diff --git "a/checkpoint-992160/trainer_state.json" "b/checkpoint-992160/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-992160/trainer_state.json"
@@ -0,0 +1,11751 @@
+{
+  "best_metric": 3.8959591388702393,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/pp-mod-subj2/transformer/0/checkpoints/checkpoint-915840",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 992160,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.9871,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8358,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1921,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.982,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8104,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7028,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.606,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5354,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4529,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.3982,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3445,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3051,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.2668,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2105,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1789,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1263,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1126,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.075,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0389,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0174,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0173,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9713,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815558291179756e-05,
+      "loss": 4.9567,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807172343669236e-05,
+      "loss": 4.933,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798786396158716e-05,
+      "loss": 4.9086,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.8983,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8628,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8611,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8439,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8151,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.8083,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7869,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7878,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 4.768,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 4.7508,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7465,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969820416244367e-05,
+      "loss": 4.7486,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.7096,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7042,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6977,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6787,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6718,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6737,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6612,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 4.648,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.639,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.63,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6284,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.5989,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95892359812181e-05,
+      "loss": 4.5984,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958085003370758e-05,
+      "loss": 4.6072,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957246408619706e-05,
+      "loss": 4.5963,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 4.5829,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5653,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5725,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5516,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5714,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 4.5394,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5529,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.533,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5206,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.524,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5103,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.5069,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4988,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5077,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 4.4839,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4794,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.466,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9421582546222625e-05,
+      "loss": 4.4656,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4834,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4741,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4553,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4549,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4593,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4533,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.4374,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.4402,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4317,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.4305,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.4249,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.421,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4253,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 4.4175,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.3952,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 4.4068,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 4.4053,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 4.3994,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.3934,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925391273242342e-05,
+      "loss": 4.3817,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 4.3895,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 4.3783,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922878764749932e-05,
+      "loss": 4.3841,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 4.3721,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.3666,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.3862,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.351,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.3497,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.3587,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.357,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.362,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.3462,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3397,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.3386,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.3421,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.3219,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.322,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.3337,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.3233,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.3163,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.3103,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3096,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3194,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3055,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.3041,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.3083,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.3068,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901920447495124e-05,
+      "loss": 4.2968,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901081852744072e-05,
+      "loss": 4.2983,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.2975,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.2957,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.28,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.2915,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968888789888125e-05,
+      "loss": 4.2849,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.2893,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.2728,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.2818,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.2706,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.2732,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918605862432463e-05,
+      "loss": 4.2537,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.2823,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.2786,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.2623,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.2518,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.2722,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.2584,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885992060866256e-05,
+      "loss": 4.2606,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885153466115204e-05,
+      "loss": 4.2528,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884314871364151e-05,
+      "loss": 4.26,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883476276613099e-05,
+      "loss": 4.2508,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.2487,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2363,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.2494,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.2441,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.2364,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2534,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877611026996855e-05,
+      "loss": 4.2301,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876772432245803e-05,
+      "loss": 4.2369,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875933837494751e-05,
+      "loss": 4.2425,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8750952427436986e-05,
+      "loss": 4.23,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.243849754333496,
+      "eval_runtime": 311.8262,
+      "eval_samples_per_second": 1223.73,
+      "eval_steps_per_second": 38.242,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8742566479926466e-05,
+      "loss": 4.2151,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.2223,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872581096370916e-05,
+      "loss": 4.2317,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717425016198635e-05,
+      "loss": 4.2216,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709039068688115e-05,
+      "loss": 4.2224,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700653121177595e-05,
+      "loss": 4.2078,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869228355247081e-05,
+      "loss": 4.2091,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.868389760496029e-05,
+      "loss": 4.2044,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675528036253507e-05,
+      "loss": 4.215,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667142088742986e-05,
+      "loss": 4.2,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865875614123246e-05,
+      "loss": 4.2081,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650386572525676e-05,
+      "loss": 4.2115,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642000625015155e-05,
+      "loss": 4.1953,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633614677504635e-05,
+      "loss": 4.1934,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862522872999411e-05,
+      "loss": 4.1806,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861684278248359e-05,
+      "loss": 4.1899,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.186,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860007088746255e-05,
+      "loss": 4.182,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859168493995203e-05,
+      "loss": 4.1814,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.2072,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.1774,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566543476224204e-05,
+      "loss": 4.1802,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558157528713684e-05,
+      "loss": 4.1861,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549771581203164e-05,
+      "loss": 4.1783,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.1713,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532999686182124e-05,
+      "loss": 4.1653,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524613738671604e-05,
+      "loss": 4.1717,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851624416996481e-05,
+      "loss": 4.1688,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.1547,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.1668,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.1549,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.1687,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847431443241221e-05,
+      "loss": 4.1643,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846592848490169e-05,
+      "loss": 4.1497,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.1636,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844918934748812e-05,
+      "loss": 4.1715,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84408033999776e-05,
+      "loss": 4.1408,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.153,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842403150495656e-05,
+      "loss": 4.1484,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841564555744604e-05,
+      "loss": 4.1378,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840725960993552e-05,
+      "loss": 4.1436,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398873662425e-05,
+      "loss": 4.1529,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839048771491447e-05,
+      "loss": 4.1469,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382118146207687e-05,
+      "loss": 4.14,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373732198697167e-05,
+      "loss": 4.1466,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365346251186646e-05,
+      "loss": 4.1362,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8356960303676126e-05,
+      "loss": 4.1424,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834859073496934e-05,
+      "loss": 4.1227,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834022116626255e-05,
+      "loss": 4.1245,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833183521875203e-05,
+      "loss": 4.141,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832344927124151e-05,
+      "loss": 4.1388,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831506332373099e-05,
+      "loss": 4.1261,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830667737622047e-05,
+      "loss": 4.119,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829830780751368e-05,
+      "loss": 4.1316,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828992186000316e-05,
+      "loss": 4.1127,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828153591249264e-05,
+      "loss": 4.1376,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827314996498212e-05,
+      "loss": 4.1192,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82647640174716e-05,
+      "loss": 4.1351,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256394448764816e-05,
+      "loss": 4.119,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248008501254296e-05,
+      "loss": 4.1113,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823962255374377e-05,
+      "loss": 4.114,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823123660623325e-05,
+      "loss": 4.1186,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822285065872273e-05,
+      "loss": 4.1051,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214481090015945e-05,
+      "loss": 4.1045,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820609514250542e-05,
+      "loss": 4.1198,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81977091949949e-05,
+      "loss": 4.1035,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818932324748438e-05,
+      "loss": 4.0977,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180953678777594e-05,
+      "loss": 4.0955,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172567731267074e-05,
+      "loss": 4.097,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164181783756554e-05,
+      "loss": 4.1152,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155795836246034e-05,
+      "loss": 4.1098,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147409888735514e-05,
+      "loss": 4.0965,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8139023941224994e-05,
+      "loss": 4.1001,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8130637993714474e-05,
+      "loss": 4.105,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812226842500768e-05,
+      "loss": 4.101,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811388247749716e-05,
+      "loss": 4.094,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810549652998664e-05,
+      "loss": 4.0993,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809711058247612e-05,
+      "loss": 4.0934,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80887246349656e-05,
+      "loss": 4.0983,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808035506625881e-05,
+      "loss": 4.0892,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807196911874829e-05,
+      "loss": 4.0933,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806358317123777e-05,
+      "loss": 4.095,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805519722372726e-05,
+      "loss": 4.0934,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804682765502047e-05,
+      "loss": 4.074,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803844170750995e-05,
+      "loss": 4.088,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803005575999943e-05,
+      "loss": 4.0871,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802166981248891e-05,
+      "loss": 4.0868,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801328386497839e-05,
+      "loss": 4.0842,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800489791746787e-05,
+      "loss": 4.0715,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996528348761076e-05,
+      "loss": 4.0831,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988142401250556e-05,
+      "loss": 4.075,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979756453740036e-05,
+      "loss": 4.085,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971370506229516e-05,
+      "loss": 4.0701,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7963000937522725e-05,
+      "loss": 4.0735,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795461499001221e-05,
+      "loss": 4.0845,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794622904250169e-05,
+      "loss": 4.0725,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793784309499117e-05,
+      "loss": 4.0559,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792947352628438e-05,
+      "loss": 4.0717,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792108757877386e-05,
+      "loss": 4.0731,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791270163126334e-05,
+      "loss": 4.0771,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790431568375282e-05,
+      "loss": 4.0698,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78959297362423e-05,
