diff --git "a/checkpoint-686880/trainer_state.json" "b/checkpoint-686880/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-686880/trainer_state.json"
@@ -0,0 +1,8143 @@
+{
+  "best_metric": 3.8975584506988525,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/pp-mod-subj2/transformer/1/checkpoints/checkpoint-686880",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 686880,
+  "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": 11.0032,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8354,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1938,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9758,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8037,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.697,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.5986,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5281,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4457,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.3904,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3359,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.2962,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.2572,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.202,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1711,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1166,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1003,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0651,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0277,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0066,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0048,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9582,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815558291179756e-05,
+      "loss": 4.9452,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807172343669236e-05,
+      "loss": 4.9205,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798786396158716e-05,
+      "loss": 4.8958,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.8853,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8514,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8489,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8297,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8011,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.7972,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7753,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7745,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 4.7553,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 4.7388,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7338,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969820416244367e-05,
+      "loss": 4.7348,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.6987,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.6905,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6858,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.666,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6603,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6616,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6505,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.636,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6272,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6182,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6196,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.5866,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.5885,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.5946,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957246408619706e-05,
+      "loss": 4.5854,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 4.572,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5541,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5619,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5414,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5591,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 4.527,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 4.5422,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5223,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5109,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5144,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 4.5001,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4955,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4865,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.4961,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 4.4721,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4678,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4566,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9421582546222625e-05,
+      "loss": 4.4574,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4751,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4639,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4461,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4442,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4485,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4432,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.4278,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.4309,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4217,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.421,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.4152,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.4121,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4151,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 4.4073,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.3863,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 4.3959,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 4.3963,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 4.3903,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.3852,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925391273242342e-05,
+      "loss": 4.3734,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92455267849129e-05,
+      "loss": 4.3792,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923714083740238e-05,
+      "loss": 4.3681,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922878764749932e-05,
+      "loss": 4.3743,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 4.3627,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.3571,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.376,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.3426,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.3397,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.3497,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.3485,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.3516,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.3365,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3312,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.3282,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912818903498054e-05,
+      "loss": 4.3338,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911980308747002e-05,
+      "loss": 4.3145,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911143351876323e-05,
+      "loss": 4.3133,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.3262,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3158,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.3076,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9077906107524886e-05,
+      "loss": 4.3027,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9069520160014366e-05,
+      "loss": 4.3012,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9061134212503846e-05,
+      "loss": 4.3112,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052748264993326e-05,
+      "loss": 4.298,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90443623174828e-05,
+      "loss": 4.2954,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.2997,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.2999,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.2885,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.2926,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.2905,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994063011223415e-05,
+      "loss": 4.2875,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985677063712895e-05,
+      "loss": 4.2709,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977291116202375e-05,
+      "loss": 4.2835,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968905168691855e-05,
+      "loss": 4.2775,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.2819,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2656,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2745,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2629,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.2666,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2477,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2743,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2717,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889348077750837e-05,
+      "loss": 4.2545,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888509482999785e-05,
+      "loss": 4.2453,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887670888248733e-05,
+      "loss": 4.2659,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886832293497681e-05,
+      "loss": 4.2508,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.2532,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515674187595e-05,
+      "loss": 4.2464,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2543,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.2449,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.2448,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2291,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8809654060010626e-05,
+      "loss": 4.2423,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8801268112500106e-05,
+      "loss": 4.2384,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8792882164989586e-05,
+      "loss": 4.2298,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2476,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877612664877228e-05,
+      "loss": 4.2253,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876774070126176e-05,
+      "loss": 4.2313,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875935475375124e-05,
+      "loss": 4.2347,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875096880624072e-05,
+      "loss": 4.2239,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.237559795379639,
+      "eval_runtime": 304.2346,
+      "eval_samples_per_second": 1254.265,
+      "eval_steps_per_second": 39.197,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87425828587302e-05,
+      "loss": 4.2103,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873421329002341e-05,
+      "loss": 4.216,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872582734251289e-05,
+      "loss": 4.2254,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871744139500237e-05,
+      "loss": 4.2138,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870905544749185e-05,
+      "loss": 4.2163,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870066949998133e-05,
+      "loss": 4.2005,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869229993127454e-05,
+      "loss": 4.2023,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683913983764027e-05,
+      "loss": 4.1989,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675528036253507e-05,
+      "loss": 4.2093,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667142088742986e-05,
+      "loss": 4.1941,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658772520036196e-05,
+      "loss": 4.2013,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650402951329405e-05,
+      "loss": 4.206,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642017003818885e-05,
+      "loss": 4.1892,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633631056308365e-05,
+      "loss": 4.1848,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8625245108797845e-05,
+      "loss": 4.1756,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8616859161287324e-05,
+      "loss": 4.1822,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608473213776804e-05,
+      "loss": 4.1785,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860008726626628e-05,
+      "loss": 4.1746,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591701318755764e-05,
+      "loss": 4.1742,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583315371245244e-05,
+      "loss": 4.201,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574929423734724e-05,
+      "loss": 4.1712,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856655985502793e-05,
+      "loss": 4.173,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855817390751741e-05,
+      "loss": 4.1797,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854978796000689e-05,
+      "loss": 4.1716,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854140201249637e-05,
+      "loss": 4.1667,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853303244378958e-05,
+      "loss": 4.1602,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852464649627906e-05,
+      "loss": 4.1648,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851626054876854e-05,
+      "loss": 4.1597,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850787460125802e-05,
+      "loss": 4.1485,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849950503255123e-05,
+      "loss": 4.1619,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849111908504072e-05,
+      "loss": 4.1493,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84827331375302e-05,
+      "loss": 4.162,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847434719001968e-05,
+      "loss": 4.1572,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846597762131289e-05,
+      "loss": 4.1419,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845759167380237e-05,
+      "loss": 4.157,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844920572629185e-05,
+      "loss": 4.1636,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844081977878133e-05,
+      "loss": 4.1367,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432450210074536e-05,
+      "loss": 4.1468,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8424064262564016e-05,
+      "loss": 4.1407,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415678315053496e-05,
+      "loss": 4.1327,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8407292367542976e-05,
+      "loss": 4.1372,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398922798836185e-05,
+      "loss": 4.1461,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839053685132567e-05,
+      "loss": 4.1433,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838215090381515e-05,
+      "loss": 4.134,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837376495630463e-05,
+      "loss": 4.1414,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836539538759784e-05,
+      "loss": 4.1304,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835700944008732e-05,
+      "loss": 4.1374,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83486234925768e-05,
+      "loss": 4.1175,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834023754506628e-05,
+      "loss": 4.1189,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833186797635949e-05,
+      "loss": 4.1338,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83234984076527e-05,
+      "loss": 4.1347,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831511246014218e-05,
+      "loss": 4.122,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830672651263166e-05,
+      "loss": 4.1137,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829834056512114e-05,
