424721383e-05,
+      "loss": 3.4307,
+      "step": 1840640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.985748147721086e-05,
+      "loss": 3.4309,
+      "step": 1841152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.984909552970034e-05,
+      "loss": 3.4464,
+      "step": 1841664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9840725960993552e-05,
+      "loss": 3.443,
+      "step": 1842176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9832340013483032e-05,
+      "loss": 3.4282,
+      "step": 1842688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9823954065972512e-05,
+      "loss": 3.4336,
+      "step": 1843200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9815568118461992e-05,
+      "loss": 3.4272,
+      "step": 1843712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9807198549755205e-05,
+      "loss": 3.4347,
+      "step": 1844224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9798812602244685e-05,
+      "loss": 3.4289,
+      "step": 1844736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.979042665473416e-05,
+      "loss": 3.4291,
+      "step": 1845248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.978204070722364e-05,
+      "loss": 3.4375,
+      "step": 1845760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9773671138516857e-05,
+      "loss": 3.422,
+      "step": 1846272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9765285191006334e-05,
+      "loss": 3.4212,
+      "step": 1846784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9756899243495814e-05,
+      "loss": 3.4244,
+      "step": 1847296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9748513295985294e-05,
+      "loss": 3.4291,
+      "step": 1847808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9740143727278506e-05,
+      "loss": 3.4355,
+      "step": 1848320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9731757779767983e-05,
+      "loss": 3.4421,
+      "step": 1848832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9723371832257466e-05,
+      "loss": 3.4246,
+      "step": 1849344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9714985884746946e-05,
+      "loss": 3.4272,
+      "step": 1849856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.970661631604016e-05,
+      "loss": 3.4386,
+      "step": 1850368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9698230368529635e-05,
+      "loss": 3.4251,
+      "step": 1850880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9689844421019115e-05,
+      "loss": 3.4214,
+      "step": 1851392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9681458473508595e-05,
+      "loss": 3.4278,
+      "step": 1851904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9673088904801807e-05,
+      "loss": 3.4182,
+      "step": 1852416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9664702957291287e-05,
+      "loss": 3.4301,
+      "step": 1852928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9656317009780767e-05,
+      "loss": 3.4327,
+      "step": 1853440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9647931062270247e-05,
+      "loss": 3.4332,
+      "step": 1853952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9639561493563456e-05,
+      "loss": 3.4259,
+      "step": 1854464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9631175546052936e-05,
+      "loss": 3.4364,
+      "step": 1854976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.962278959854242e-05,
+      "loss": 3.4309,
+      "step": 1855488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.96144036510319e-05,
+      "loss": 3.4214,
+      "step": 1856000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.960603408232511e-05,
+      "loss": 3.3985,
+      "step": 1856512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.959764813481459e-05,
+      "loss": 3.4378,
+      "step": 1857024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.958926218730407e-05,
+      "loss": 3.4251,
+      "step": 1857536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.958087623979355e-05,
+      "loss": 3.4319,
+      "step": 1858048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.957250667108676e-05,
+      "loss": 3.4242,
+      "step": 1858560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.956412072357624e-05,
+      "loss": 3.4216,
+      "step": 1859072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.955573477606572e-05,
+      "loss": 3.4139,
+      "step": 1859584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.95473488285552e-05,
+      "loss": 3.4226,
+      "step": 1860096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.953897925984841e-05,
+      "loss": 3.441,
+      "step": 1860608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.953059331233789e-05,
+      "loss": 3.4129,
+      "step": 1861120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9522207364827373e-05,
+      "loss": 3.4443,
+      "step": 1861632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9513821417316853e-05,
+      "loss": 3.4258,
+      "step": 1862144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9505451848610062e-05,
+      "loss": 3.4096,
+      "step": 1862656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9497065901099542e-05,
+      "loss": 3.4257,
+      "step": 1863168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9488679953589022e-05,
+      "loss": 3.4154,
+      "step": 1863680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9480294006078502e-05,
+      "loss": 3.4137,
+      "step": 1864192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9471924437371715e-05,
+      "loss": 3.4269,
+      "step": 1864704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9463538489861195e-05,
+      "loss": 3.4325,
+      "step": 1865216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9455152542350675e-05,
+      "loss": 3.4182,
+      "step": 1865728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9446766594840155e-05,
+      "loss": 3.4316,
+      "step": 1866240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9438397026133364e-05,
+      "loss": 3.4076,
+      "step": 1866752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9430011078622844e-05,
+      "loss": 3.4247,
+      "step": 1867264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9421625131112327e-05,
+      "loss": 3.4232,
+      "step": 1867776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9413239183601807e-05,
+      "loss": 3.4219,
+      "step": 1868288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9404869614895016e-05,
+      "loss": 3.4316,
+      "step": 1868800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9396483667384496e-05,
+      "loss": 3.4252,
+      "step": 1869312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9388097719873976e-05,
+      "loss": 3.4425,
+      "step": 1869824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9379711772363456e-05,
+      "loss": 3.4168,
+      "step": 1870336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.937134220365667e-05,
+      "loss": 3.4316,
+      "step": 1870848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.936295625614615e-05,
+      "loss": 3.421,
+      "step": 1871360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.935457030863563e-05,
+      "loss": 3.417,
+      "step": 1871872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.934618436112511e-05,
+      "loss": 3.4349,
+      "step": 1872384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9337814792418317e-05,
+      "loss": 3.4328,
+      "step": 1872896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9329428844907797e-05,
+      "loss": 3.4341,
+      "step": 1873408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9321042897397277e-05,
+      "loss": 3.4398,
+      "step": 1873920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9312656949886757e-05,
+      "loss": 3.4207,
+      "step": 1874432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.930428738117997e-05,