+      "loss": 4.0639,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788756016753551e-05,
+      "loss": 4.06,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787917422002499e-05,
+      "loss": 4.0701,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787078827251447e-05,
+      "loss": 4.0489,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786241870380768e-05,
+      "loss": 4.0561,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854032756297166e-05,
+      "loss": 4.0637,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845646808786646e-05,
+      "loss": 4.0628,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837260861276126e-05,
+      "loss": 4.0431,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828874913765606e-05,
+      "loss": 4.0567,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782048896625508e-05,
+      "loss": 4.0455,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781210301874456e-05,
+      "loss": 4.0639,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780371707123404e-05,
+      "loss": 4.0506,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779534750252725e-05,
+      "loss": 4.0432,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778696155501673e-05,
+      "loss": 4.0615,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777857560750621e-05,
+      "loss": 4.0554,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770206038799423e-05,
+      "loss": 4.0466,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761820091288903e-05,
+      "loss": 4.0522,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7753434143778383e-05,
+      "loss": 4.0486,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774504819626786e-05,
+      "loss": 4.0519,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773666224875734e-05,
+      "loss": 4.032,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772827630124682e-05,
+      "loss": 4.0534,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77198903537363e-05,
+      "loss": 4.0415,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771150440622578e-05,
+      "loss": 4.0495,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770313483751899e-05,
+      "loss": 4.0409,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769474889000847e-05,
+      "loss": 4.0441,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768636294249795e-05,
+      "loss": 4.0374,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767797699498743e-05,
+      "loss": 4.0393,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766960742628064e-05,
+      "loss": 4.0226,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766122147877012e-05,
+      "loss": 4.0552,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76528355312596e-05,
+      "loss": 4.0448,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764444958374909e-05,
+      "loss": 4.0372,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76360800150423e-05,
+      "loss": 4.0277,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762769406753178e-05,
+      "loss": 4.0455,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761930812002126e-05,
+      "loss": 4.0421,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761092217251074e-05,
+      "loss": 4.0339,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760253622500022e-05,
+      "loss": 4.0299,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7594166656293426e-05,
+      "loss": 4.0384,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7585780708782906e-05,
+      "loss": 4.0326,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577394761272386e-05,
+      "loss": 4.0412,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569008813761866e-05,
+      "loss": 4.0186,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560639245055075e-05,
+      "loss": 4.0356,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552253297544555e-05,
+      "loss": 4.0305,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543867350034035e-05,
+      "loss": 4.0249,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753548140252352e-05,
+      "loss": 4.0429,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752711183381673e-05,
+      "loss": 4.0228,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751872588630621e-05,
+      "loss": 4.0277,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751033993879569e-05,
+      "loss": 4.03,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750195399128517e-05,
+      "loss": 4.0263,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.07372522354126,
+      "eval_runtime": 302.1401,
+      "eval_samples_per_second": 1262.96,
+      "eval_steps_per_second": 39.468,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749356804377465e-05,
+      "loss": 4.0124,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748518209626413e-05,
+      "loss": 4.0208,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747679614875361e-05,
+      "loss": 4.0292,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468410201243084e-05,
+      "loss": 4.0219,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460024253732563e-05,
+      "loss": 4.0269,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451638306222043e-05,
+      "loss": 4.0063,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744325235871152e-05,
+      "loss": 4.0127,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434866411201e-05,
+      "loss": 4.0101,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742649684249422e-05,
+      "loss": 4.0213,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74181108949837e-05,
+      "loss": 4.0084,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740972494747318e-05,
+      "loss": 4.0162,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740133899996266e-05,
+      "loss": 4.0195,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739296943125587e-05,
+      "loss": 4.0024,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738458348374535e-05,
+      "loss": 4.0077,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737619753623483e-05,
+      "loss": 3.991,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736781158872431e-05,
+      "loss": 4.0027,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735942564121379e-05,
+      "loss": 4.0019,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735103969370327e-05,
+      "loss": 3.9987,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734265374619275e-05,
+      "loss": 3.9978,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733426779868223e-05,
+      "loss": 4.0235,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732588185117171e-05,
+      "loss": 3.9992,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731749590366119e-05,
+      "loss": 3.9996,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7309109956150674e-05,
+      "loss": 4.0068,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300740387443884e-05,
+      "loss": 4.0022,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292354439933364e-05,
+      "loss": 3.9933,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7283968492422844e-05,
+      "loss": 3.9888,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275582544912323e-05,
+      "loss": 3.9993,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726721297620553e-05,
+      "loss": 3.9888,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725882702869501e-05,
+      "loss": 3.9831,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725044108118449e-05,
+      "loss": 3.9961,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724205513367397e-05,
+      "loss": 3.9845,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723368556496718e-05,
+      "loss": 3.9975,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722529961745666e-05,
+      "loss": 3.9916,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721691366994614e-05,
+      "loss": 3.9813,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720852772243563e-05,
+      "loss": 3.9944,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720015815372884e-05,
+      "loss": 4.0059,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719177220621832e-05,
+      "loss": 3.9731,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71833862587078e-05,
+      "loss": 3.9897,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717500031119727e-05,
+      "loss": 3.9778,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7166630742490486e-05,
+      "loss": 3.9798,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158244794979966e-05,
+      "loss": 3.9782,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149858847469446e-05,
+      "loss": 3.9932,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714147289995892e-05,
+      "loss": 3.9853,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133103331252135e-05,
+      "loss": 3.9801,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124717383741615e-05,
+      "loss": 3.9848,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116331436231095e-05,
+      "loss": 3.9754,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7107945488720575e-05,
+      "loss": 3.9868,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709957592001379e-05,
+      "loss": 3.9673,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709118997250327e-05,
+      "loss": 3.969,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708282040379648e-05,
+      "loss": 3.9836,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707443445628596e-05,
+      "loss": 3.9805,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706604850877544e-05,
+      "loss": 3.9721,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705766256126492e-05,
+      "loss": 3.9693,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704927661375439e-05,
+      "loss": 3.9788,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704089066624387e-05,
+      "loss": 3.958,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703250471873335e-05,
+      "loss": 3.9824,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702413515002657e-05,
+      "loss": 3.9705,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701574920251605e-05,
+      "loss": 3.9828,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700736325500553e-05,
+      "loss": 3.9743,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699897730749501e-05,
+      "loss": 3.9621,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699060773878822e-05,
+      "loss": 3.9642,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69822217912777e-05,
+      "loss": 3.9711,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697383584376718e-05,
+      "loss": 3.9564,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696544989625666e-05,
+      "loss": 3.9589,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695708032754987e-05,
+      "loss": 3.9716,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694869438003935e-05,
+      "loss": 3.9605,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694030843252883e-05,
+      "loss": 3.9502,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931922485018307e-05,
+      "loss": 3.9529,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6923536537507787e-05,
+      "loss": 3.9515,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915150589997267e-05,
+      "loss": 3.9731,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690676464248675e-05,
+      "loss": 3.9657,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689837869497623e-05,
+      "loss": 3.9555,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689000912626944e-05,
+      "loss": 3.9629,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688162317875892e-05,
+      "loss": 3.9597,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68732372312484e-05,
+      "loss": 3.9637,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686485128373788e-05,
+      "loss": 3.9536,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685648171503109e-05,
+      "loss": 3.9646,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684809576752057e-05,
+      "loss": 3.959,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683970982001005e-05,
+      "loss": 3.9565,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683134025130326e-05,
+      "loss": 3.9531,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682295430379274e-05,
+      "loss": 3.9579,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681456835628222e-05,
+      "loss": 3.9589,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680618240877171e-05,
+      "loss": 3.9576,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679779646126119e-05,
+      "loss": 3.9432,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678941051375067e-05,
+      "loss": 3.9519,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678102456624015e-05,
+      "loss": 3.9523,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677263861872963e-05,
+      "loss": 3.951,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764269050022836e-05,
+      "loss": 3.9547,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755883102512316e-05,
+      "loss": 3.9382,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747497155001796e-05,
+      "loss": 3.9502,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739111207491276e-05,
+      "loss": 3.9465,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730741638784485e-05,
+      "loss": 3.9557,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722355691273965e-05,
+      "loss": 3.9374,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6713986122567174e-05,
+      "loss": 3.9463,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670560017505666e-05,
+      "loss": 3.9589,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669721422754614e-05,
+      "loss": 3.943,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668882828003562e-05,
+      "loss": 3.9281,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66804423325251e-05,
+      "loss": 3.9417,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667205638501458e-05,
+      "loss": 3.9461,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663670437504054e-05,
+      "loss": 3.9475,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655284489993534e-05,
+      "loss": 3.9448,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664691492128675e-05,
+      "loss": 3.9347,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663852897377623e-05,
+      "loss": 3.9339,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66301430262657e-05,
+      "loss": 3.9455,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662175707875518e-05,
+      "loss": 3.9271,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66133875100484e-05,
+      "loss": 3.9323,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660500156253788e-05,
+      "loss": 3.9412,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659661561502736e-05,
+      "loss": 3.94,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588246046320574e-05,
+      "loss": 3.9178,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579860098810054e-05,
+      "loss": 3.9358,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657147415129953e-05,
+      "loss": 3.9235,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656308820378901e-05,
+      "loss": 3.9382,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655470225627849e-05,
+      "loss": 3.9337,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654631630876797e-05,
+      "loss": 3.9254,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653793036125745e-05,
+      "loss": 3.9403,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652954441374693e-05,
+      "loss": 3.9376,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652119122384387e-05,
+      "loss": 3.9248,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651280527633335e-05,
+      "loss": 3.933,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650441932882283e-05,
+      "loss": 3.9317,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649603338131231e-05,
+      "loss": 3.9316,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648764743380179e-05,
+      "loss": 3.9187,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647926148629127e-05,
+      "loss": 3.9331,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647087553878075e-05,
+      "loss": 3.9269,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646250597007396e-05,
+      "loss": 3.9314,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645412002256344e-05,
+      "loss": 3.9265,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644573407505292e-05,
+      "loss": 3.931,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64373481275424e-05,
+      "loss": 3.919,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642897855883561e-05,
+      "loss": 3.9243,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642059261132509e-05,
+      "loss": 3.9104,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641220666381457e-05,
+      "loss": 3.9398,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640382071630405e-05,
+      "loss": 3.9288,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395434768793536e-05,
+      "loss": 3.9253,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387048821283016e-05,
+      "loss": 3.9163,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378662873772496e-05,
+      "loss": 3.9308,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370276926261976e-05,
+      "loss": 3.9307,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361907357555185e-05,
+      "loss": 3.9204,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353521410044665e-05,
+      "loss": 3.9191,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345135462534145e-05,
+      "loss": 3.9255,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336749515023625e-05,
+      "loss": 3.9235,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328379946316834e-05,
+      "loss": 3.9309,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6319993998806314e-05,
+      "loss": 3.9125,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311608051295794e-05,
+      "loss": 3.9234,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303222103785274e-05,
+      "loss": 3.9185,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629485253507848e-05,
+      "loss": 3.9139,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628646658756797e-05,
+      "loss": 3.9362,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627808064005745e-05,
+      "loss": 3.9134,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626969469254693e-05,
+      "loss": 3.9192,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626132512384014e-05,
+      "loss": 3.9227,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625293917632962e-05,
+      "loss": 3.9157,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.000222206115723,
+      "eval_runtime": 309.5243,
+      "eval_samples_per_second": 1232.83,
+      "eval_steps_per_second": 38.527,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624456960762283e-05,
+      "loss": 3.9122,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623618366011231e-05,
+      "loss": 3.9115,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622779771260179e-05,
+      "loss": 3.9219,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621941176509127e-05,
+      "loss": 3.9171,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621102581758075e-05,