+      "loss": 4.125,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289954617610625e-05,
+      "loss": 4.1067,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281568670100105e-05,
+      "loss": 4.1323,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273182722589585e-05,
+      "loss": 4.1138,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264796775079065e-05,
+      "loss": 4.1286,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256427206372274e-05,
+      "loss": 4.1149,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248041258861754e-05,
+      "loss": 4.1054,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239655311351234e-05,
+      "loss": 4.1097,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823128574264444e-05,
+      "loss": 4.1119,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822289979513392e-05,
+      "loss": 4.0998,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82145138476234e-05,
+      "loss": 4.0985,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820612790011288e-05,
+      "loss": 4.1141,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819775833140609e-05,
+      "loss": 4.0992,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818937238389558e-05,
+      "loss": 4.0901,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818098643638506e-05,
+      "loss": 4.0903,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817260048887454e-05,
+      "loss": 4.0904,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816423092016775e-05,
+      "loss": 4.1104,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815584497265723e-05,
+      "loss": 4.1066,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814745902514671e-05,
+      "loss": 4.0912,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813907307763619e-05,
+      "loss": 4.0936,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81307035089294e-05,
+      "loss": 4.0973,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812231756141888e-05,
+      "loss": 4.0952,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811393161390836e-05,
+      "loss": 4.0889,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810554566639784e-05,
+      "loss": 4.0936,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8097176097691046e-05,
+      "loss": 4.0882,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808879015018053e-05,
+      "loss": 4.0931,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808040420267001e-05,
+      "loss": 4.083,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807201825515949e-05,
+      "loss": 4.088,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80636486864527e-05,
+      "loss": 4.0891,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805526273894218e-05,
+      "loss": 4.0885,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804687679143166e-05,
+      "loss": 4.0718,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803849084392114e-05,
+      "loss": 4.0818,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803012127521435e-05,
+      "loss": 4.0802,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802173532770383e-05,
+      "loss": 4.0831,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801336575899704e-05,
+      "loss": 4.0799,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800497981148652e-05,
+      "loss": 4.0655,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996593863976e-05,
+      "loss": 4.0784,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988207916465486e-05,
+      "loss": 4.069,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979821968954966e-05,
+      "loss": 4.0789,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971436021444446e-05,
+      "loss": 4.0646,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7963050073933926e-05,
+      "loss": 4.0699,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954680505227135e-05,
+      "loss": 4.0802,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946294557716615e-05,
+      "loss": 4.0673,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937908610206095e-05,
+      "loss": 4.0503,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929539041499304e-05,
+      "loss": 4.0654,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7921153093988784e-05,
+      "loss": 4.0689,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912767146478264e-05,
+      "loss": 4.0704,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7904381198967744e-05,
+      "loss": 4.0627,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895995251457224e-05,
+      "loss": 4.0591,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887609303946704e-05,
+      "loss": 4.0524,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879223356436184e-05,
+      "loss": 4.066,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870837408925664e-05,
+      "loss": 4.0453,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786248421902261e-05,
+      "loss": 4.0499,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785409827151209e-05,
+      "loss": 4.0592,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784571232400157e-05,
+      "loss": 4.0601,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783732637649105e-05,
+      "loss": 4.0372,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782895680778426e-05,
+      "loss": 4.051,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782057086027374e-05,
+      "loss": 4.0407,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781218491276322e-05,
+      "loss": 4.0572,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78037989652527e-05,
+      "loss": 4.0438,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779541301774218e-05,
+      "loss": 4.0396,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778702707023166e-05,
+      "loss": 4.0582,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777864112272114e-05,
+      "loss": 4.0515,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770271554014354e-05,
+      "loss": 4.0433,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761885606503834e-05,
+      "loss": 4.0488,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775349965899331e-05,
+      "loss": 4.0443,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774511371148279e-05,
+      "loss": 4.0491,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7736744142776e-05,
+      "loss": 4.0281,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772835819526548e-05,
+      "loss": 4.0489,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719972247754956e-05,
+      "loss": 4.0364,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7711586300244436e-05,
+      "loss": 4.0442,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703200352733916e-05,
+      "loss": 4.0356,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694814405223396e-05,
+      "loss": 4.0406,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686428457712876e-05,
+      "loss": 4.0337,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7678042510202355e-05,
+      "loss": 4.0351,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766967294149557e-05,
+      "loss": 4.0172,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766128699398505e-05,
+      "loss": 4.0489,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765290104647453e-05,
+      "loss": 4.0399,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764453147776774e-05,
+      "loss": 4.0336,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763614553025722e-05,
+      "loss": 4.0236,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76277595827467e-05,
+      "loss": 4.0418,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761937363523618e-05,
+      "loss": 4.0365,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761098768772566e-05,
+      "loss": 4.029,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760260174021514e-05,
+      "loss": 4.0259,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759423217150835e-05,
+      "loss": 4.036,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758584622399783e-05,
+      "loss": 4.0296,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757746027648731e-05,
+      "loss": 4.039,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569074328976796e-05,
+      "loss": 4.015,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560704760270005e-05,
+      "loss": 4.0307,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552318812759485e-05,
+      "loss": 4.0278,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543932865248965e-05,
+      "loss": 4.021,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535546917738445e-05,
+      "loss": 4.0403,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527177349031654e-05,
+      "loss": 4.0192,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518791401521134e-05,
+      "loss": 4.0241,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510405454010614e-05,
+      "loss": 4.0243,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7502019506500094e-05,
+      "loss": 4.0224,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.068605422973633,
+      "eval_runtime": 305.2926,
+      "eval_samples_per_second": 1249.919,
+      "eval_steps_per_second": 39.061,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493633558989574e-05,
+      "loss": 4.0074,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485247611479054e-05,
+      "loss": 4.0163,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476861663968534e-05,
+      "loss": 4.0245,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468475716458014e-05,
+      "loss": 4.0164,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460089768947494e-05,
+      "loss": 4.0231,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451703821436974e-05,
+      "loss": 4.0026,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443317873926453e-05,
+      "loss": 4.0095,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743494830521967e-05,
+      "loss": 4.0054,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742656235770914e-05,
+      "loss": 4.0167,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741817641019862e-05,
+      "loss": 4.0047,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74097904626881e-05,
+      "loss": 4.0129,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740140451517758e-05,
+      "loss": 4.0161,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739303494647079e-05,
+      "loss": 3.9994,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738464899896027e-05,
+      "loss": 4.0034,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737626305144975e-05,
+      "loss": 3.9861,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736789348274297e-05,
+      "loss": 3.9984,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735950753523245e-05,
+      "loss": 3.9985,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735112158772193e-05,
+      "loss": 3.9927,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734273564021141e-05,
+      "loss": 3.9936,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733434969270089e-05,
+      "loss": 4.0183,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732596374519037e-05,
+      "loss": 3.9955,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7317594176483576e-05,
+      "loss": 3.9925,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7309208228973056e-05,
+      "loss": 4.0012,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300822281462536e-05,
+      "loss": 3.9992,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292436333952016e-05,
+      "loss": 3.9889,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7284066765245225e-05,
+      "loss": 3.9852,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275680817734705e-05,
+      "loss": 3.9942,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267294870224185e-05,
+      "loss": 3.9825,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258908922713665e-05,
+      "loss": 3.9804,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725053935400688e-05,
+      "loss": 3.9926,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724215340649636e-05,
+      "loss": 3.9818,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723376745898584e-05,
+      "loss": 3.9921,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722538151147532e-05,
+      "loss": 3.9875,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721701194276853e-05,
+      "loss": 3.9756,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720862599525801e-05,
+      "loss": 3.9894,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720024004774749e-05,
+      "loss": 4.0013,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719185410023697e-05,
+      "loss": 3.9715,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718346815272645e-05,
+      "loss": 3.9842,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717509858401966e-05,
+      "loss": 3.9737,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716671263650914e-05,
+      "loss": 3.9757,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715832668899862e-05,
+      "loss": 3.9726,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149940741488105e-05,
+      "loss": 3.9884,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141571172781314e-05,
+      "loss": 3.9832,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133185225270794e-05,
+      "loss": 3.9763,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124799277760274e-05,
+      "loss": 3.9822,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116413330249754e-05,
+      "loss": 3.9718,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7108043761542963e-05,
+      "loss": 3.984,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099657814032443e-05,
+      "loss": 3.9642,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709127186652192e-05,
+      "loss": 3.9648,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70828859190114e-05,
+      "loss": 3.9772,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707451635030461e-05,
+      "loss": 3.9776,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706613040279409e-05,
+      "loss": 3.9688,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705774445528357e-05,
+      "loss": 3.9634,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704935850777306e-05,
+      "loss": 3.9728,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704098893906627e-05,
+      "loss": 3.9542,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703260299155575e-05,
+      "loss": 3.9796,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702421704404523e-05,
+      "loss": 3.9671,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701583109653471e-05,
+      "loss": 3.9784,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700746152782792e-05,
+      "loss": 3.9716,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69990755803174e-05,
+      "loss": 3.9573,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699068963280688e-05,
+      "loss": 3.9619,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698230368529636e-05,
+      "loss": 3.9666,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697391773778584e-05,
+      "loss": 3.9537,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965548169079046e-05,
+      "loss": 3.9544,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6957162221568526e-05,
+      "loss": 3.9666,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694877627405801e-05,
+      "loss": 3.9583,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694039032654749e-05,
+      "loss": 3.9452,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69320207578407e-05,
+      "loss": 3.9479,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692363481033018e-05,