+      "loss": 3.4154,
+      "step": 1874944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.929590143366945e-05,
+      "loss": 3.4251,
+      "step": 1875456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.928751548615893e-05,
+      "loss": 3.4363,
+      "step": 1875968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9279129538648406e-05,
+      "loss": 3.4232,
+      "step": 1876480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9270759969941622e-05,
+      "loss": 3.4302,
+      "step": 1876992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9262374022431102e-05,
+      "loss": 3.4296,
+      "step": 1877504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9253988074920582e-05,
+      "loss": 3.4253,
+      "step": 1878016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.924561850621379e-05,
+      "loss": 3.4275,
+      "step": 1878528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.923723255870327e-05,
+      "loss": 3.419,
+      "step": 1879040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.922884661119275e-05,
+      "loss": 3.4287,
+      "step": 1879552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.922046066368223e-05,
+      "loss": 3.4382,
+      "step": 1880064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9212091094975443e-05,
+      "loss": 3.4285,
+      "step": 1880576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9203705147464923e-05,
+      "loss": 3.4132,
+      "step": 1881088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9195319199954403e-05,
+      "loss": 3.4233,
+      "step": 1881600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.918693325244388e-05,
+      "loss": 3.424,
+      "step": 1882112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9178563683737092e-05,
+      "loss": 3.4303,
+      "step": 1882624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9170177736226576e-05,
+      "loss": 3.4293,
+      "step": 1883136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9161791788716056e-05,
+      "loss": 3.4206,
+      "step": 1883648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9153405841205532e-05,
+      "loss": 3.4186,
+      "step": 1884160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9145036272498745e-05,
+      "loss": 3.4143,
+      "step": 1884672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9136650324988225e-05,
+      "loss": 3.4261,
+      "step": 1885184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9128264377477705e-05,
+      "loss": 3.4245,
+      "step": 1885696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9119878429967185e-05,
+      "loss": 3.4142,
+      "step": 1886208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9111508861260397e-05,
+      "loss": 3.4331,
+      "step": 1886720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9103122913749877e-05,
+      "loss": 3.4187,
+      "step": 1887232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9094736966239354e-05,
+      "loss": 3.4235,
+      "step": 1887744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9086351018728834e-05,
+      "loss": 3.4173,
+      "step": 1888256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9077981450022046e-05,
+      "loss": 3.4182,
+      "step": 1888768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.906959550251153e-05,
+      "loss": 3.4307,
+      "step": 1889280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9061209555001006e-05,
+      "loss": 3.4173,
+      "step": 1889792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9052823607490486e-05,
+      "loss": 3.4301,
+      "step": 1890304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.90444540387837e-05,
+      "loss": 3.4328,
+      "step": 1890816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9036068091273175e-05,
+      "loss": 3.4242,
+      "step": 1891328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.902768214376266e-05,
+      "loss": 3.4289,
+      "step": 1891840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.901929619625214e-05,
+      "loss": 3.4197,
+      "step": 1892352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.901092662754535e-05,
+      "loss": 3.4319,
+      "step": 1892864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9002540680034827e-05,
+      "loss": 3.4225,
+      "step": 1893376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8994154732524307e-05,
+      "loss": 3.4241,
+      "step": 1893888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8985768785013787e-05,
+      "loss": 3.4244,
+      "step": 1894400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8977399216307e-05,
+      "loss": 3.4299,
+      "step": 1894912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.896901326879648e-05,
+      "loss": 3.4329,
+      "step": 1895424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.896062732128596e-05,
+      "loss": 3.4176,
+      "step": 1895936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.895224137377544e-05,
+      "loss": 3.4127,
+      "step": 1896448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.894387180506865e-05,
+      "loss": 3.4179,
+      "step": 1896960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.893548585755813e-05,
+      "loss": 3.4219,
+      "step": 1897472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8927099910047612e-05,
+      "loss": 3.4289,
+      "step": 1897984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8918713962537092e-05,
+      "loss": 3.4214,
+      "step": 1898496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.89103443938303e-05,
+      "loss": 3.4336,
+      "step": 1899008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.890195844631978e-05,
+      "loss": 3.4142,
+      "step": 1899520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.889357249880926e-05,
+      "loss": 3.436,
+      "step": 1900032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.888518655129874e-05,
+      "loss": 3.4214,
+      "step": 1900544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8876816982591953e-05,
+      "loss": 3.4303,
+      "step": 1901056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8868431035081433e-05,
+      "loss": 3.424,
+      "step": 1901568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8860045087570913e-05,
+      "loss": 3.4205,
+      "step": 1902080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8851659140060393e-05,
+      "loss": 3.4221,
+      "step": 1902592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8843289571353602e-05,
+      "loss": 3.4392,
+      "step": 1903104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8834903623843082e-05,
+      "loss": 3.4156,
+      "step": 1903616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8826517676332566e-05,
+      "loss": 3.4304,
+      "step": 1904128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8818131728822046e-05,
+      "loss": 3.4246,
+      "step": 1904640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8809762160115255e-05,
+      "loss": 3.4247,
+      "step": 1905152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8801392591408467e-05,
+      "loss": 3.4266,
+      "step": 1905664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8793006643897944e-05,
+      "loss": 3.4255,
+      "step": 1906176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8784620696387427e-05,
+      "loss": 3.4229,
+      "step": 1906688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8776234748876907e-05,
+      "loss": 3.4208,
+      "step": 1907200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8767848801366387e-05,
+      "loss": 3.4181,
+      "step": 1907712
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8569862842559814,
+      "eval_runtime": 309.0612,
+      "eval_samples_per_second": 1234.678,
+      "eval_steps_per_second": 38.585,
+      "step": 1908000
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.3224327233854464e+18,
+  "trial_name": null,
+  "trial_params": null
+}