+      "loss": 3.9221,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620263987007023e-05,
+      "loss": 3.9028,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619425392255971e-05,
+      "loss": 3.9114,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185884353852924e-05,
+      "loss": 3.8994,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617751478514613e-05,
+      "loss": 3.921,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616912883763561e-05,
+      "loss": 3.9025,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616074289012509e-05,
+      "loss": 3.9136,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615235694261457e-05,
+      "loss": 3.9156,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614398737390778e-05,
+      "loss": 3.9039,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613560142639726e-05,
+      "loss": 3.8977,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612721547888674e-05,
+      "loss": 3.8959,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611882953137622e-05,
+      "loss": 3.8947,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61104435838657e-05,
+      "loss": 3.901,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610205763635518e-05,
+      "loss": 3.8983,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609367168884466e-05,
+      "loss": 3.8982,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608528574133414e-05,
+      "loss": 3.9197,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607691617262736e-05,
+      "loss": 3.9013,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606853022511684e-05,
+      "loss": 3.8984,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606014427760631e-05,
+      "loss": 3.907,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605175833009579e-05,
+      "loss": 3.9038,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604337238258527e-05,
+      "loss": 3.8934,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603498643507475e-05,
+      "loss": 3.8899,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602660048756423e-05,
+      "loss": 3.9047,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601821454005371e-05,
+      "loss": 3.8851,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600984497134692e-05,
+      "loss": 3.8893,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.60014590238364e-05,
+      "loss": 3.8929,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599307307632588e-05,
+      "loss": 3.8884,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598468712881536e-05,
+      "loss": 3.9036,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976317560108575e-05,
+      "loss": 3.8939,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967931612598055e-05,
+      "loss": 3.8867,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959545665087535e-05,
+      "loss": 3.8986,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5951159717577015e-05,
+      "loss": 3.9078,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5942790148870224e-05,
+      "loss": 3.8835,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5934404201359704e-05,
+      "loss": 3.8936,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5926018253849184e-05,
+      "loss": 3.878,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5917632306338664e-05,
+      "loss": 3.894,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5909246358828144e-05,
+      "loss": 3.8802,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5900860411317624e-05,
+      "loss": 3.9016,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5892474463807104e-05,
+      "loss": 3.8907,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588410489510031e-05,
+      "loss": 3.8856,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58757189475898e-05,
+      "loss": 3.8924,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586733300007928e-05,
+      "loss": 3.8825,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585894705256876e-05,
+      "loss": 3.8928,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585057748386197e-05,
+      "loss": 3.8776,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584219153635145e-05,
+      "loss": 3.8755,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583380558884093e-05,
+      "loss": 3.8921,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582541964133041e-05,
+      "loss": 3.8865,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581703369381989e-05,
+      "loss": 3.8853,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580864774630937e-05,
+      "loss": 3.8765,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580027817760258e-05,
+      "loss": 3.885,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579189223009206e-05,
+      "loss": 3.8702,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578350628258154e-05,
+      "loss": 3.8905,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577512033507102e-05,
+      "loss": 3.8815,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57667343875605e-05,
+      "loss": 3.8952,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575834844004998e-05,
+      "loss": 3.8853,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574996249253946e-05,
+      "loss": 3.8761,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574157654502894e-05,
+      "loss": 3.8717,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573322335512588e-05,
+      "loss": 3.882,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572483740761536e-05,
+      "loss": 3.8698,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571645146010484e-05,
+      "loss": 3.8687,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5708065512594315e-05,
+      "loss": 3.881,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699679565083795e-05,
+      "loss": 3.8783,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569130999637701e-05,
+      "loss": 3.8616,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568292404886649e-05,
+      "loss": 3.8657,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567453810135597e-05,
+      "loss": 3.8623,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566615215384545e-05,
+      "loss": 3.8884,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565776620633493e-05,
+      "loss": 3.8763,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564938025882441e-05,
+      "loss": 3.8707,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564101069011762e-05,
+      "loss": 3.8772,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56326247426071e-05,
+      "loss": 3.8691,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562423879509658e-05,
+      "loss": 3.8829,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561585284758606e-05,
+      "loss": 3.863,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560748327887927e-05,
+      "loss": 3.8848,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559909733136875e-05,
+      "loss": 3.868,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559071138385823e-05,
+      "loss": 3.873,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558232543634771e-05,
+      "loss": 3.8739,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557393948883719e-05,
+      "loss": 3.8687,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556555354132667e-05,
+      "loss": 3.8795,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557183972619885e-05,
+      "loss": 3.8691,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548798025109365e-05,
+      "loss": 3.8597,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5540412077598845e-05,
+      "loss": 3.8703,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5532026130088325e-05,
+      "loss": 3.861,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523640182577804e-05,
+      "loss": 3.8733,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5515254235067284e-05,
+      "loss": 3.8684,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5506868287556764e-05,
+      "loss": 3.8561,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498498718849973e-05,
+      "loss": 3.8695,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5490112771339453e-05,
+      "loss": 3.8609,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481726823828933e-05,
+      "loss": 3.8698,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547334087631841e-05,
+      "loss": 3.8579,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546497130761162e-05,
+      "loss": 3.8675,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545658536010111e-05,
+      "loss": 3.8771,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544819941259059e-05,
+      "loss": 3.8607,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543981346508007e-05,
+      "loss": 3.8507,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543142751756955e-05,
+      "loss": 3.8614,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542304157005903e-05,
+      "loss": 3.8583,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541467200135224e-05,
+      "loss": 3.8711,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540628605384172e-05,
+      "loss": 3.862,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53979001063312e-05,
+      "loss": 3.856,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538951415882068e-05,
+      "loss": 3.857,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538112821131015e-05,
+      "loss": 3.8669,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537274226379963e-05,
+      "loss": 3.8447,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536435631628911e-05,
+      "loss": 3.8571,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535597036877859e-05,
+      "loss": 3.8596,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534760080007181e-05,
+      "loss": 3.8617,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533923123136502e-05,
+      "loss": 3.8345,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53308452838545e-05,
+      "loss": 3.8563,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5322459336343976e-05,
+      "loss": 3.8523,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314073388833456e-05,
+      "loss": 3.852,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305687441322936e-05,
+      "loss": 3.8587,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297301493812416e-05,
+      "loss": 3.845,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5288915546301896e-05,
+      "loss": 3.8646,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5280529598791376e-05,
+      "loss": 3.8589,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527217640888832e-05,
+      "loss": 3.8486,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52637904613778e-05,
+      "loss": 3.8573,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525540451386728e-05,
+      "loss": 3.8517,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524701856635676e-05,
+      "loss": 3.8575,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523863261884624e-05,
+      "loss": 3.8385,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523026305013945e-05,
+      "loss": 3.861,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522187710262893e-05,
+      "loss": 3.8483,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521349115511841e-05,
+      "loss": 3.8543,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520510520760789e-05,
+      "loss": 3.8525,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519671926009737e-05,
+      "loss": 3.8581,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518833331258685e-05,
+      "loss": 3.8482,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517996374388006e-05,
+      "loss": 3.8438,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517157779636954e-05,
+      "loss": 3.8381,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516319184885902e-05,
+      "loss": 3.8608,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51548059013485e-05,
+      "loss": 3.8528,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146436332641714e-05,
+      "loss": 3.8513,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5138050385131194e-05,
+      "loss": 3.8421,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129664437620674e-05,
+      "loss": 3.8559,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121278490110154e-05,
+      "loss": 3.8531,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112892542599634e-05,
+      "loss": 3.8448,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104506595089114e-05,
+      "loss": 3.8438,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509613702638232e-05,
+      "loss": 3.8509,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50877510788718e-05,
+      "loss": 3.8512,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507936513136128e-05,
+      "loss": 3.8542,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507097918385076e-05,
+      "loss": 3.8386,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506259323634024e-05,
+      "loss": 3.8509,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505420728882972e-05,
+      "loss": 3.8429,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504583772012293e-05,
+      "loss": 3.8439,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503745177261242e-05,
+      "loss": 3.8571,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50290658251019e-05,
+      "loss": 3.844,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502067987759138e-05,
+      "loss": 3.848,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501231030888459e-05,
+      "loss": 3.8439,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500392436137407e-05,
+      "loss": 3.8485,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9593608379364014,
+      "eval_runtime": 309.269,
+      "eval_samples_per_second": 1233.848,
+      "eval_steps_per_second": 38.559,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499553841386355e-05,
+      "loss": 3.8323,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498715246635303e-05,
+      "loss": 3.8421,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497876651884251e-05,
+      "loss": 3.8487,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497038057133199e-05,
+      "loss": 3.8442,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496199462382146e-05,
+      "loss": 3.8539,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495360867631094e-05,
+      "loss": 3.8306,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494522272880042e-05,
+      "loss": 3.8388,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49368367812899e-05,
+      "loss": 3.8302,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928467212583116e-05,
+      "loss": 3.8482,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4920081265072596e-05,
+      "loss": 3.8392,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911695317562076e-05,
+      "loss": 3.8395,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903309370051556e-05,
+      "loss": 3.841,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4894939801344765e-05,
+      "loss": 3.8348,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886553853834245e-05,
+      "loss": 3.8284,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878167906323725e-05,
+      "loss": 3.83,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869781958813205e-05,
+      "loss": 3.8207,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861396011302685e-05,
+      "loss": 3.834,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4853010063792165e-05,
+      "loss": 3.8288,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844624116281645e-05,
+      "loss": 3.8254,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4836238168771125e-05,
+      "loss": 3.8569,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482786860006434e-05,
+      "loss": 3.8293,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481948265255382e-05,
+      "loss": 3.8325,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48110967050433e-05,
+      "loss": 3.8379,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480271075753278e-05,
+      "loss": 3.832,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479434118882599e-05,
+      "loss": 3.8264,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478595524131547e-05,
+      "loss": 3.8239,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477756929380495e-05,
+      "loss": 3.8369,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476918334629443e-05,
+      "loss": 3.8135,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476079739878391e-05,
+      "loss": 3.8239,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475242783007712e-05,
+      "loss": 3.8208,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47440418825666e-05,
+      "loss": 3.8237,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473565593505608e-05,
+      "loss": 3.8367,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472726998754556e-05,
+      "loss": 3.8269,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4718900418838775e-05,
+      "loss": 3.8236,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710514471328255e-05,
+      "loss": 3.8304,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4702128523817734e-05,
+      "loss": 3.8378,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693742576307214e-05,
+      "loss": 3.8188,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685356628796694e-05,
+      "loss": 3.829,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676987060089903e-05,
+      "loss": 3.8133,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668601112579383e-05,
+      "loss": 3.829,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466021516506886e-05,
+      "loss": 3.8096,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465182921755834e-05,
+      "loss": 3.8387,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464345964885155e-05,
+      "loss": 3.8258,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463507370134103e-05,
+      "loss": 3.8206,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462668775383051e-05,
+      "loss": 3.8281,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461830180631999e-05,
+      "loss": 3.8194,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460991585880947e-05,
+      "loss": 3.8274,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460154629010269e-05,
+      "loss": 3.8136,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459316034259217e-05,
+      "loss": 3.8096,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458477439508165e-05,
+      "loss": 3.829,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457638844757112e-05,
+      "loss": 3.8192,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45680025000606e-05,
+      "loss": 3.8204,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455963293135382e-05,
+      "loss": 3.8142,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455124698384329e-05,
+      "loss": 3.8202,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454286103633277e-05,
+      "loss": 3.8085,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453447508882225e-05,
+      "loss": 3.8242,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4526105520115466e-05,
+      "loss": 3.8171,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517719572604946e-05,