+      "loss": 3.9486,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691524886281966e-05,
+      "loss": 3.9685,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690686291530914e-05,
+      "loss": 3.9632,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689849334660235e-05,
+      "loss": 3.9523,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689010739909183e-05,
+      "loss": 3.9583,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688172145158131e-05,
+      "loss": 3.9555,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687333550407079e-05,
+      "loss": 3.9594,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864965935364e-05,
+      "loss": 3.9485,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685657998785348e-05,
+      "loss": 3.9598,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848194040342966e-05,
+      "loss": 3.9543,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6839808092832446e-05,
+      "loss": 3.9561,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831438524125655e-05,
+      "loss": 3.9465,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6823052576615135e-05,
+      "loss": 3.9542,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6814666629104615e-05,
+      "loss": 3.9546,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806280681594095e-05,
+      "loss": 3.9531,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6797911112887304e-05,
+      "loss": 3.9413,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789525165376784e-05,
+      "loss": 3.9469,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6781139217866264e-05,
+      "loss": 3.948,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772753270355744e-05,
+      "loss": 3.9491,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676438370164895e-05,
+      "loss": 3.9514,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675599775413843e-05,
+      "loss": 3.9336,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674761180662792e-05,
+      "loss": 3.9466,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.67392258591174e-05,
+      "loss": 3.9423,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673085629041061e-05,
+      "loss": 3.95,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672247034290009e-05,
+      "loss": 3.9323,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671408439538957e-05,
+      "loss": 3.9421,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670569844787904e-05,
+      "loss": 3.9535,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669732887917226e-05,
+      "loss": 3.9391,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668894293166174e-05,
+      "loss": 3.9237,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668055698415122e-05,
+      "loss": 3.9388,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667217103664069e-05,
+      "loss": 3.9433,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666380146793391e-05,
+      "loss": 3.9443,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665541552042339e-05,
+      "loss": 3.941,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6647029572912874e-05,
+      "loss": 3.9305,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663864362540235e-05,
+      "loss": 3.9297,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663025767789183e-05,
+      "loss": 3.9413,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662188810918504e-05,
+      "loss": 3.9234,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613502161674516e-05,
+      "loss": 3.9288,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6605116214163996e-05,
+      "loss": 3.9364,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596730266653476e-05,
+      "loss": 3.938,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658836069794669e-05,
+      "loss": 3.9127,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579974750436165e-05,
+      "loss": 3.9301,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571588802925645e-05,
+      "loss": 3.9208,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563202855415125e-05,
+      "loss": 3.9341,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655483328670834e-05,
+      "loss": 3.9291,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654644733919782e-05,
+      "loss": 3.9209,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65380613916873e-05,
+      "loss": 3.9381,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652967544417678e-05,
+      "loss": 3.9349,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652130587546999e-05,
+      "loss": 3.9232,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651291992795947e-05,
+      "loss": 3.9304,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650453398044895e-05,
+      "loss": 3.9295,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649614803293843e-05,
+      "loss": 3.9271,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648777846423164e-05,
+      "loss": 3.9145,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647939251672112e-05,
+      "loss": 3.9287,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64710065692106e-05,
+      "loss": 3.9226,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646262062170008e-05,
+      "loss": 3.9257,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454251052993294e-05,
+      "loss": 3.9243,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445865105482774e-05,
+      "loss": 3.9288,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437479157972254e-05,
+      "loss": 3.9142,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429093210461734e-05,
+      "loss": 3.919,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642072364175494e-05,
+      "loss": 3.9066,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641233769424442e-05,
+      "loss": 3.9355,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64039517467339e-05,
+      "loss": 3.9231,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639556579922338e-05,
+      "loss": 3.9227,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638719623051659e-05,
+      "loss": 3.9126,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637881028300607e-05,
+      "loss": 3.929,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637042433549555e-05,
+      "loss": 3.926,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636203838798503e-05,
+      "loss": 3.9155,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635366881927825e-05,
+      "loss": 3.9154,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634528287176773e-05,
+      "loss": 3.9233,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633689692425721e-05,
+      "loss": 3.9195,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632851097674669e-05,
+      "loss": 3.9266,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63201414080399e-05,
+      "loss": 3.9084,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631175546052938e-05,
+      "loss": 3.9195,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630336951301886e-05,
+      "loss": 3.9163,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629498356550834e-05,
+      "loss": 3.9104,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286613996801546e-05,
+      "loss": 3.9307,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6278228049291026e-05,
+      "loss": 3.9114,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269842101780506e-05,
+      "loss": 3.9154,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6261456154269986e-05,
+      "loss": 3.9171,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62530865855632e-05,
+      "loss": 3.9119,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9972167015075684,
+      "eval_runtime": 302.4616,
+      "eval_samples_per_second": 1261.618,
+      "eval_steps_per_second": 39.426,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624470063805268e-05,
+      "loss": 3.9094,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623631469054216e-05,
+      "loss": 3.9073,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622792874303164e-05,
+      "loss": 3.9171,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621955917432485e-05,
+      "loss": 3.9116,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621117322681433e-05,
+      "loss": 3.9173,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620278727930381e-05,
+      "loss": 3.9006,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619440133179329e-05,
+      "loss": 3.9078,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61860317630865e-05,
+      "loss": 3.8956,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617764581557598e-05,
+      "loss": 3.9176,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616925986806546e-05,
+      "loss": 3.8987,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160890299358675e-05,
+      "loss": 3.9098,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152504351848155e-05,
+      "loss": 3.9113,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6144134783141364e-05,
+      "loss": 3.9017,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135748835630844e-05,
+      "loss": 3.8943,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127362888120324e-05,
+      "loss": 3.8921,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6118976940609804e-05,
+      "loss": 3.8924,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110590993099284e-05,
+      "loss": 3.8994,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6102205045588764e-05,
+      "loss": 3.8941,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093819098078244e-05,
+      "loss": 3.8942,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085433150567724e-05,
+      "loss": 3.9147,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607706358186093e-05,
+      "loss": 3.8976,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606867763435041e-05,
+      "loss": 3.893,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606029168683989e-05,
+      "loss": 3.9009,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605190573932937e-05,
+      "loss": 3.9001,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604353617062259e-05,
+      "loss": 3.8916,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603515022311207e-05,
+      "loss": 3.8864,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602676427560155e-05,
+      "loss": 3.9007,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601837832809103e-05,
+      "loss": 3.8799,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601000875938424e-05,
+      "loss": 3.8857,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600162281187372e-05,
+      "loss": 3.8913,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59932368643632e-05,
+      "loss": 3.8852,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598485091685268e-05,
+      "loss": 3.8984,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597648134814589e-05,
+      "loss": 3.8898,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596809540063537e-05,
+      "loss": 3.8822,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595970945312485e-05,
+      "loss": 3.8947,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595133988441806e-05,
+      "loss": 3.9022,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594295393690754e-05,
+      "loss": 3.8807,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593456798939702e-05,
+      "loss": 3.89,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59261820418865e-05,
+      "loss": 3.875,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591781247317971e-05,
+      "loss": 3.8893,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590942652566919e-05,
+      "loss": 3.8749,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590104057815867e-05,
+      "loss": 3.8977,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589265463064815e-05,
+      "loss": 3.8884,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588428506194136e-05,
+      "loss": 3.8847,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587589911443084e-05,
+      "loss": 3.8883,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586751316692032e-05,
+      "loss": 3.8815,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58591272194098e-05,
+      "loss": 3.8892,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5850757650703016e-05,
+      "loss": 3.873,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5842371703192496e-05,
+      "loss": 3.8709,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5833985755681976e-05,
+      "loss": 3.8881,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5825616186975185e-05,
+      "loss": 3.8842,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5817230239464665e-05,
+      "loss": 3.8822,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808844291954145e-05,
+      "loss": 3.8711,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800458344443625e-05,
+      "loss": 3.8799,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5792072396933105e-05,
+      "loss": 3.8643,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783686449422585e-05,
+      "loss": 3.8885,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5775300501912065e-05,
+      "loss": 3.8769,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576691455440154e-05,
+      "loss": 3.8921,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758544985694754e-05,
+      "loss": 3.8825,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5750159038184234e-05,
+      "loss": 3.8707,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741773090673714e-05,
+      "loss": 3.8689,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733387143163194e-05,
+      "loss": 3.8783,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572501757445641e-05,
+      "loss": 3.8677,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571663162694589e-05,
+      "loss": 3.8667,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570824567943537e-05,
+      "loss": 3.8762,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569985973192484e-05,
+      "loss": 3.8755,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569149016321806e-05,
+      "loss": 3.8564,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568310421570754e-05,
+      "loss": 3.8609,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567471826819701e-05,
+      "loss": 3.8594,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566633232068649e-05,
+      "loss": 3.8837,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565796275197971e-05,
+      "loss": 3.8749,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564957680446919e-05,
+      "loss": 3.8648,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564119085695867e-05,
+      "loss": 3.8726,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632821288251884e-05,
+      "loss": 3.8676,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624435340741364e-05,
+      "loss": 3.8783,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5616049393230843e-05,
+      "loss": 3.8588,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560766344572032e-05,
+      "loss": 3.8794,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559929387701353e-05,
+      "loss": 3.8662,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559090792950301e-05,
+      "loss": 3.8707,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582521981992486e-05,
+      "loss": 3.868,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5574136034481966e-05,
+      "loss": 3.866,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556576646577518e-05,
+      "loss": 3.8749,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555738051826466e-05,
+      "loss": 3.8662,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554899457075414e-05,
+      "loss": 3.8574,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554060862324362e-05,
+      "loss": 3.8657,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553223905453684e-05,
+      "loss": 3.8609,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552385310702631e-05,
+      "loss": 3.8722,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551546715951579e-05,
+      "loss": 3.8672,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550708121200527e-05,
+      "loss": 3.8507,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498711643298486e-05,
+      "loss": 3.8659,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549032569578796e-05,
+      "loss": 3.855,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548193974827744e-05,
+      "loss": 3.8664,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547355380076692e-05,
+      "loss": 3.8543,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5465184232060135e-05,
+      "loss": 3.8626,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5456798284549615e-05,
+      "loss": 3.871,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5448412337039095e-05,
+      "loss": 3.8563,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5440026389528575e-05,
+      "loss": 3.8464,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431656820821784e-05,
+      "loss": 3.8592,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5423270873311264e-05,
+      "loss": 3.8559,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414884925800744e-05,
+      "loss": 3.8693,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406498978290224e-05,
+      "loss": 3.8597,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539812940958343e-05,
+      "loss": 3.8511,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538974346207291e-05,
+      "loss": 3.8515,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538135751456239e-05,
+      "loss": 3.8628,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537297156705187e-05,
+      "loss": 3.8415,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536460199834508e-05,
+      "loss": 3.8529,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535621605083457e-05,
+      "loss": 3.8531,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534783010332405e-05,
+      "loss": 3.8598,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533944415581353e-05,
+      "loss": 3.8305,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533107458710674e-05,
+      "loss": 3.8524,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532268863959622e-05,
+      "loss": 3.8502,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53143026920857e-05,
+      "loss": 3.8487,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530591674457518e-05,
+      "loss": 3.8543,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529754717586839e-05,
+      "loss": 3.8405,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528916122835787e-05,
+      "loss": 3.8619,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528077528084735e-05,
+      "loss": 3.8552,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527238933333683e-05,
+      "loss": 3.8464,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5264019764630036e-05,
+      "loss": 3.8533,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525563381711952e-05,
+      "loss": 3.8526,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5247247869609e-05,
+      "loss": 3.8515,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523886192209848e-05,
+      "loss": 3.833,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523049235339169e-05,
+      "loss": 3.8556,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522210640588117e-05,
+      "loss": 3.8438,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521372045837065e-05,
+      "loss": 3.8479,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520533451086013e-05,
+      "loss": 3.8474,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519696494215334e-05,
+      "loss": 3.8534,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518857899464282e-05,
+      "loss": 3.8433,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51801930471323e-05,
+      "loss": 3.8389,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517180709962178e-05,
+      "loss": 3.8356,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516343753091499e-05,
+      "loss": 3.8564,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5155051583404476e-05,
+      "loss": 3.8476,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146665635893956e-05,
+      "loss": 3.8486,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5138279688383436e-05,
+      "loss": 3.8377,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129910119676645e-05,
+      "loss": 3.8541,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121524172166125e-05,
+      "loss": 3.85,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5113138224655605e-05,
+      "loss": 3.8419,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104752277145085e-05,
+      "loss": 3.8414,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096382708438294e-05,
+      "loss": 3.848,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087996760927774e-05,
+      "loss": 3.8485,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079610813417254e-05,
+      "loss": 3.8525,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5071224865906734e-05,
+      "loss": 3.8354,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506287167600368e-05,
+      "loss": 3.847,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505448572849316e-05,
+      "loss": 3.8422,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504609978098264e-05,
+      "loss": 3.8411,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503771383347212e-05,
+      "loss": 3.8554,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50293278859616e-05,
+      "loss": 3.8402,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502094193845108e-05,
+      "loss": 3.8447,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501255599094056e-05,
+      "loss": 3.8381,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500417004343004e-05,
+      "loss": 3.844,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.955796003341675,
+      "eval_runtime": 301.8628,
+      "eval_samples_per_second": 1264.121,
+      "eval_steps_per_second": 39.505,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499580047472325e-05,
+      "loss": 3.8275,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498741452721273e-05,
+      "loss": 3.8387,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497902857970221e-05,
+      "loss": 3.8439,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497064263219169e-05,
+      "loss": 3.8407,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496225668468117e-05,
+      "loss": 3.8494,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495387073717065e-05,
+      "loss": 3.8277,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945501168463863e-05,
+      "loss": 3.8378,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493711522095334e-05,
+      "loss": 3.8264,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492872927344282e-05,
+      "loss": 3.8434,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49203433259323e-05,
+      "loss": 3.8352,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491195737842178e-05,
+      "loss": 3.8365,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490357143091126e-05,
+      "loss": 3.8365,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489520186220447e-05,
+      "loss": 3.8321,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488681591469395e-05,
+      "loss": 3.8241,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487842996718343e-05,
+      "loss": 3.828,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487004401967291e-05,
+      "loss": 3.8171,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861658072162385e-05,
+      "loss": 3.8307,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4853272124651865e-05,
+      "loss": 3.8239,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484488617714135e-05,
+      "loss": 3.823,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483651660843456e-05,
+      "loss": 3.8511,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482813066092404e-05,
+      "loss": 3.8241,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481976109221726e-05,
+      "loss": 3.8278,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481137514470674e-05,
+      "loss": 3.8317,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480298919719622e-05,
+      "loss": 3.8286,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479460324968569e-05,
+      "loss": 3.8251,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478621730217517e-05,
+      "loss": 3.821,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477783135466465e-05,
+      "loss": 3.8324,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476944540715413e-05,
+      "loss": 3.8072,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476107583844734e-05,
+      "loss": 3.8196,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475268989093682e-05,
+      "loss": 3.8191,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47443039434263e-05,
+      "loss": 3.8201,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4735917995915786e-05,
+      "loss": 3.83,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4727532048405266e-05,
+      "loss": 3.8257,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4719146100894746e-05,
+      "loss": 3.8186,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710760153384226e-05,
+      "loss": 3.8274,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4702374205873705e-05,
+      "loss": 3.8302,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4694004637166915e-05,
+      "loss": 3.8157,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685618689656395e-05,
+      "loss": 3.8243,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4677232742145874e-05,
+      "loss": 3.8098,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668846794635354e-05,
+      "loss": 3.8236,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660477225928564e-05,
+      "loss": 3.8059,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4652091278418043e-05,
+      "loss": 3.8345,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643705330907523e-05,
+      "loss": 3.8233,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4635319383397003e-05,
+      "loss": 3.8181,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462694981469022e-05,
+      "loss": 3.8235,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461858024598343e-05,
+      "loss": 3.8171,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461019429847291e-05,
+      "loss": 3.8239,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460180835096239e-05,
+      "loss": 3.8094,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459342240345187e-05,
+      "loss": 3.8058,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458503645594135e-05,
+      "loss": 3.825,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457666688723456e-05,
+      "loss": 3.8184,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456828093972404e-05,
+      "loss": 3.818,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455989499221352e-05,
+      "loss": 3.8093,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551509044703e-05,
+      "loss": 3.8151,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454312309719248e-05,
+      "loss": 3.8062,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453473714968196e-05,
+      "loss": 3.82,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452635120217144e-05,
+      "loss": 3.8132,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517965254660924e-05,
+      "loss": 3.8299,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450959568595413e-05,
+      "loss": 3.818,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450120973844361e-05,
+      "loss": 3.8066,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449282379093309e-05,
+      "loss": 3.8098,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448443784342257e-05,
+      "loss": 3.8141,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447606827471578e-05,
+      "loss": 3.804,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446768232720526e-05,
+      "loss": 3.8067,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445931275849847e-05,
+      "loss": 3.8111,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445092681098795e-05,
+      "loss": 3.8135,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444254086347743e-05,
+      "loss": 3.7938,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443415491596691e-05,
+      "loss": 3.8015,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442576896845639e-05,
+      "loss": 3.7982,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441738302094588e-05,
+      "loss": 3.8153,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440899707343535e-05,
+      "loss": 3.8123,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400627504728566e-05,
+      "loss": 3.8057,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4392241557218046e-05,
+      "loss": 3.8103,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4383855609707526e-05,
+      "loss": 3.8065,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4375469662197e-05,
+      "loss": 3.8214,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436708371468648e-05,
+      "loss": 3.7972,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435869776717596e-05,
+      "loss": 3.8138,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435031181966544e-05,
+      "loss": 3.8081,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434194225095865e-05,
+      "loss": 3.8074,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433355630344813e-05,
+      "loss": 3.8089,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4325170355937615e-05,
+      "loss": 3.8063,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316784408427095e-05,
+      "loss": 3.8107,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308398460916575e-05,
+      "loss": 3.8073,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4300012513406055e-05,
+      "loss": 3.7915,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291626565895535e-05,
+      "loss": 3.8067,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4283240618385015e-05,
+      "loss": 3.8025,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274871049678224e-05,
+      "loss": 3.8097,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426650148097143e-05,
+      "loss": 3.8052,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425811553346091e-05,
+      "loss": 3.7911,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424972958595039e-05,
+      "loss": 3.806,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424134363843987e-05,
+      "loss": 3.7991,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423295769092935e-05,
+      "loss": 3.8091,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422457174341883e-05,
+      "loss": 3.7907,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421618579590831e-05,
+      "loss": 3.8044,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420779984839779e-05,
+      "loss": 3.818,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419943027969101e-05,
+      "loss": 3.7957,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419104433218049e-05,