+      "loss": 3.8324,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509333625094426e-05,
+      "loss": 3.8214,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4500947677583906e-05,
+      "loss": 3.8087,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492561730073386e-05,
+      "loss": 3.8138,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4484192161366595e-05,
+      "loss": 3.8179,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475806213856075e-05,
+      "loss": 3.8068,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467420266345555e-05,
+      "loss": 3.8096,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4459034318835035e-05,
+      "loss": 3.8155,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4450648371324515e-05,
+      "loss": 3.8183,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442278802617724e-05,
+      "loss": 3.7978,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4433892855107204e-05,
+      "loss": 3.8055,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425506907596684e-05,
+      "loss": 3.8017,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4417120960086164e-05,
+      "loss": 3.8211,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440875139137938e-05,
+      "loss": 3.814,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440036544386886e-05,
+      "loss": 3.8121,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439197949635834e-05,
+      "loss": 3.8151,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438359354884782e-05,
+      "loss": 3.8071,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43752076013373e-05,
+      "loss": 3.8265,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436682165382678e-05,
+      "loss": 3.8,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435845208511999e-05,
+      "loss": 3.8184,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435006613760947e-05,
+      "loss": 3.8116,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434168019009895e-05,
+      "loss": 3.8108,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433329424258843e-05,
+      "loss": 3.8132,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432492467388164e-05,
+      "loss": 3.8104,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431653872637112e-05,
+      "loss": 3.8151,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308152778860604e-05,
+      "loss": 3.8118,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299766831350084e-05,
+      "loss": 3.7955,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429139726264329e-05,
+      "loss": 3.8125,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428301131513277e-05,
+      "loss": 3.8031,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427462536762225e-05,
+      "loss": 3.812,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426623942011173e-05,
+      "loss": 3.8072,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425785347260121e-05,
+      "loss": 3.7969,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424948390389442e-05,
+      "loss": 3.8093,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42410979563839e-05,
+      "loss": 3.8035,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423271200887338e-05,
+      "loss": 3.8119,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422432606136286e-05,
+      "loss": 3.7937,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421597287145981e-05,
+      "loss": 3.8089,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420758692394929e-05,
+      "loss": 3.8226,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419920097643877e-05,
+      "loss": 3.7999,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419081502892825e-05,
+      "loss": 3.7866,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418242908141773e-05,
+      "loss": 3.8007,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417404313390721e-05,
+      "loss": 3.8045,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416565718639669e-05,
+      "loss": 3.8095,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415727123888617e-05,
+      "loss": 3.8064,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4148901670179376e-05,
+      "loss": 3.7976,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4140515722668856e-05,
+      "loss": 3.7992,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4132129775158336e-05,
+      "loss": 3.8069,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123743827647816e-05,
+      "loss": 3.7863,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115357880137296e-05,
+      "loss": 3.7976,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410698831143051e-05,
+      "loss": 3.8019,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409860236391999e-05,
+      "loss": 3.806,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40902327952132e-05,
+      "loss": 3.7783,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408184684770268e-05,
+      "loss": 3.7981,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407346090019216e-05,
+      "loss": 3.7924,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406507495268164e-05,
+      "loss": 3.7993,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405668900517112e-05,
+      "loss": 3.7985,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40483030576606e-05,
+      "loss": 3.7916,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4039917110150073e-05,
+      "loss": 3.8025,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031531162639553e-05,
+      "loss": 3.8046,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402316159393277e-05,
+      "loss": 3.7911,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401477564642225e-05,
+      "loss": 3.8026,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400638969891173e-05,
+      "loss": 3.7931,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399800375140121e-05,
+      "loss": 3.8014,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3989634182694425e-05,
+      "loss": 3.7826,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981248235183905e-05,
+      "loss": 3.8051,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397286228767338e-05,
+      "loss": 3.7849,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396447634016286e-05,
+      "loss": 3.7991,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3956106771456074e-05,
+      "loss": 3.7891,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394772082394555e-05,
+      "loss": 3.8039,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393933487643503e-05,
+      "loss": 3.7951,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393094892892451e-05,
+      "loss": 3.7884,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392256298141399e-05,
+      "loss": 3.7787,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39141934127072e-05,
+      "loss": 3.809,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390580746519668e-05,
+      "loss": 3.7913,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389742151768616e-05,
+      "loss": 3.7981,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388903557017564e-05,
+      "loss": 3.7858,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388066600146885e-05,
+      "loss": 3.7985,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387228005395833e-05,
+      "loss": 3.8004,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386389410644781e-05,
+      "loss": 3.7888,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385550815893729e-05,
+      "loss": 3.7889,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38471385902305e-05,
+      "loss": 3.797,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383875264271998e-05,
+      "loss": 3.7988,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383036669520946e-05,
+      "loss": 3.8014,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382198074769894e-05,
+      "loss": 3.7814,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381359480018842e-05,
+      "loss": 3.7966,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805225231481637e-05,
+      "loss": 3.7886,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796839283971116e-05,
+      "loss": 3.7923,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788453336460596e-05,
+      "loss": 3.7988,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3780067388950076e-05,
+      "loss": 3.7865,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771697820243285e-05,
+      "loss": 3.8008,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763311872732765e-05,
+      "loss": 3.785,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3754925925222245e-05,
+      "loss": 3.8001,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.934128761291504,
+      "eval_runtime": 303.936,
+      "eval_samples_per_second": 1255.498,
+      "eval_steps_per_second": 39.235,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3746539977711725e-05,
+      "loss": 3.7815,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3738170409004934e-05,
+      "loss": 3.7913,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3729784461494414e-05,
+      "loss": 3.7939,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3721398513983894e-05,
+      "loss": 3.7882,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3713012566473374e-05,
+      "loss": 3.7987,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370464299776659e-05,
+      "loss": 3.7795,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369625705025607e-05,
+      "loss": 3.7829,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368787110274555e-05,
+      "loss": 3.7758,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367948515523503e-05,
+      "loss": 3.7948,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367111558652824e-05,
+      "loss": 3.789,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366272963901772e-05,
+      "loss": 3.7854,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36543436915072e-05,
+      "loss": 3.7895,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364597412280041e-05,
+      "loss": 3.778,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363758817528989e-05,
+      "loss": 3.7785,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362920222777937e-05,
+      "loss": 3.7778,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362081628026885e-05,
+      "loss": 3.7671,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361243033275833e-05,
+      "loss": 3.7785,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360404438524781e-05,
+      "loss": 3.7738,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359565843773729e-05,
+      "loss": 3.7736,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3587272490226775e-05,
+      "loss": 3.8078,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3578902921519984e-05,
+      "loss": 3.7738,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3570516974009464e-05,
+      "loss": 3.7836,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3562131026498944e-05,
+      "loss": 3.7833,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3553745078988424e-05,
+      "loss": 3.7798,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354537551028163e-05,
+      "loss": 3.7721,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353698956277111e-05,
+      "loss": 3.7761,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352860361526059e-05,
+      "loss": 3.7831,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352021766775007e-05,
+      "loss": 3.7634,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351184809904328e-05,
+      "loss": 3.7708,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350346215153276e-05,
+      "loss": 3.769,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349507620402224e-05,
+      "loss": 3.7771,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348669025651173e-05,
+      "loss": 3.7803,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347832068780494e-05,
+      "loss": 3.7775,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346993474029442e-05,
+      "loss": 3.7745,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34615487927839e-05,
+      "loss": 3.7771,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345316284527338e-05,
+      "loss": 3.7833,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344477689776286e-05,
+      "loss": 3.7664,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3436407329056066e-05,
+      "loss": 3.7793,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3428021381545546e-05,
+      "loss": 3.7595,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419635434035026e-05,
+      "loss": 3.7817,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3411249486524506e-05,
+      "loss": 3.7559,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402879917817715e-05,
+      "loss": 3.7909,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3394493970307195e-05,
+      "loss": 3.7747,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338610802279668e-05,
+      "loss": 3.771,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377722075286155e-05,
+      "loss": 3.7753,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336935250657937e-05,
+      "loss": 3.7739,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336096655906885e-05,
+      "loss": 3.7739,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335258061155833e-05,
+      "loss": 3.7634,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3344194664047804e-05,
+      "loss": 3.7546,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333582509534102e-05,
+      "loss": 3.7846,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33274391478305e-05,
+      "loss": 3.7705,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331905320031998e-05,
+      "loss": 3.7763,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331066725280945e-05,
+      "loss": 3.7645,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330229768410267e-05,
+      "loss": 3.7706,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329391173659215e-05,
+      "loss": 3.7594,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328552578908163e-05,
+      "loss": 3.7743,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327713984157111e-05,
+      "loss": 3.7658,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3268770272864325e-05,
+      "loss": 3.782,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3260384325353805e-05,
+      "loss": 3.7768,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325199837784328e-05,
+      "loss": 3.7597,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324361243033276e-05,
+      "loss": 3.7643,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3235242861625974e-05,
+      "loss": 3.7726,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322685691411545e-05,
+      "loss": 3.7572,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321847096660493e-05,
+      "loss": 3.7599,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321008501909441e-05,
+      "loss": 3.769,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320171545038762e-05,
+      "loss": 3.7657,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31933295028771e-05,
+      "loss": 3.7506,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318494355536658e-05,
+      "loss": 3.7587,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317655760785606e-05,
+      "loss": 3.7526,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316818803914928e-05,
+      "loss": 3.7717,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315980209163875e-05,
+      "loss": 3.765,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315141614412823e-05,
+      "loss": 3.762,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314303019661771e-05,
+      "loss": 3.772,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313466062791092e-05,
+      "loss": 3.7567,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31262746804004e-05,
+      "loss": 3.7775,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311788873288988e-05,
+      "loss": 3.7521,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310950278537936e-05,
+      "loss": 3.7713,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3101133216672576e-05,
+      "loss": 3.7648,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3092747269162056e-05,
+      "loss": 3.7626,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3084361321651536e-05,
+      "loss": 3.7679,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3075975374141016e-05,
+      "loss": 3.7615,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3067605805434225e-05,
+      "loss": 3.7654,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3059219857923705e-05,
+      "loss": 3.7655,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050833910413185e-05,
+      "loss": 3.7493,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042447962902665e-05,
+      "loss": 3.7647,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3034078394195874e-05,
+      "loss": 3.7572,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025692446685354e-05,
+      "loss": 3.7632,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3017306499174834e-05,
+      "loss": 3.7565,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008920551664314e-05,
+      "loss": 3.7512,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300055098295753e-05,
+      "loss": 3.7604,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299216503544701e-05,
+      "loss": 3.7616,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.298377908793649e-05,
+      "loss": 3.7621,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297539314042597e-05,
+      "loss": 3.748,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296702357171918e-05,
+      "loss": 3.7628,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295863762420866e-05,
+      "loss": 3.7735,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295025167669814e-05,
+      "loss": 3.7567,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294186572918762e-05,
+      "loss": 3.7391,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293349616048083e-05,
+      "loss": 3.7558,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292511021297031e-05,
+      "loss": 3.7585,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291672426545979e-05,
+      "loss": 3.7595,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290833831794927e-05,
+      "loss": 3.7636,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2899968749242484e-05,
+      "loss": 3.7525,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2891582801731964e-05,
+      "loss": 3.7523,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2883196854221444e-05,
+      "loss": 3.7594,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2874810906710923e-05,
+      "loss": 3.7433,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286644133800413e-05,
+      "loss": 3.7524,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285805539049361e-05,
+      "loss": 3.7533,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284966944298309e-05,
+      "loss": 3.7597,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28412998742763e-05,
+      "loss": 3.736,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283291392676578e-05,