+      "loss": 3.7831,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418265838466997e-05,
+      "loss": 3.7986,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417427243715945e-05,
+      "loss": 3.8005,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416590286845266e-05,
+      "loss": 3.8068,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415751692094214e-05,
+      "loss": 3.8021,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414913097343162e-05,
+      "loss": 3.7924,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414076140472483e-05,
+      "loss": 3.7934,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413237545721431e-05,
+      "loss": 3.8047,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412398950970379e-05,
+      "loss": 3.7826,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115619940997e-05,
+      "loss": 3.7944,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410723399348648e-05,
+      "loss": 3.7963,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409884804597596e-05,
+      "loss": 3.8019,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409046209846544e-05,
+      "loss": 3.7748,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408207615095492e-05,
+      "loss": 3.7953,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407370658224813e-05,
+      "loss": 3.7895,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406532063473761e-05,
+      "loss": 3.7952,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405693468722709e-05,
+      "loss": 3.7936,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404854873971657e-05,
+      "loss": 3.7865,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404016279220605e-05,
+      "loss": 3.8008,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403177684469553e-05,
+      "loss": 3.8012,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402340727598874e-05,
+      "loss": 3.7882,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401502132847822e-05,
+      "loss": 3.7994,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40066353809677e-05,
+      "loss": 3.7916,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399824943345719e-05,
+      "loss": 3.7972,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398986348594666e-05,
+      "loss": 3.7796,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981493917239876e-05,
+      "loss": 3.798,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3973107969729356e-05,
+      "loss": 3.7819,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396472202221883e-05,
+      "loss": 3.7936,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395633607470831e-05,
+      "loss": 3.7856,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394795012719779e-05,
+      "loss": 3.7999,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393956417968727e-05,
+      "loss": 3.7901,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393117823217675e-05,
+      "loss": 3.7843,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392280866346996e-05,
+      "loss": 3.7758,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391442271595944e-05,
+      "loss": 3.8053,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3906036768448925e-05,
+      "loss": 3.7859,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897650820938405e-05,
+      "loss": 3.7934,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3889281252231614e-05,
+      "loss": 3.7826,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880895304721094e-05,
+      "loss": 3.798,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872509357210574e-05,
+      "loss": 3.7985,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3864123409700054e-05,
+      "loss": 3.7837,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385575384099326e-05,
+      "loss": 3.7858,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384736789348274e-05,
+      "loss": 3.7943,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383898194597222e-05,
+      "loss": 3.7953,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38305959984617e-05,
+      "loss": 3.7993,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382221005095118e-05,
+      "loss": 3.7781,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381382410344066e-05,
+      "loss": 3.7912,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380543815593014e-05,
+      "loss": 3.7854,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379706858722336e-05,
+      "loss": 3.7903,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378868263971284e-05,
+      "loss": 3.797,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378029669220232e-05,
+      "loss": 3.784,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37719107446918e-05,
+      "loss": 3.7963,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376352479718128e-05,
+      "loss": 3.7803,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375513884967076e-05,
+      "loss": 3.7955,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.931006669998169,
+      "eval_runtime": 306.1997,
+      "eval_samples_per_second": 1246.216,
+      "eval_steps_per_second": 38.945,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374675290216024e-05,
+      "loss": 3.777,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373836695464972e-05,
+      "loss": 3.7869,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372999738594293e-05,
+      "loss": 3.7876,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372161143843241e-05,
+      "loss": 3.7853,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371322549092189e-05,
+      "loss": 3.793,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370483954341137e-05,
+      "loss": 3.7778,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369648635350831e-05,
+      "loss": 3.781,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368810040599779e-05,
+      "loss": 3.7722,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367971445848727e-05,
+      "loss": 3.7903,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367132851097675e-05,
+      "loss": 3.7858,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366294256346623e-05,
+      "loss": 3.7793,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365455661595571e-05,
+      "loss": 3.7863,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364618704724892e-05,
+      "loss": 3.7757,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36378010997384e-05,
+      "loss": 3.7737,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362941515222788e-05,
+      "loss": 3.7755,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362102920471736e-05,
+      "loss": 3.7635,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361264325720684e-05,
+      "loss": 3.7764,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360427368850005e-05,
+      "loss": 3.7692,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359588774098953e-05,
+      "loss": 3.7712,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358750179347901e-05,
+      "loss": 3.8039,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357911584596849e-05,
+      "loss": 3.7719,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357072989845797e-05,
+      "loss": 3.7768,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356234395094745e-05,
+      "loss": 3.7773,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355395800343693e-05,
+      "loss": 3.7784,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354558843473014e-05,
+      "loss": 3.7708,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353720248721962e-05,
+      "loss": 3.7735,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35288165397091e-05,
+      "loss": 3.7779,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352043059219858e-05,
+      "loss": 3.7577,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351204464468806e-05,
+      "loss": 3.768,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350367507598127e-05,
+      "loss": 3.7668,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349528912847075e-05,
+      "loss": 3.7735,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486903180960234e-05,
+      "loss": 3.7759,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478517233449714e-05,
+      "loss": 3.7754,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3470131285939194e-05,
+      "loss": 3.7704,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461745338428674e-05,
+      "loss": 3.7757,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3453359390918154e-05,
+      "loss": 3.778,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344498982221136e-05,
+      "loss": 3.7669,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343660387470084e-05,
+      "loss": 3.7771,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342821792719032e-05,
+      "loss": 3.7561,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34198319796798e-05,
+      "loss": 3.7758,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341146241097301e-05,
+      "loss": 3.7519,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340309284226622e-05,
+      "loss": 3.7875,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33947068947557e-05,
+      "loss": 3.7718,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338632094724519e-05,
+      "loss": 3.7686,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337793499973467e-05,
+      "loss": 3.7724,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336956543102788e-05,
+      "loss": 3.7709,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336117948351736e-05,
+      "loss": 3.7708,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335279353600684e-05,
+      "loss": 3.7572,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334440758849632e-05,
+      "loss": 3.7499,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3336038019789526e-05,
+      "loss": 3.7813,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3327652072279006e-05,
+      "loss": 3.7669,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3319266124768486e-05,
+      "loss": 3.7726,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3310880177257966e-05,
+      "loss": 3.7591,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3302494229747446e-05,
+      "loss": 3.7658,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3294124661040655e-05,
+      "loss": 3.7559,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328573871353014e-05,
+      "loss": 3.7714,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327735276601962e-05,
+      "loss": 3.7617,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32689668185091e-05,
+      "loss": 3.7775,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326058087099858e-05,
+      "loss": 3.7725,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325219492348806e-05,
+      "loss": 3.756,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324380897597754e-05,
+      "loss": 3.7615,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323543940727075e-05,
+      "loss": 3.7683,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322705345976023e-05,
+      "loss": 3.7563,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321866751224971e-05,
+      "loss": 3.7552,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321028156473919e-05,
+      "loss": 3.7654,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32019119960324e-05,
+      "loss": 3.7632,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319352604852188e-05,
+      "loss": 3.7449,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318514010101136e-05,
+      "loss": 3.7536,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317675415350084e-05,
+      "loss": 3.7493,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3168368205990326e-05,
+      "loss": 3.7676,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3159998637283535e-05,
+      "loss": 3.7617,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151612689773015e-05,
+      "loss": 3.7582,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3143226742262495e-05,
+      "loss": 3.7669,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3134840794751975e-05,
+      "loss": 3.7542,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312645484724145e-05,
+      "loss": 3.7748,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3118085278534664e-05,
+      "loss": 3.7484,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3109699331024144e-05,
+      "loss": 3.7666,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310131338351362e-05,
+      "loss": 3.7596,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30929274360031e-05,
+      "loss": 3.7604,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308454148849258e-05,
+      "loss": 3.7631,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3076155540982064e-05,
+      "loss": 3.76,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3067769593471544e-05,
+      "loss": 3.7615,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3059383645961024e-05,
+      "loss": 3.7618,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305101407725423e-05,
+      "loss": 3.7459,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304262812974371e-05,
+      "loss": 3.7596,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303424218223319e-05,
+      "loss": 3.7554,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302585623472267e-05,
+      "loss": 3.7647,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301750304481962e-05,
+      "loss": 3.7548,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300911709730909e-05,
+      "loss": 3.747,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300073114979857e-05,
+      "loss": 3.7567,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299234520228805e-05,
+      "loss": 3.7543,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.298395925477753e-05,
+      "loss": 3.759,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297558968607075e-05,
+      "loss": 3.7461,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2967203738560227e-05,
+      "loss": 3.7576,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2958817791049707e-05,
+      "loss": 3.7683,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2950431843539186e-05,
+      "loss": 3.7529,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2942062274832396e-05,
+      "loss": 3.7348,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2933676327321876e-05,
+      "loss": 3.7527,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2925290379811355e-05,
+      "loss": 3.7544,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2916904432300835e-05,
+      "loss": 3.7571,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2908518484790315e-05,
+      "loss": 3.7609,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2900148916083525e-05,
+      "loss": 3.7486,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2891762968573004e-05,
+      "loss": 3.7461,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2883377021062484e-05,
+      "loss": 3.7569,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287499107355197e-05,
+      "loss": 3.7389,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286660512604145e-05,
+      "loss": 3.7498,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285823555733466e-05,
+      "loss": 3.7462,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284984960982414e-05,
+      "loss": 3.7571,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284146366231362e-05,
+      "loss": 3.7321,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28330777148031e-05,
+      "loss": 3.7448,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282469176729258e-05,