+      "loss": 3.7495,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282452797925526e-05,
+      "loss": 3.7512,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281614203174474e-05,
+      "loss": 3.7515,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280775608423422e-05,
+      "loss": 3.7552,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27993701367237e-05,
+      "loss": 3.7429,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279098418921318e-05,
+      "loss": 3.7583,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278259824170267e-05,
+      "loss": 3.7596,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277422867299588e-05,
+      "loss": 3.7443,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276584272548536e-05,
+      "loss": 3.7615,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275745677797484e-05,
+      "loss": 3.7467,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274907083046432e-05,
+      "loss": 3.7582,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2740701261757526e-05,
+      "loss": 3.7431,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2732315314247006e-05,
+      "loss": 3.7542,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2723929366736486e-05,
+      "loss": 3.7417,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2715543419225966e-05,
+      "loss": 3.7527,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2707173850519175e-05,
+      "loss": 3.7485,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2698787903008655e-05,
+      "loss": 3.7555,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2690401955498135e-05,
+      "loss": 3.7479,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268201600798762e-05,
+      "loss": 3.7501,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267364643928083e-05,
+      "loss": 3.7316,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266526049177031e-05,
+      "loss": 3.7662,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265687454425979e-05,
+      "loss": 3.7459,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264848859674927e-05,
+      "loss": 3.753,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264011902804248e-05,
+      "loss": 3.7415,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263173308053196e-05,
+      "loss": 3.7565,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262334713302144e-05,
+      "loss": 3.7528,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261496118551092e-05,
+      "loss": 3.7464,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260659161680413e-05,
+      "loss": 3.7461,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259820566929361e-05,
+      "loss": 3.7513,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258981972178309e-05,
+      "loss": 3.7517,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2581433774272575e-05,
+      "loss": 3.7632,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2573064205565784e-05,
+      "loss": 3.7353,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2564678258055264e-05,
+      "loss": 3.7548,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2556292310544744e-05,
+      "loss": 3.7415,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2547906363034224e-05,
+      "loss": 3.7523,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253955317313116e-05,
+      "loss": 3.7528,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253116722562064e-05,
+      "loss": 3.7398,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252278127811012e-05,
+      "loss": 3.7592,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25143953305996e-05,
+      "loss": 3.7459,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250600938308908e-05,
+      "loss": 3.7527,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.919607162475586,
+      "eval_runtime": 303.7413,
+      "eval_samples_per_second": 1256.303,
+      "eval_steps_per_second": 39.26,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249762343557856e-05,
+      "loss": 3.74,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248923748806804e-05,
+      "loss": 3.741,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248085154055753e-05,
+      "loss": 3.7522,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472465593047e-05,
+      "loss": 3.7488,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246407964553648e-05,
+      "loss": 3.7542,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245569369802596e-05,
+      "loss": 3.7426,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244730775051544e-05,
+      "loss": 3.7353,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243892180300492e-05,
+      "loss": 3.7387,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243055223429813e-05,
+      "loss": 3.7483,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242216628678761e-05,
+      "loss": 3.7479,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241378033927709e-05,
+      "loss": 3.7412,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240539439176657e-05,
+      "loss": 3.7481,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239702482305978e-05,
+      "loss": 3.7313,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238863887554927e-05,
+      "loss": 3.7406,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238025292803875e-05,
+      "loss": 3.7383,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237186698052823e-05,
+      "loss": 3.7252,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363497411821436e-05,
+      "loss": 3.74,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2355111464310916e-05,
+      "loss": 3.729,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346725516800396e-05,
+      "loss": 3.7322,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2338339569289876e-05,
+      "loss": 3.7687,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2329970000583085e-05,
+      "loss": 3.7347,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321584053072565e-05,
+      "loss": 3.7379,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2313198105562045e-05,
+      "loss": 3.7412,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304812158051525e-05,
+      "loss": 3.745,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2296442589344734e-05,
+      "loss": 3.7237,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228805664183422e-05,
+      "loss": 3.736,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22796706943237e-05,
+      "loss": 3.7385,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227130112561691e-05,
+      "loss": 3.7259,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226291517810639e-05,
+      "loss": 3.7313,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225452923059587e-05,
+      "loss": 3.7262,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224614328308535e-05,
+      "loss": 3.7365,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223777371437856e-05,
+      "loss": 3.7429,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222938776686804e-05,
+      "loss": 3.7363,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222100181935752e-05,
+      "loss": 3.7333,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2212615871847e-05,
+      "loss": 3.7344,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220424630314021e-05,
+      "loss": 3.7429,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219586035562969e-05,
+      "loss": 3.7272,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2187474408119174e-05,
+      "loss": 3.7425,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2179088460608654e-05,
+      "loss": 3.7175,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217071889190186e-05,
+      "loss": 3.7356,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216233294439134e-05,
+      "loss": 3.7172,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215394699688082e-05,
+      "loss": 3.75,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21455610493703e-05,
+      "loss": 3.7316,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213719148066351e-05,
+      "loss": 3.7335,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212880553315299e-05,
+      "loss": 3.7344,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212041958564247e-05,
+      "loss": 3.7347,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211203363813195e-05,
+      "loss": 3.7361,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210366406942516e-05,
+      "loss": 3.7249,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209527812191464e-05,
+      "loss": 3.7106,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208689217440413e-05,
+      "loss": 3.7452,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207850622689361e-05,
+      "loss": 3.727,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207013665818682e-05,
+      "loss": 3.7395,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20617507106763e-05,
+      "loss": 3.7224,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205336476316578e-05,
+      "loss": 3.7294,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204497881565526e-05,
+      "loss": 3.716,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036609246948466e-05,
+      "loss": 3.7357,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2028223299437946e-05,
+      "loss": 3.7299,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2019837351927426e-05,
+      "loss": 3.7381,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011451404416906e-05,
+      "loss": 3.7363,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2003081835710115e-05,
+      "loss": 3.7193,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1994695888199595e-05,
+      "loss": 3.7234,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1986309940689075e-05,
+      "loss": 3.7398,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197792399317856e-05,
+      "loss": 3.7162,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196955442447177e-05,
+      "loss": 3.72,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196116847696125e-05,
+      "loss": 3.7282,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195278252945073e-05,
+      "loss": 3.7334,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194439658194021e-05,
+      "loss": 3.7101,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193601063442969e-05,
+      "loss": 3.718,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19276410657229e-05,
+      "loss": 3.718,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191925511821238e-05,
+      "loss": 3.7236,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191086917070186e-05,
+      "loss": 3.73,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190249960199507e-05,
+      "loss": 3.7253,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189411365448455e-05,
+      "loss": 3.7298,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188572770697403e-05,
+      "loss": 3.7182,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1877341759463515e-05,
+      "loss": 3.742,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1868972190756724e-05,
+      "loss": 3.7144,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860586243246204e-05,
+      "loss": 3.7314,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1852200295735684e-05,
+      "loss": 3.7248,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843814348225164e-05,
+      "loss": 3.7221,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18354611583221e-05,
+      "loss": 3.7301,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182707521081158e-05,
+      "loss": 3.7264,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181868926330106e-05,
+      "loss": 3.7228,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181030331579054e-05,
+      "loss": 3.7268,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180191736828002e-05,
+      "loss": 3.7136,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17935314207695e-05,
+      "loss": 3.7224,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178514547325898e-05,
+      "loss": 3.7205,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177675952574847e-05,
+      "loss": 3.7262,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176838995704168e-05,
+      "loss": 3.7185,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176000400953116e-05,
+      "loss": 3.7133,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175161806202064e-05,
+      "loss": 3.722,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174323211451012e-05,
+      "loss": 3.7259,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173486254580333e-05,
+      "loss": 3.7245,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172647659829281e-05,
+      "loss": 3.7085,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171809065078229e-05,
+      "loss": 3.7245,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170970470327177e-05,
+      "loss": 3.7338,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1701335134564976e-05,
+      "loss": 3.7219,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1692949187054456e-05,
+      "loss": 3.7018,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1684563239543936e-05,
+      "loss": 3.7187,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167617729203342e-05,
+      "loss": 3.7148,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166780772332663e-05,
+      "loss": 3.7243,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165942177581611e-05,
+      "loss": 3.7266,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165103582830559e-05,
+      "loss": 3.7139,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164264988079507e-05,
+      "loss": 3.7135,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163428031208828e-05,
+      "loss": 3.7236,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162589436457776e-05,
+      "loss": 3.7055,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161750841706724e-05,
+      "loss": 3.7194,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160912246955672e-05,
+      "loss": 3.7115,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160075290084993e-05,
+      "loss": 3.7266,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1592383332143145e-05,
+      "loss": 3.698,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1583997384632625e-05,
+      "loss": 3.7092,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1575611437122105e-05,
+      "loss": 3.7115,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1567225489611585e-05,
+      "loss": 3.7199,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558839542101065e-05,
+      "loss": 3.7141,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1550453594590545e-05,
+      "loss": 3.709,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1542067647080025e-05,
+      "loss": 3.7174,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15336816995695e-05,
+      "loss": 3.7221,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1525312130862714e-05,
+      "loss": 3.7071,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1516926183352194e-05,
+      "loss": 3.7261,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1508540235841674e-05,
+      "loss": 3.7101,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150015428833115e-05,
+      "loss": 3.717,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149178471962436e-05,
+      "loss": 3.708,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148339877211384e-05,
+      "loss": 3.7161,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147501282460333e-05,
+      "loss": 3.7072,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14666268770928e-05,
+      "loss": 3.713,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145825730838602e-05,
+      "loss": 3.7168,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14498713608755e-05,
+      "loss": 3.7148,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144148541336497e-05,
+      "loss": 3.7106,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143309946585445e-05,
+      "loss": 3.7129,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142472989714767e-05,
+      "loss": 3.6953,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141634394963715e-05,
+      "loss": 3.7284,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140795800212662e-05,
+      "loss": 3.7095,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13995720546161e-05,
+      "loss": 3.7151,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139120248590932e-05,
+      "loss": 3.7075,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13828165383988e-05,
+      "loss": 3.7251,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137443059088828e-05,
+      "loss": 3.7124,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136604464337776e-05,
+      "loss": 3.7081,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135767507467097e-05,
+      "loss": 3.7097,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1349289127160446e-05,
+      "loss": 3.7143,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340903179649926e-05,
+      "loss": 3.7156,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332517232139406e-05,
+      "loss": 3.7279,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132414766343262e-05,
+      "loss": 3.6982,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315761715922095e-05,
+      "loss": 3.7169,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1307375768411575e-05,
+      "loss": 3.7076,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1298989820901055e-05,
+      "loss": 3.7193,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129062025219427e-05,
+      "loss": 3.7118,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128223430468375e-05,
+      "loss": 3.7086,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127384835717323e-05,
+      "loss": 3.7192,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126546240966271e-05,
+      "loss": 3.7113,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125709284095592e-05,
+      "loss": 3.7166,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9099810123443604,
+      "eval_runtime": 304.112,
+      "eval_samples_per_second": 1254.771,
+      "eval_steps_per_second": 39.213,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12487068934454e-05,
+      "loss": 3.7103,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124032094593488e-05,
+      "loss": 3.7025,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1231951377228095e-05,
+      "loss": 3.7141,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122356542971757e-05,
+      "loss": 3.7144,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121517948220705e-05,
+      "loss": 3.7208,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120679353469653e-05,
+      "loss": 3.704,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1198423965989744e-05,
+      "loss": 3.7003,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1190038018479224e-05,
+      "loss": 3.6992,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1181652070968704e-05,
+      "loss": 3.715,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1173266123458184e-05,
+      "loss": 3.7147,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116489655475139e-05,
+      "loss": 3.7013,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115651060724087e-05,
+      "loss": 3.7109,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114814103853409e-05,
+      "loss": 3.7,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113975509102357e-05,
+      "loss": 3.7038,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113136914351304e-05,
+      "loss": 3.6978,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112298319600252e-05,
+      "loss": 3.694,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1114597248492e-05,
+      "loss": 3.7004,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110621130098148e-05,
+      "loss": 3.6986,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109782535347096e-05,
+      "loss": 3.6943,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108943940596044e-05,
+      "loss": 3.7309,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108106983725366e-05,
+      "loss": 3.704,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107268388974314e-05,
+      "loss": 3.7014,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106429794223262e-05,
+      "loss": 3.704,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10559119947221e-05,
+      "loss": 3.708,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104754242601531e-05,
+      "loss": 3.6899,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103915647850479e-05,
+      "loss": 3.7029,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103077053099427e-05,
+      "loss": 3.7048,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1022384583483747e-05,
+      "loss": 3.6937,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1014015014776956e-05,
+      "loss": 3.6929,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005629067266436e-05,
+      "loss": 3.6928,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0997243119755916e-05,
+      "loss": 3.7041,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988857172245396e-05,
+      "loss": 3.7079,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098048760353861e-05,
+      "loss": 3.7022,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097210165602809e-05,
+      "loss": 3.6978,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096371570851757e-05,
+      "loss": 3.6993,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095532976100705e-05,
+      "loss": 3.7063,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094696019230026e-05,
+      "loss": 3.6985,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093857424478974e-05,
+      "loss": 3.7041,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093018829727922e-05,
+      "loss": 3.6862,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092181872857243e-05,
+      "loss": 3.7001,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091343278106191e-05,
+      "loss": 3.685,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090504683355139e-05,
+      "loss": 3.714,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08966772648446e-05,
+      "loss": 3.7007,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0888291317334085e-05,
+      "loss": 3.6995,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0879905369823565e-05,
+      "loss": 3.7028,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0871519422313045e-05,
+      "loss": 3.6974,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0863133474802525e-05,
+      "loss": 3.7012,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0854747527292005e-05,
+      "loss": 3.6938,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0846361579781485e-05,
+      "loss": 3.6716,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0837975632270965e-05,
+      "loss": 3.7117,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0829606063564174e-05,
+      "loss": 3.6943,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082123649485738e-05,
+      "loss": 3.7077,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081285054734686e-05,
+      "loss": 3.6914,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080446459983634e-05,
+      "loss": 3.6957,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079607865232582e-05,
+      "loss": 3.6799,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07876927048153e-05,
+      "loss": 3.7023,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077930675730478e-05,
+      "loss": 3.6978,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077092080979427e-05,
+      "loss": 3.7023,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076253486228375e-05,
+      "loss": 3.7076,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075416529357696e-05,
+      "loss": 3.6856,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074577934606644e-05,
+      "loss": 3.6882,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073739339855592e-05,
+      "loss": 3.7068,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07290074510454e-05,
+      "loss": 3.6847,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072063788233861e-05,
+      "loss": 3.6842,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071225193482809e-05,
+      "loss": 3.6958,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070386598731757e-05,
+      "loss": 3.6963,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069548003980705e-05,
+      "loss": 3.6819,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0687110471100257e-05,
+      "loss": 3.6851,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0678724523589736e-05,
+      "loss": 3.6802,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067033857607922e-05,
+      "loss": 3.6917,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066196900737243e-05,
+      "loss": 3.6974,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065358305986191e-05,
+      "loss": 3.6926,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064519711235139e-05,
+      "loss": 3.6929,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.063681116484087e-05,
+      "loss": 3.6897,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062844159613408e-05,
+      "loss": 3.7069,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062005564862356e-05,
+      "loss": 3.6812,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061166970111304e-05,
+      "loss": 3.7019,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060328375360252e-05,
+      "loss": 3.6909,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059491418489573e-05,
+      "loss": 3.6853,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0586544616188946e-05,
+      "loss": 3.7007,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0578158668678426e-05,
+      "loss": 3.6936,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0569772721167906e-05,
+      "loss": 3.6928,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0561386773657386e-05,
+      "loss": 3.696,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0553000826146866e-05,
+      "loss": 3.681,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0544614878636346e-05,
+      "loss": 3.6896,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053622893112582e-05,
+      "loss": 3.6897,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05278429836153e-05,
+      "loss": 3.6949,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0519473414908515e-05,
+      "loss": 3.6854,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0511087467397995e-05,
+      "loss": 3.6836,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050270151988747e-05,
+      "loss": 3.6844,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049431557237695e-05,
+      "loss": 3.6945,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0485946003670164e-05,
+      "loss": 3.6934,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0477560056159644e-05,
+      "loss": 3.6788,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0469174108649124e-05,
+      "loss": 3.6945,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0460788161138604e-05,
+      "loss": 3.6997,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045241859243182e-05,
+      "loss": 3.6892,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044403264492129e-05,
+      "loss": 3.6709,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043564669741077e-05,
+      "loss": 3.6843,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042726074990025e-05,
+      "loss": 3.6801,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041889118119347e-05,
+      "loss": 3.6974,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041050523368294e-05,
+      "loss": 3.6936,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040211928617242e-05,
+      "loss": 3.6821,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03937333386619e-05,
+      "loss": 3.683,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038536376995512e-05,
+      "loss": 3.6845,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03769778224446e-05,
+      "loss": 3.679,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036859187493408e-05,
+      "loss": 3.6864,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036020592742356e-05,
+      "loss": 3.6773,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0351836358716766e-05,
+      "loss": 3.6982,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034346679000998e-05,
+      "loss": 3.6678,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033508084249946e-05,
+      "loss": 3.6768,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032669489498894e-05,
+      "loss": 3.6782,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0318308947478415e-05,
+      "loss": 3.6879,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0309922999967895e-05,
+      "loss": 3.679,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0301537052457375e-05,
+      "loss": 3.6791,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0293151104946855e-05,
+      "loss": 3.6865,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284765157436335e-05,
+      "loss": 3.6923,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027639558872955e-05,
+      "loss": 3.6733,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026800964121903e-05,
+      "loss": 3.6904,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025962369370851e-05,
+      "loss": 3.6816,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025123774619799e-05,
+      "loss": 3.6849,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02428681774912e-05,
+      "loss": 3.6788,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023448222998068e-05,
+      "loss": 3.68,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022609628247016e-05,
+      "loss": 3.6752,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021771033495964e-05,
+      "loss": 3.6847,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020934076625285e-05,
+      "loss": 3.688,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020095481874233e-05,
+      "loss": 3.6813,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019256887123181e-05,
+      "loss": 3.6799,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018418292372129e-05,
+      "loss": 3.6802,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0175813355014505e-05,
+      "loss": 3.6652,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0167427407503985e-05,
+      "loss": 3.6978,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0159041459993465e-05,
+      "loss": 3.6771,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0150671891286674e-05,
+      "loss": 3.6834,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0142285943776154e-05,
+      "loss": 3.6748,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0133899996265634e-05,
+      "loss": 3.693,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0125514048755114e-05,
+      "loss": 3.6813,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0117128101244594e-05,
+      "loss": 3.6784,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0108742153734074e-05,
+      "loss": 3.6813,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0100356206223554e-05,
+      "loss": 3.68,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009198663751676e-05,
+      "loss": 3.6821,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008360069000624e-05,
+      "loss": 3.7007,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007521474249572e-05,
+      "loss": 3.6639,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006682879498521e-05,
+      "loss": 3.6895,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005845922627842e-05,
+      "loss": 3.6763,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00500732787679e-05,
+      "loss": 3.6861,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004168733125738e-05,
+      "loss": 3.6787,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003330138374686e-05,
+      "loss": 3.6812,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002493181504007e-05,
+      "loss": 3.6864,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0016562246333276e-05,
+      "loss": 3.6798,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0008176298822756e-05,
+      "loss": 3.6843,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9040579795837402,
+      "eval_runtime": 302.0975,
+      "eval_samples_per_second": 1263.139,
+      "eval_steps_per_second": 39.474,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9999790351312236e-05,
+      "loss": 3.6787,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9991404403801716e-05,
+      "loss": 3.6714,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9983018456291196e-05,
+      "loss": 3.6847,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9974632508780676e-05,
+      "loss": 3.6812,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996624656127016e-05,
+      "loss": 3.6908,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995786061375964e-05,
+      "loss": 3.6724,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994947466624912e-05,
+      "loss": 3.6713,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99410887187386e-05,
+      "loss": 3.6736,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993271915003181e-05,
+      "loss": 3.6811,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992433320252129e-05,
+      "loss": 3.6848,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991594725501077e-05,
+      "loss": 3.6712,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990756130750025e-05,
+      "loss": 3.6795,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989920811759719e-05,
+      "loss": 3.674,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989082217008667e-05,
+      "loss": 3.67,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988243622257615e-05,
+      "loss": 3.6719,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987405027506563e-05,
+      "loss": 3.6613,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986566432755512e-05,
+      "loss": 3.671,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9857278380044597e-05,
+      "loss": 3.6652,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848892432534077e-05,
+      "loss": 3.6666,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984050648502355e-05,
+      "loss": 3.6968,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9832136916316766e-05,
+      "loss": 3.6774,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823750968806246e-05,
+      "loss": 3.6722,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9815365021295726e-05,
+      "loss": 3.6754,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98069790737852e-05,
+      "loss": 3.6788,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9798609505078415e-05,
+      "loss": 3.6556,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9790223557567895e-05,
+      "loss": 3.6752,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978183761005737e-05,
+      "loss": 3.6772,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9773451662546854e-05,
+      "loss": 3.6617,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976508209384007e-05,
+      "loss": 3.661,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975669614632955e-05,
+      "loss": 3.662,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9748310198819023e-05,
+      "loss": 3.6724,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9739924251308503e-05,
+      "loss": 3.6792,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973155468260172e-05,
+      "loss": 3.6705,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.97231687350912e-05,
+      "loss": 3.6697,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971478278758067e-05,
+      "loss": 3.6708,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970641321887389e-05,
+      "loss": 3.6738,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969802727136337e-05,
+      "loss": 3.6705,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968964132385284e-05,
+      "loss": 3.6718,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968125537634232e-05,
+      "loss": 3.6561,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967288580763554e-05,
+      "loss": 3.669,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966451623892875e-05,
+      "loss": 3.6581,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965613029141823e-05,
+      "loss": 3.681,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964774434390771e-05,
+      "loss": 3.6728,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963935839639719e-05,
+      "loss": 3.6681,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963097244888667e-05,
+      "loss": 3.675,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9622586501376146e-05,
+      "loss": 3.6679,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9614200553865626e-05,
+      "loss": 3.6738,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9605814606355106e-05,
+      "loss": 3.6645,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9597445037648315e-05,
+      "loss": 3.6408,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9589059090137795e-05,
+      "loss": 3.6817,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9580673142627275e-05,
+      "loss": 3.6639,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957228719511676e-05,
+      "loss": 3.6813,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956391762640997e-05,
+      "loss": 3.6625,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955553167889945e-05,
+      "loss": 3.6639,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954714573138893e-05,
+      "loss": 3.6548,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953875978387841e-05,
+      "loss": 3.6706,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953039021517162e-05,
+      "loss": 3.6716,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95220042676611e-05,
+      "loss": 3.6708,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951361832015058e-05,
+      "loss": 3.6797,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950523237264006e-05,
+      "loss": 3.6588,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949686280393327e-05,
+      "loss": 3.6557,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948847685642275e-05,
+      "loss": 3.6782,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948009090891223e-05,
+      "loss": 3.6549,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947170496140171e-05,
+      "loss": 3.6551,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9463335392694925e-05,