+      "loss": 3.7485,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281630581978206e-05,
+      "loss": 3.7473,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280791987227154e-05,
+      "loss": 3.7507,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279953392476102e-05,
+      "loss": 3.7402,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279116435605423e-05,
+      "loss": 3.7572,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278279478734744e-05,
+      "loss": 3.7556,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277440883983692e-05,
+      "loss": 3.7423,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2766022892326405e-05,
+      "loss": 3.7567,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757636944815885e-05,
+      "loss": 3.7436,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2749267376109094e-05,
+      "loss": 3.7548,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2740881428598574e-05,
+      "loss": 3.7397,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2732495481088054e-05,
+      "loss": 3.7499,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2724109533577534e-05,
+      "loss": 3.7375,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2715723586067014e-05,
+      "loss": 3.7477,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2707337638556494e-05,
+      "loss": 3.7441,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26989680698497e-05,
+      "loss": 3.7517,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269058212233918e-05,
+      "loss": 3.7441,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268219617482866e-05,
+      "loss": 3.7436,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267381022731814e-05,
+      "loss": 3.7288,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266542427980762e-05,
+      "loss": 3.7619,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265705471110084e-05,
+      "loss": 3.7418,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264866876359032e-05,
+      "loss": 3.75,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26402828160798e-05,
+      "loss": 3.7371,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263189686856928e-05,
+      "loss": 3.7536,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262351092105876e-05,
+      "loss": 3.7517,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261512497354823e-05,
+      "loss": 3.7415,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260675540484145e-05,
+      "loss": 3.7459,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259836945733093e-05,
+      "loss": 3.7487,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25899835098204e-05,
+      "loss": 3.75,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258159756230988e-05,
+      "loss": 3.7614,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257321161479936e-05,
+      "loss": 3.7306,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2564842046092576e-05,
+      "loss": 3.7515,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2556456098582056e-05,
+      "loss": 3.7373,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2548070151071536e-05,
+      "loss": 3.7509,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2539684203561016e-05,
+      "loss": 3.7502,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2531298256050496e-05,
+      "loss": 3.7391,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2522912308539976e-05,
+      "loss": 3.7536,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2514542739833185e-05,
+      "loss": 3.7397,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2506156792322665e-05,
+      "loss": 3.7466,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9170572757720947,
+      "eval_runtime": 309.1665,
+      "eval_samples_per_second": 1234.257,
+      "eval_steps_per_second": 38.571,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2497770844812145e-05,
+      "loss": 3.7334,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2489384897301625e-05,
+      "loss": 3.739,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480998949791105e-05,
+      "loss": 3.7483,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472613002280585e-05,
+      "loss": 3.7434,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2464227054770065e-05,
+      "loss": 3.7482,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245585748606328e-05,
+      "loss": 3.7397,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244747153855276e-05,
+      "loss": 3.7307,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243908559104224e-05,
+      "loss": 3.7358,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243069964353172e-05,
+      "loss": 3.7434,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24223136960212e-05,
+      "loss": 3.7449,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241392774851068e-05,
+      "loss": 3.7367,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240554180100016e-05,
+      "loss": 3.7452,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239717223229337e-05,
+      "loss": 3.7292,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238878628478285e-05,
+      "loss": 3.7336,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238040033727233e-05,
+      "loss": 3.734,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237201438976181e-05,
+      "loss": 3.7221,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236362844225129e-05,
+      "loss": 3.7355,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23552588735445e-05,
+      "loss": 3.7242,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234687292603398e-05,
+      "loss": 3.7298,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233848697852346e-05,
+      "loss": 3.7626,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2330101031012945e-05,
+      "loss": 3.732,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321731462306154e-05,
+      "loss": 3.7337,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2313345514795634e-05,
+      "loss": 3.735,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304959567285114e-05,
+      "loss": 3.7418,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229657361977459e-05,
+      "loss": 3.7215,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22882040510678e-05,
+      "loss": 3.7333,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227981810355728e-05,
+      "loss": 3.7343,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227143215604676e-05,
+      "loss": 3.7203,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2263046208536236e-05,
+      "loss": 3.727,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254660261025716e-05,
+      "loss": 3.7248,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2246274313515196e-05,
+      "loss": 3.7334,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223790474480841e-05,
+      "loss": 3.7371,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222951879729789e-05,
+      "loss": 3.7318,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222113284978737e-05,
+      "loss": 3.7292,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221274690227685e-05,
+      "loss": 3.7322,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220436095476633e-05,
+      "loss": 3.7373,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219597500725581e-05,
+      "loss": 3.7259,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218758905974529e-05,
+      "loss": 3.7392,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21792194910385e-05,
+      "loss": 3.7154,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217083354352798e-05,
+      "loss": 3.7332,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216244759601746e-05,
+      "loss": 3.7116,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215406164850694e-05,
+      "loss": 3.7452,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214567570099642e-05,
+      "loss": 3.7304,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2137306132289637e-05,
+      "loss": 3.7316,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128920184779116e-05,
+      "loss": 3.7314,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2120534237268596e-05,
+      "loss": 3.7312,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2112148289758076e-05,
+      "loss": 3.7337,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103762342247556e-05,
+      "loss": 3.7175,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2095392773540765e-05,
+      "loss": 3.7073,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2087006826030245e-05,
+      "loss": 3.7415,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2078620878519725e-05,
+      "loss": 3.7253,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2070234931009205e-05,
+      "loss": 3.7364,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061848983498685e-05,
+      "loss": 3.7179,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2053463035988165e-05,
+      "loss": 3.7256,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2045093467281374e-05,
+      "loss": 3.7121,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036707519770854e-05,
+      "loss": 3.7339,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2028321572260334e-05,
+      "loss": 3.7273,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201993562474982e-05,
+      "loss": 3.7331,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20115496772393e-05,
+      "loss": 3.7347,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200316372972878e-05,
+      "loss": 3.7143,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1994777782218254e-05,
+      "loss": 3.7201,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198640821351147e-05,
+      "loss": 3.734,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197802226600095e-05,
+      "loss": 3.7125,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196963631849042e-05,
+      "loss": 3.7172,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19612503709799e-05,
+      "loss": 3.724,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195288080227312e-05,
+      "loss": 3.7311,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19444948547626e-05,
+      "loss": 3.7042,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193610890725207e-05,
+      "loss": 3.7132,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192772295974156e-05,
+      "loss": 3.7121,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1919353391034775e-05,
+      "loss": 3.7183,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191096744352425e-05,
+      "loss": 3.7278,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190258149601373e-05,
+      "loss": 3.7207,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189419554850321e-05,
+      "loss": 3.7242,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188580960099269e-05,
+      "loss": 3.7154,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187742365348217e-05,
+      "loss": 3.7378,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186905408477538e-05,
+      "loss": 3.7118,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186066813726486e-05,
+      "loss": 3.7269,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185228218975434e-05,
+      "loss": 3.7213,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843896242243817e-05,
+      "loss": 3.7198,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1835526673537026e-05,
+      "loss": 3.7253,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182714072602651e-05,
+      "loss": 3.7229,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181875477851599e-05,
+      "loss": 3.7188,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181036883100547e-05,
+      "loss": 3.7249,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180199926229868e-05,
+      "loss": 3.7095,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179361331478816e-05,
+      "loss": 3.7178,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178522736727764e-05,
+      "loss": 3.7203,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177684141976712e-05,
+      "loss": 3.7266,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17684554722566e-05,
+      "loss": 3.7154,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176008590354981e-05,
+      "loss": 3.7084,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175169995603929e-05,
+      "loss": 3.7169,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174331400852877e-05,
+      "loss": 3.7211,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173492806101825e-05,
+      "loss": 3.7206,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1726558492311466e-05,
+      "loss": 3.7068,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1718172544800946e-05,
+      "loss": 3.7201,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1709786597290426e-05,
+      "loss": 3.7288,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1701400649779906e-05,
+      "loss": 3.717,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1693031081073115e-05,
+      "loss": 3.6994,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1684645133562595e-05,
+      "loss": 3.7159,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1676259186052075e-05,
+      "loss": 3.7114,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1667873238541555e-05,
+      "loss": 3.7243,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1659503669834764e-05,
+      "loss": 3.723,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1651117722324244e-05,
+      "loss": 3.7101,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642731774813724e-05,
+      "loss": 3.7103,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1634345827303204e-05,
+      "loss": 3.7194,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162597625859642e-05,
+      "loss": 3.7024,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16175903110859e-05,
+      "loss": 3.7146,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160920436357538e-05,
+      "loss": 3.7067,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160081841606486e-05,
+      "loss": 3.7246,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159244884735807e-05,
+      "loss": 3.6946,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158406289984755e-05,
+      "loss": 3.7051,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157567695233703e-05,
+      "loss": 3.7082,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156729100482651e-05,
+      "loss": 3.7156,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155890505731599e-05,
+      "loss": 3.7117,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15505354886092e-05,
+      "loss": 3.7035,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154214954109868e-05,
+      "loss": 3.7149,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153376359358816e-05,
+      "loss": 3.7171,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152537764607764e-05,
+      "loss": 3.7047,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151699169856712e-05,
+      "loss": 3.7238,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15086057510566e-05,
+      "loss": 3.7067,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150023618234981e-05,
+      "loss": 3.7134,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149185023483929e-05,
+      "loss": 3.7039,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148346428732877e-05,
+      "loss": 3.7113,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147507833981825e-05,
+      "loss": 3.7027,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146669239230773e-05,
+      "loss": 3.7107,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145832282360094e-05,