+      "loss": 3.6631,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454949445184404e-05,
+      "loss": 3.6716,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944657987647762e-05,
+      "loss": 3.6501,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9438193928967094e-05,
+      "loss": 3.6572,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429807981456573e-05,
+      "loss": 3.6542,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9421422033946053e-05,
+      "loss": 3.6596,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9413036086435533e-05,
+      "loss": 3.67,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940465013892501e-05,
+      "loss": 3.6618,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939626419141449e-05,
+      "loss": 3.6648,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938787824390397e-05,
+      "loss": 3.6595,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937950867519718e-05,
+      "loss": 3.6804,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937112272768666e-05,
+      "loss": 3.6527,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936273678017615e-05,
+      "loss": 3.6711,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935435083266563e-05,
+      "loss": 3.6634,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934598126395884e-05,
+      "loss": 3.6597,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933761169525205e-05,
+      "loss": 3.6707,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932922574774153e-05,
+      "loss": 3.6665,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932083980023101e-05,
+      "loss": 3.6647,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931245385272049e-05,
+      "loss": 3.6626,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930406790520997e-05,
+      "loss": 3.6571,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929568195769945e-05,
+      "loss": 3.6571,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.928729601018893e-05,
+      "loss": 3.6624,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927891006267841e-05,
+      "loss": 3.6653,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9270540493971616e-05,
+      "loss": 3.6549,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92621545464611e-05,
+      "loss": 3.6574,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925376859895058e-05,
+      "loss": 3.6567,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924539903024379e-05,
+      "loss": 3.6688,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923701308273327e-05,
+      "loss": 3.6631,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922862713522275e-05,
+      "loss": 3.6524,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922024118771223e-05,
+      "loss": 3.665,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921187161900544e-05,
+      "loss": 3.6707,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920348567149492e-05,
+      "loss": 3.6626,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91950997239844e-05,
+      "loss": 3.6437,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918671377647388e-05,
+      "loss": 3.6556,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917834420776709e-05,
+      "loss": 3.653,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916995826025657e-05,
+      "loss": 3.6671,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9161572312746056e-05,
+      "loss": 3.666,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9153186365235536e-05,
+      "loss": 3.6619,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9144816796528745e-05,
+      "loss": 3.6554,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9136430849018225e-05,
+      "loss": 3.6556,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9128044901507705e-05,
+      "loss": 3.654,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9119658953997185e-05,
+      "loss": 3.6576,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9111289385290394e-05,
+      "loss": 3.6507,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9102903437779874e-05,
+      "loss": 3.6689,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9094533869073083e-05,
+      "loss": 3.644,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9086147921562563e-05,
+      "loss": 3.6455,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907776197405204e-05,
+      "loss": 3.6531,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906937602654152e-05,
+      "loss": 3.655,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906099007903101e-05,
+      "loss": 3.6558,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905260413152049e-05,
+      "loss": 3.6513,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904421818400997e-05,
+      "loss": 3.6586,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903583223649945e-05,
+      "loss": 3.667,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902746266779266e-05,
+      "loss": 3.645,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901907672028214e-05,
+      "loss": 3.6601,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901069077277162e-05,
+      "loss": 3.6556,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90023048252611e-05,
+      "loss": 3.6586,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899391887775057e-05,
+      "loss": 3.6527,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898553293024005e-05,
+      "loss": 3.6498,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897714698272953e-05,
+      "loss": 3.6502,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896877741402275e-05,
+      "loss": 3.6562,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896039146651223e-05,
+      "loss": 3.6596,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895200551900171e-05,
+      "loss": 3.6536,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894361957149119e-05,
+      "loss": 3.6542,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893523362398067e-05,
+      "loss": 3.653,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892684767647015e-05,
+      "loss": 3.638,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891846172895963e-05,
+      "loss": 3.6682,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891007578144911e-05,
+      "loss": 3.6475,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890170621274232e-05,
+      "loss": 3.6627,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8893320265231797e-05,
+      "loss": 3.6445,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8884934317721277e-05,
+      "loss": 3.6652,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8876548370210757e-05,
+      "loss": 3.6548,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8868162422700236e-05,
+      "loss": 3.6508,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8859776475189716e-05,
+      "loss": 3.6528,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8851390527679196e-05,
+      "loss": 3.6548,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8843004580168676e-05,
+      "loss": 3.6579,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883463501146189e-05,
+      "loss": 3.671,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882624906395137e-05,
+      "loss": 3.6395,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881786311644085e-05,
+      "loss": 3.6633,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880949354773406e-05,
+      "loss": 3.6482,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880110760022354e-05,
+      "loss": 3.6574,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879272165271302e-05,
+      "loss": 3.6548,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878435208400623e-05,
+      "loss": 3.652,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877596613649571e-05,
+      "loss": 3.6589,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876758018898519e-05,
+      "loss": 3.6567,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875919424147467e-05,
+      "loss": 3.6545,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8986892700195312,
+      "eval_runtime": 303.1501,
+      "eval_samples_per_second": 1258.753,
+      "eval_steps_per_second": 39.337,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.875080829396415e-05,
+      "loss": 3.6493,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.874242234645363e-05,
+      "loss": 3.648,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.873403639894311e-05,
+      "loss": 3.656,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.872565045143259e-05,
+      "loss": 3.6533,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.871726450392207e-05,
+      "loss": 3.6629,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870887855641155e-05,
+      "loss": 3.6464,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870049260890103e-05,
+      "loss": 3.6465,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869210666139051e-05,
+      "loss": 3.6492,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.868373709268372e-05,
+      "loss": 3.6536,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.86753511451732e-05,
+      "loss": 3.6563,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.866696519766268e-05,
+      "loss": 3.6487,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865857925015216e-05,
+      "loss": 3.6508,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8650226060249104e-05,
+      "loss": 3.6495,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8641840112738584e-05,
+      "loss": 3.642,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8633454165228064e-05,
+      "loss": 3.6448,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8625068217717544e-05,
+      "loss": 3.6365,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8616682270207024e-05,
+      "loss": 3.6398,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8608296322696503e-05,
+      "loss": 3.6412,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8599910375185983e-05,
+      "loss": 3.641,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8591524427675463e-05,
+      "loss": 3.6655,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.858315485896867e-05,
+      "loss": 3.6532,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.857476891145815e-05,
+      "loss": 3.6479,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.856638296394763e-05,
+      "loss": 3.6488,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.855799701643711e-05,
+      "loss": 3.6533,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854962744773032e-05,
+      "loss": 3.6322,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.85412415002198e-05,
+      "loss": 3.6472,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.853285555270929e-05,
+      "loss": 3.6493,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.852446960519877e-05,
+      "loss": 3.6319,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.851610003649198e-05,
+      "loss": 3.6407,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.850771408898146e-05,
+      "loss": 3.6355,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849932814147094e-05,
+      "loss": 3.6478,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8490958572764146e-05,
+      "loss": 3.6468,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8482572625253626e-05,
+      "loss": 3.6459,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8474186677743106e-05,
+      "loss": 3.6503,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8465800730232586e-05,
+      "loss": 3.638,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8457414782722066e-05,
+      "loss": 3.6489,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8449028835211546e-05,
+      "loss": 3.6512,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8440642887701026e-05,
+      "loss": 3.6405,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8432256940190506e-05,
+      "loss": 3.6327,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842388737148372e-05,
+      "loss": 3.6418,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.841551780277693e-05,
+      "loss": 3.6321,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.840713185526641e-05,
+      "loss": 3.6551,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839874590775589e-05,
+      "loss": 3.6481,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839035996024537e-05,
+      "loss": 3.64,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838197401273485e-05,
+      "loss": 3.6484,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837358806522433e-05,
+      "loss": 3.6396,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836520211771381e-05,
+      "loss": 3.6475,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835683254900702e-05,
+      "loss": 3.6455,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.83484466014965e-05,
+      "loss": 3.6096,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834006065398598e-05,
+      "loss": 3.6573,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833167470647546e-05,
+      "loss": 3.6407,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8323305137768675e-05,
+      "loss": 3.6494,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8314919190258155e-05,
+      "loss": 3.6396,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8306533242747635e-05,
+      "loss": 3.639,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8298147295237115e-05,
+      "loss": 3.6284,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8289761347726595e-05,
+      "loss": 3.6461,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828137540021607e-05,
+      "loss": 3.646,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.827298945270555e-05,
+      "loss": 3.6434,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.826460350519503e-05,
+      "loss": 3.6594,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8256233936488244e-05,
+      "loss": 3.6372,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.824784798897772e-05,
+      "loss": 3.6261,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82394620414672e-05,
+      "loss": 3.6566,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823107609395668e-05,
+      "loss": 3.631,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.822270652524989e-05,
+      "loss": 3.6241,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821432057773937e-05,
+      "loss": 3.6427,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.820593463022885e-05,
+      "loss": 3.6434,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.819756506152207e-05,
+      "loss": 3.6276,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818917911401154e-05,
+      "loss": 3.6324,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818079316650102e-05,
+      "loss": 3.627,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.81724072189905e-05,
+      "loss": 3.6333,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.816402127147998e-05,
+      "loss": 3.6459,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.815563532396946e-05,
+      "loss": 3.6366,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.814724937645894e-05,
+      "loss": 3.6402,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813886342894842e-05,
+      "loss": 3.6318,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813049386024163e-05,
+      "loss": 3.6541,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812210791273111e-05,
+      "loss": 3.6311,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.81137219652206e-05,
+      "loss": 3.6479,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810533601771008e-05,
+      "loss": 3.6378,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.809696644900329e-05,
+      "loss": 3.6367,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8088596880296496e-05,
+      "loss": 3.6433,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8080210932785976e-05,
+      "loss": 3.6453,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8071824985275456e-05,
+      "loss": 3.6369,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8063439037764936e-05,
+      "loss": 3.6407,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8055053090254416e-05,
+      "loss": 3.6318,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8046667142743896e-05,
+      "loss": 3.6313,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8038281195233376e-05,
+      "loss": 3.637,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8029895247722855e-05,
+      "loss": 3.6411,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8021525679016065e-05,
+      "loss": 3.6302,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.801313973150555e-05,
+      "loss": 3.6334,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.800477016279876e-05,
+      "loss": 3.6328,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799638421528824e-05,
+      "loss": 3.6392,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.798799826777772e-05,
+      "loss": 3.6409,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79796123202672e-05,
+      "loss": 3.6275,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797122637275668e-05,
+      "loss": 3.6408,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796284042524616e-05,
+      "loss": 3.641,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.795445447773564e-05,
+      "loss": 3.6441,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794608490902885e-05,
+      "loss": 3.6177,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793769896151833e-05,
+      "loss": 3.629,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792931301400781e-05,
+      "loss": 3.6285,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792092706649729e-05,
+      "loss": 3.6433,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7912557497790505e-05,
+      "loss": 3.6385,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7904171550279985e-05,
+      "loss": 3.6381,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7895785602769465e-05,
+      "loss": 3.6323,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7887399655258945e-05,
+      "loss": 3.63,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7879030086552154e-05,
+      "loss": 3.6269,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7870644139041634e-05,
+      "loss": 3.6353,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7862258191531114e-05,
+      "loss": 3.6288,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7853872244020594e-05,
+      "loss": 3.6368,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784551905411753e-05,
+      "loss": 3.6238,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.783713310660701e-05,
+      "loss": 3.6186,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782874715909649e-05,
+      "loss": 3.6283,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782036121158597e-05,
+      "loss": 3.6305,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781197526407546e-05,
+      "loss": 3.6313,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.780358931656494e-05,
+      "loss": 3.6262,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.779520336905442e-05,
+      "loss": 3.6367,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.77868174215439e-05,
+      "loss": 3.6411,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777844785283711e-05,