+      "loss": 3.7142,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144993687609042e-05,
+      "loss": 3.7109,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14415509285799e-05,
+      "loss": 3.7074,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143316498106938e-05,
+      "loss": 3.709,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142479541236259e-05,
+      "loss": 3.693,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141640946485207e-05,
+      "loss": 3.7214,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140802351734155e-05,
+      "loss": 3.7066,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139963756983103e-05,
+      "loss": 3.7141,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139126800112425e-05,
+      "loss": 3.7029,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138288205361373e-05,
+      "loss": 3.7207,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137449610610321e-05,
+      "loss": 3.7122,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136611015859268e-05,
+      "loss": 3.7056,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357740589885896e-05,
+      "loss": 3.7087,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1349354642375376e-05,
+      "loss": 3.712,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340968694864856e-05,
+      "loss": 3.7128,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133258274735433e-05,
+      "loss": 3.7249,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132419679984381e-05,
+      "loss": 3.6954,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131581085233329e-05,
+      "loss": 3.714,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1307441283626505e-05,
+      "loss": 3.7047,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1299055336115985e-05,
+      "loss": 3.7158,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1290669388605465e-05,
+      "loss": 3.711,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1282283441094945e-05,
+      "loss": 3.7051,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1273897493584425e-05,
+      "loss": 3.716,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1265527924877634e-05,
+      "loss": 3.7053,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1257141977367114e-05,
+      "loss": 3.7114,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9071083068847656,
+      "eval_runtime": 306.1144,
+      "eval_samples_per_second": 1246.563,
+      "eval_steps_per_second": 38.956,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1248756029856594e-05,
+      "loss": 3.7061,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1240370082346074e-05,
+      "loss": 3.6995,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123200051363928e-05,
+      "loss": 3.7099,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122361456612876e-05,
+      "loss": 3.7098,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121522861861824e-05,
+      "loss": 3.7155,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120684267110773e-05,
+      "loss": 3.7013,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119845672359721e-05,
+      "loss": 3.696,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119008715489042e-05,
+      "loss": 3.6967,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11817012073799e-05,
+      "loss": 3.7103,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117331525986938e-05,
+      "loss": 3.7111,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116492931235886e-05,
+      "loss": 3.6981,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115655974365207e-05,
+      "loss": 3.7088,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114817379614155e-05,
+      "loss": 3.6998,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113978784863103e-05,
+      "loss": 3.6972,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113140190112051e-05,
+      "loss": 3.6961,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1123032332413716e-05,
+      "loss": 3.6919,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1114646384903196e-05,
+      "loss": 3.6984,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110626043739268e-05,
+      "loss": 3.6947,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109787448988216e-05,
+      "loss": 3.6928,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108948854237164e-05,
+      "loss": 3.7227,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108111897366485e-05,
+      "loss": 3.7,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107273302615433e-05,
+      "loss": 3.6964,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106434707864381e-05,
+      "loss": 3.6987,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105596113113329e-05,
+      "loss": 3.707,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10475915624265e-05,
+      "loss": 3.6879,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103920561491598e-05,
+      "loss": 3.7006,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103081966740546e-05,
+      "loss": 3.7001,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102243371989494e-05,
+      "loss": 3.6883,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101404777238442e-05,
+      "loss": 3.6892,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005678203677637e-05,
+      "loss": 3.6899,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0997292256167117e-05,
+      "loss": 3.6998,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988906308656597e-05,
+      "loss": 3.7033,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0980520361146076e-05,
+      "loss": 3.6996,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0972134413635556e-05,
+      "loss": 3.6932,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0963748466125036e-05,
+      "loss": 3.6966,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0955362518614516e-05,
+      "loss": 3.7005,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0946992949907725e-05,
+      "loss": 3.6962,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0938623381200935e-05,
+      "loss": 3.6987,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0930237433690414e-05,
+      "loss": 3.6839,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0921851486179894e-05,
+      "loss": 3.6972,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0913465538669374e-05,
+      "loss": 3.6811,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0905079591158854e-05,
+      "loss": 3.7091,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0896693643648334e-05,
+      "loss": 3.6979,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088832407494155e-05,
+      "loss": 3.697,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087993812743103e-05,
+      "loss": 3.7011,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087155217992051e-05,
+      "loss": 3.6951,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086316623240999e-05,
+      "loss": 3.6983,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085478028489946e-05,
+      "loss": 3.6875,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084641071619268e-05,
+      "loss": 3.6684,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083802476868216e-05,
+      "loss": 3.7091,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082963882117163e-05,
+      "loss": 3.6903,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082125287366111e-05,
+      "loss": 3.7037,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081288330495433e-05,
+      "loss": 3.687,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080449735744381e-05,
+      "loss": 3.691,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079611140993329e-05,
+      "loss": 3.6764,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078772546242277e-05,
+      "loss": 3.6987,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0779355893715984e-05,
+      "loss": 3.6945,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0770969946205464e-05,
+      "loss": 3.6958,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076258399869494e-05,
+      "loss": 3.704,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075419805118442e-05,
+      "loss": 3.6836,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074582848247763e-05,
+      "loss": 3.6861,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0737442534967106e-05,
+      "loss": 3.7017,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0729056587456586e-05,
+      "loss": 3.6797,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0720670639946066e-05,
+      "loss": 3.6824,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0712284692435546e-05,
+      "loss": 3.6928,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070391512372876e-05,
+      "loss": 3.695,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069552917621824e-05,
+      "loss": 3.6759,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068714322870772e-05,
+      "loss": 3.6799,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06787572811972e-05,
+      "loss": 3.6767,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067038771249041e-05,
+      "loss": 3.686,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066200176497989e-05,
+      "loss": 3.6946,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065361581746937e-05,
+      "loss": 3.6892,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064522986995885e-05,
+      "loss": 3.6887,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.063686030125206e-05,
+      "loss": 3.6856,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062847435374154e-05,
+      "loss": 3.7038,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062008840623102e-05,
+      "loss": 3.678,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06117024587205e-05,
+      "loss": 3.6999,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060331651120998e-05,
+      "loss": 3.6863,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0594946942503195e-05,
+      "loss": 3.6824,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0586560994992675e-05,
+      "loss": 3.6955,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0578175047482155e-05,
+      "loss": 3.6924,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0569789099971635e-05,
+      "loss": 3.6856,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0561419531264844e-05,
+      "loss": 3.6927,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0553033583754324e-05,
+      "loss": 3.6776,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0544647636243804e-05,
+      "loss": 3.6834,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0536261688733284e-05,
+      "loss": 3.6859,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052789212002649e-05,
+      "loss": 3.6964,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051950617251597e-05,
+      "loss": 3.6823,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051112022500545e-05,
+      "loss": 3.6815,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050273427749493e-05,
+      "loss": 3.678,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049436470878815e-05,
+      "loss": 3.6901,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048597876127763e-05,
+      "loss": 3.6887,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047759281376711e-05,
+      "loss": 3.6739,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046920686625659e-05,
+      "loss": 3.6909,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04608372975498e-05,
+      "loss": 3.6951,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045245135003928e-05,
+      "loss": 3.6826,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044406540252876e-05,
+      "loss": 3.6673,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043567945501824e-05,
+      "loss": 3.6836,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042730988631145e-05,
+      "loss": 3.6777,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041892393880093e-05,
+      "loss": 3.6933,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041053799129041e-05,
+      "loss": 3.6904,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040215204377989e-05,
+      "loss": 3.678,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03937824750731e-05,
+      "loss": 3.6794,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038539652756258e-05,
+      "loss": 3.6819,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037701058005206e-05,
+      "loss": 3.674,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036862463254154e-05,
+      "loss": 3.6811,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036023868503102e-05,
+      "loss": 3.6719,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035186911632423e-05,
+      "loss": 3.6929,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034348316881371e-05,
+      "loss": 3.6638,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033509722130319e-05,
+      "loss": 3.6732,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032671127379267e-05,
+      "loss": 3.6739,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031834170508588e-05,
+      "loss": 3.6848,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030995575757536e-05,
+      "loss": 3.6772,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030156981006484e-05,
+      "loss": 3.6742,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029318386255432e-05,
+      "loss": 3.6845,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284814293847536e-05,
+      "loss": 3.689,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0276428346337016e-05,
+      "loss": 3.6726,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0268042398826496e-05,
+      "loss": 3.6884,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0259656451315976e-05,
+      "loss": 3.6799,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0251286882609185e-05,
+      "loss": 3.6805,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0242900935098665e-05,
+      "loss": 3.6736,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0234514987588145e-05,
+      "loss": 3.6761,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0226129040077625e-05,
+      "loss": 3.6718,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0217759471370834e-05,
+      "loss": 3.6826,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0209373523860314e-05,
+      "loss": 3.6841,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0200987576349794e-05,
+      "loss": 3.6786,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0192601628839274e-05,
+      "loss": 3.677,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0184215681328754e-05,
+      "loss": 3.6762,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017584611262197e-05,
+      "loss": 3.6634,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016746016511145e-05,
+      "loss": 3.6925,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015907421760093e-05,
+      "loss": 3.6732,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015068827009041e-05,
+      "loss": 3.6822,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014231870138362e-05,
+      "loss": 3.6715,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01339327538731e-05,
+      "loss": 3.6887,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012554680636258e-05,
+      "loss": 3.6824,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011716085885206e-05,
+      "loss": 3.6739,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010879129014527e-05,
+      "loss": 3.68,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010040534263475e-05,
+      "loss": 3.6768,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009201939512423e-05,
+      "loss": 3.6806,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008363344761371e-05,
+      "loss": 3.6996,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007524750010319e-05,
+      "loss": 3.6626,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0066877931396404e-05,
+      "loss": 3.6872,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0058491983885883e-05,
+      "loss": 3.6725,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0050106036375363e-05,
+      "loss": 3.6834,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0041720088864837e-05,
+      "loss": 3.6764,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003335052015805e-05,
+      "loss": 3.6785,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002496457264753e-05,
+      "loss": 3.6838,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0016578625137006e-05,
+      "loss": 3.6763,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0008192677626486e-05,
+      "loss": 3.678,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.901350259780884,
+      "eval_runtime": 305.4695,
+      "eval_samples_per_second": 1249.195,
+      "eval_steps_per_second": 39.038,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9999806730115966e-05,
+      "loss": 3.6756,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9991420782605445e-05,
+      "loss": 3.6686,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998303483509493e-05,
+      "loss": 3.6803,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997464888758441e-05,
+      "loss": 3.6773,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996626294007389e-05,
+      "loss": 3.6856,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99578933713671e-05,
+      "loss": 3.6708,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994950742385658e-05,
+      "loss": 3.6699,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994112147634606e-05,
+      "loss": 3.6715,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993273552883554e-05,
+      "loss": 3.6771,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992436596012875e-05,
+      "loss": 3.6796,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991598001261823e-05,
+      "loss": 3.6678,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990759406510771e-05,
+      "loss": 3.676,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989920811759719e-05,
+      "loss": 3.6734,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98908385488904e-05,
+      "loss": 3.6631,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9882452601379886e-05,
+      "loss": 3.6717,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9874066653869366e-05,
+      "loss": 3.6594,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865680706358846e-05,
+      "loss": 3.6673,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9857311137652055e-05,
+      "loss": 3.6625,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848925190141535e-05,
+      "loss": 3.6638,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9840539242631015e-05,
+      "loss": 3.6894,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9832153295120495e-05,
+      "loss": 3.6741,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823783726413704e-05,
+      "loss": 3.6695,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9815397778903184e-05,
+      "loss": 3.6681,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9807011831392664e-05,
+      "loss": 3.6765,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9798625883882144e-05,
+      "loss": 3.6532,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979025631517535e-05,
+      "loss": 3.6727,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978187036766484e-05,
+      "loss": 3.6726,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977348442015432e-05,
+      "loss": 3.6566,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97650984726438e-05,
+      "loss": 3.6568,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975672890393701e-05,
+      "loss": 3.6621,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974834295642649e-05,
+      "loss": 3.6686,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973995700891597e-05,
+      "loss": 3.6734,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973157106140545e-05,
+      "loss": 3.6665,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972320149269866e-05,
+      "loss": 3.6666,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971481554518814e-05,
+      "loss": 3.6695,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970642959767762e-05,
+      "loss": 3.6689,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96980436501671e-05,
+      "loss": 3.6687,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9689674081460306e-05,
+      "loss": 3.6688,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968128813394979e-05,
+      "loss": 3.6531,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967290218643927e-05,
+      "loss": 3.6675,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966451623892875e-05,
+      "loss": 3.6553,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965613029141823e-05,
+      "loss": 3.678,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964776072271144e-05,
+      "loss": 3.6707,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963937477520092e-05,
+      "loss": 3.6661,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96309888276904e-05,
+      "loss": 3.671,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962261925898361e-05,
+      "loss": 3.6668,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.961423331147309e-05,
+      "loss": 3.6722,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960584736396257e-05,
+      "loss": 3.6587,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959746141645205e-05,
+      "loss": 3.6357,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958909184774526e-05,
+      "loss": 3.6778,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958070590023475e-05,
+      "loss": 3.6608,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957231995272423e-05,
+      "loss": 3.6761,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956393400521371e-05,
+      "loss": 3.6577,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955554805770319e-05,
+      "loss": 3.6594,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9547178488996396e-05,
+      "loss": 3.6505,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9538792541485876e-05,
+      "loss": 3.667,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9530406593975356e-05,
+      "loss": 3.6696,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9522020646464836e-05,
+      "loss": 3.6664,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9513651077758045e-05,
+      "loss": 3.6757,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9505265130247525e-05,
+      "loss": 3.6562,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9496879182737005e-05,
+      "loss": 3.6534,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9488493235226485e-05,
+      "loss": 3.6738,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9480107287715965e-05,
+      "loss": 3.6532,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947173771900918e-05,
+      "loss": 3.6534,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946335177149866e-05,
+      "loss": 3.6623,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945496582398814e-05,
+      "loss": 3.6688,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944657987647762e-05,
+      "loss": 3.6469,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943821030777083e-05,
+      "loss": 3.6551,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942982436026031e-05,
+      "loss": 3.6483,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942143841274979e-05,
+      "loss": 3.6559,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941305246523926e-05,
+      "loss": 3.6663,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940466651772874e-05,
+      "loss": 3.6578,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939629694902196e-05,
+      "loss": 3.6596,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938791100151144e-05,
+      "loss": 3.6585,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937952505400092e-05,
+      "loss": 3.6768,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93711391064904e-05,
+      "loss": 3.6506,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9362769537783614e-05,
+      "loss": 3.6682,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935438359027309e-05,
+      "loss": 3.6585,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934599764276257e-05,
+      "loss": 3.655,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933761169525205e-05,
+      "loss": 3.6673,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932924212654526e-05,
+      "loss": 3.6638,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9320856179034736e-05,
+      "loss": 3.6614,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9312470231524216e-05,
+      "loss": 3.6602,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9304084284013696e-05,
+      "loss": 3.6544,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929571471530691e-05,
+      "loss": 3.652,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9287328767796385e-05,
+      "loss": 3.6601,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927894282028587e-05,
+      "loss": 3.6657,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927055687277535e-05,
+      "loss": 3.6517,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926218730406856e-05,
+      "loss": 3.6532,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925380135655804e-05,
+      "loss": 3.6509,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924541540904752e-05,
+      "loss": 3.6652,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9237029461537e-05,
+      "loss": 3.6594,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922865989283021e-05,
+      "loss": 3.6464,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922027394531969e-05,
+      "loss": 3.6602,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921188799780917e-05,
+      "loss": 3.6672,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920350205029865e-05,
+      "loss": 3.6567,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919513248159186e-05,
+      "loss": 3.6396,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918674653408134e-05,
+      "loss": 3.6536,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9178360586570826e-05,
+      "loss": 3.6496,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9169974639060306e-05,
+      "loss": 3.6652,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9161605070353515e-05,
+      "loss": 3.6617,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9153219122842995e-05,
+      "loss": 3.6585,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9144833175332475e-05,
+      "loss": 3.6513,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9136447227821955e-05,
+      "loss": 3.6527,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9128077659115164e-05,
+      "loss": 3.6491,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9119691711604644e-05,
+      "loss": 3.6511,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9111305764094124e-05,
+      "loss": 3.6455,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9102919816583604e-05,
+      "loss": 3.6646,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909455024787681e-05,
+      "loss": 3.641,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908616430036629e-05,
+      "loss": 3.6412,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907777835285578e-05,
+      "loss": 3.6499,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906939240534526e-05,
+      "loss": 3.6512,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906100645783474e-05,
+      "loss": 3.654,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905263688912795e-05,
+      "loss": 3.6469,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904425094161743e-05,
+      "loss": 3.6549,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903586499410691e-05,
+      "loss": 3.6641,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902747904659639e-05,
+      "loss": 3.6443,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90191094778896e-05,
+      "loss": 3.6577,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901072353037908e-05,
+      "loss": 3.6528,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900233758286856e-05,
+      "loss": 3.6549,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899395163535804e-05,
+      "loss": 3.648,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8985582066651246e-05,
+      "loss": 3.6466,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897719611914073e-05,
+      "loss": 3.6464,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896881017163021e-05,
+      "loss": 3.6543,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896042422411969e-05,
+      "loss": 3.6557,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89520546554129e-05,
+      "loss": 3.65,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894366870790238e-05,
+      "loss": 3.6506,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893528276039186e-05,
+      "loss": 3.6487,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892689681288134e-05,
+      "loss": 3.6366,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891852724417455e-05,
+      "loss": 3.6629,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891014129666403e-05,
+      "loss": 3.6442,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890175534915351e-05,
+      "loss": 3.6594,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889336940164299e-05,
+      "loss": 3.6416,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88849998329362e-05,
+      "loss": 3.6617,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887661388542569e-05,
+      "loss": 3.6543,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8868227937915167e-05,
+      "loss": 3.6474,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8859841990404647e-05,
+      "loss": 3.6503,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8851472421697856e-05,
+      "loss": 3.6512,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8843086474187336e-05,
+      "loss": 3.6551,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8834700526676816e-05,
+      "loss": 3.67,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8826314579166295e-05,
+      "loss": 3.6361,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8817945010459505e-05,
+      "loss": 3.6632,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8809559062948985e-05,
+      "loss": 3.6427,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8801173115438465e-05,
+      "loss": 3.6559,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8792787167927944e-05,
+      "loss": 3.6526,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8784417599221154e-05,
+      "loss": 3.6512,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877603165171064e-05,
+      "loss": 3.6557,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876764570420012e-05,
+      "loss": 3.6527,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.87592597566896e-05,
+      "loss": 3.6505,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8975584506988525,
+      "eval_runtime": 306.3478,
+      "eval_samples_per_second": 1245.614,
+      "eval_steps_per_second": 38.926,
+      "step": 686880
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 4.4566382516737536e+17,
+  "trial_name": null,
+  "trial_params": null
+}