+      "loss": 3.6237,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777006190532659e-05,
+      "loss": 3.6341,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776167595781607e-05,
+      "loss": 3.6351,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.775330638910928e-05,
+      "loss": 3.631,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7744920441598757e-05,
+      "loss": 3.6305,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7736534494088237e-05,
+      "loss": 3.6233,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7728148546577716e-05,
+      "loss": 3.6296,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7719778977870926e-05,
+      "loss": 3.6304,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771139303036041e-05,
+      "loss": 3.6346,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.770300708284989e-05,
+      "loss": 3.6309,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.769462113533937e-05,
+      "loss": 3.6263,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768625156663258e-05,
+      "loss": 3.6302,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.767786561912206e-05,
+      "loss": 3.6137,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766947967161154e-05,
+      "loss": 3.6443,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766109372410102e-05,
+      "loss": 3.6237,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7652707776590494e-05,
+      "loss": 3.6405,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.764433820788371e-05,
+      "loss": 3.6224,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.763595226037319e-05,
+      "loss": 3.64,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.762756631286267e-05,
+      "loss": 3.6336,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761918036535215e-05,
+      "loss": 3.6228,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761079441784163e-05,
+      "loss": 3.6299,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760240847033111e-05,
+      "loss": 3.6267,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.759402252282059e-05,
+      "loss": 3.6393,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758563657531007e-05,
+      "loss": 3.6437,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757726700660328e-05,
+      "loss": 3.6155,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756888105909276e-05,
+      "loss": 3.6341,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756049511158224e-05,
+      "loss": 3.6278,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755212554287545e-05,
+      "loss": 3.6348,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754373959536493e-05,
+      "loss": 3.6281,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753535364785441e-05,
+      "loss": 3.6334,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.752696770034389e-05,
+      "loss": 3.6319,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751858175283337e-05,
+      "loss": 3.6316,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751019580532285e-05,
+      "loss": 3.6331,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8964552879333496,
+      "eval_runtime": 303.302,
+      "eval_samples_per_second": 1258.122,
+      "eval_steps_per_second": 39.317,
+      "step": 763200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.750180985781233e-05,
+      "loss": 3.6257,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7493440289105544e-05,
+      "loss": 3.6219,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7485054341595024e-05,
+      "loss": 3.6328,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7476668394084504e-05,
+      "loss": 3.6307,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7468282446573984e-05,
+      "loss": 3.6395,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7459896499063463e-05,
+      "loss": 3.6229,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.745152693035667e-05,
+      "loss": 3.6243,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.744314098284615e-05,
+      "loss": 3.6245,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.743475503533563e-05,
+      "loss": 3.627,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.742636908782511e-05,
+      "loss": 3.6342,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.741798314031459e-05,
+      "loss": 3.6227,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740959719280407e-05,
+      "loss": 3.6244,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740122762409728e-05,
+      "loss": 3.6301,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.739284167658677e-05,
+      "loss": 3.6183,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.738445572907625e-05,
+      "loss": 3.6241,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.737606978156573e-05,
+      "loss": 3.6115,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.736768383405521e-05,
+      "loss": 3.6156,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735929788654468e-05,
+      "loss": 3.6219,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735091193903416e-05,
+      "loss": 3.6142,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.734252599152364e-05,
+      "loss": 3.6405,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.733415642281686e-05,
+      "loss": 3.6337,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.732577047530633e-05,
+      "loss": 3.6239,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.731738452779581e-05,
+      "loss": 3.6278,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730899858028529e-05,
+      "loss": 3.626,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7300629011578506e-05,
+      "loss": 3.6126,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7292243064067986e-05,
+      "loss": 3.6244,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7283857116557466e-05,
+      "loss": 3.625,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.727548754785068e-05,
+      "loss": 3.6081,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7267101600340155e-05,
+      "loss": 3.6159,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7258715652829635e-05,
+      "loss": 3.613,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7250329705319115e-05,
+      "loss": 3.6274,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7241943757808595e-05,
+      "loss": 3.6235,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7233574189101804e-05,
+      "loss": 3.6235,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7225188241591284e-05,
+      "loss": 3.6257,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7216802294080764e-05,
+      "loss": 3.6131,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7208416346570244e-05,
+      "loss": 3.6283,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7200030399059724e-05,
+      "loss": 3.6307,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7191644451549204e-05,
+      "loss": 3.6137,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.718325850403869e-05,
+      "loss": 3.616,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.717487255652817e-05,
+      "loss": 3.6152,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.716651936662511e-05,
+      "loss": 3.6104,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.715813341911459e-05,
+      "loss": 3.6292,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714974747160407e-05,
+      "loss": 3.626,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714136152409355e-05,
+      "loss": 3.6166,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.713297557658303e-05,
+      "loss": 3.6243,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.712458962907251e-05,
+      "loss": 3.6191,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.711620368156199e-05,
+      "loss": 3.6226,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.710781773405147e-05,
+      "loss": 3.6234,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709946454414841e-05,
+      "loss": 3.5922,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709107859663789e-05,
+      "loss": 3.6308,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.708269264912737e-05,
+      "loss": 3.6151,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.707430670161685e-05,
+      "loss": 3.6262,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.706592075410633e-05,
+      "loss": 3.62,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.705753480659581e-05,
+      "loss": 3.6141,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704916523788902e-05,
+      "loss": 3.607,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.70407792903785e-05,
+      "loss": 3.6222,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703239334286798e-05,
+      "loss": 3.6229,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.702400739535746e-05,
+      "loss": 3.6188,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.701562144784694e-05,
+      "loss": 3.6353,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.700723550033642e-05,
+      "loss": 3.6209,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69988495528259e-05,
+      "loss": 3.5989,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.699046360531538e-05,
+      "loss": 3.6311,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69820940366086e-05,
+      "loss": 3.609,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.697370808909808e-05,
+      "loss": 3.6072,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.696532214158756e-05,
+      "loss": 3.6206,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.695693619407704e-05,
+      "loss": 3.6199,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6948583004173976e-05,
+      "loss": 3.6047,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6940197056663456e-05,
+      "loss": 3.6099,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6931811109152936e-05,
+      "loss": 3.6036,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6923425161642416e-05,
+      "loss": 3.6114,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6915039214131896e-05,
+      "loss": 3.6223,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6906653266621376e-05,
+      "loss": 3.6153,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6898267319110856e-05,
+      "loss": 3.6149,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6889881371600336e-05,
+      "loss": 3.6098,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6881511802893545e-05,
+      "loss": 3.6292,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.687312585538303e-05,
+      "loss": 3.6119,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.686473990787251e-05,
+      "loss": 3.6259,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.685635396036199e-05,
+      "loss": 3.6107,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68479843916552e-05,
+      "loss": 3.615,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683961482294841e-05,
+      "loss": 3.6219,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683122887543789e-05,
+      "loss": 3.6226,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.682284292792737e-05,
+      "loss": 3.6145,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.681445698041685e-05,
+      "loss": 3.6206,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.680607103290633e-05,
+      "loss": 3.604,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.679768508539581e-05,
+      "loss": 3.6087,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678929913788529e-05,
+      "loss": 3.6162,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678091319037477e-05,
+      "loss": 3.6199,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6772543621667985e-05,
+      "loss": 3.6073,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6764157674157465e-05,
+      "loss": 3.611,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.675577172664694e-05,
+      "loss": 3.6114,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.674738577913642e-05,
+      "loss": 3.6138,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6739016210429634e-05,
+      "loss": 3.6202,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6730630262919114e-05,
+      "loss": 3.6038,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672224431540859e-05,
+      "loss": 3.6171,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.67138747467018e-05,
+      "loss": 3.6217,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.670548879919128e-05,
+      "loss": 3.6221,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6697102851680756e-05,
+      "loss": 3.5925,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6688716904170236e-05,
+      "loss": 3.6097,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668034733546345e-05,
+      "loss": 3.608,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667196138795294e-05,
+      "loss": 3.6191,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.666357544044241e-05,
+      "loss": 3.6197,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.665520587173563e-05,
+      "loss": 3.614,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664681992422511e-05,
+      "loss": 3.6078,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663843397671459e-05,
+      "loss": 3.6122,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663004802920406e-05,
+      "loss": 3.6058,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662166208169354e-05,
+      "loss": 3.6135,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.661327613418302e-05,
+      "loss": 3.6041,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.66048901866725e-05,
+      "loss": 3.6141,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.659652061796571e-05,
+      "loss": 3.6033,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.658813467045519e-05,
+      "loss": 3.5975,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6579748722944676e-05,
+      "loss": 3.6033,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6571362775434156e-05,
+      "loss": 3.6072,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6562976827923636e-05,
+      "loss": 3.6123,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6554590880413116e-05,
+      "loss": 3.6043,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6546204932902596e-05,
+      "loss": 3.6114,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6537818985392076e-05,
+      "loss": 3.6149,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6529449416685285e-05,
+      "loss": 3.607,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6521063469174765e-05,
+      "loss": 3.612,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6512677521664245e-05,
+      "loss": 3.6122,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6504291574153725e-05,
+      "loss": 3.6096,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6495905626643205e-05,
+      "loss": 3.608,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6487519679132685e-05,
+      "loss": 3.6001,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6479133731622165e-05,
+      "loss": 3.6088,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6470764162915374e-05,
+      "loss": 3.6055,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646237821540486e-05,
+      "loss": 3.6133,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.645399226789434e-05,
+      "loss": 3.6124,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.644560632038382e-05,
+      "loss": 3.6033,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.64372203728733e-05,
+      "loss": 3.6071,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6428834425362774e-05,
+      "loss": 3.593,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6420448477852254e-05,
+      "loss": 3.622,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6412062530341734e-05,
+      "loss": 3.6028,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.640369296163495e-05,
+      "loss": 3.6186,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.639530701412442e-05,
+      "loss": 3.6025,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63869210666139e-05,
+      "loss": 3.6158,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637853511910338e-05,
+      "loss": 3.6114,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637014917159286e-05,
+      "loss": 3.6011,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636177960288608e-05,
+      "loss": 3.6107,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.635339365537556e-05,
+      "loss": 3.6001,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.634500770786504e-05,
+      "loss": 3.6185,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.633662176035452e-05,
+      "loss": 3.6169,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6328235812844e-05,
+      "loss": 3.601,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631984986533348e-05,
+      "loss": 3.6088,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631146391782296e-05,
+      "loss": 3.6059,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.630307797031244e-05,
+      "loss": 3.6154,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.629470840160565e-05,
+      "loss": 3.6033,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.628632245409513e-05,
+      "loss": 3.6179,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.627793650658461e-05,
+      "loss": 3.6041,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6269566937877816e-05,
+      "loss": 3.6139,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6261180990367296e-05,
+      "loss": 3.6082,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.896568775177002,
+      "eval_runtime": 303.9645,
+      "eval_samples_per_second": 1255.38,
+      "eval_steps_per_second": 39.232,
+      "step": 839520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6252795042856776e-05,
+      "loss": 3.6034,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.624440909534626e-05,
+      "loss": 3.6007,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.623602314783574e-05,
+      "loss": 3.6085,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.622763720032522e-05,
+      "loss": 3.6136,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.62192512528147e-05,
+      "loss": 3.6192,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621086530530418e-05,
+      "loss": 3.6002,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.620249573659739e-05,
+      "loss": 3.6035,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.619410978908687e-05,
+      "loss": 3.603,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.618572384157635e-05,
+      "loss": 3.6068,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.617733789406583e-05,
+      "loss": 3.6093,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6168951946555305e-05,
+      "loss": 3.5997,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6160565999044785e-05,
+      "loss": 3.6036,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6152196430338e-05,
+      "loss": 3.6113,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.614381048282748e-05,
+      "loss": 3.5958,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.613542453531696e-05,
+      "loss": 3.5986,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.612703858780644e-05,
+      "loss": 3.5904,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611865264029592e-05,
+      "loss": 3.5904,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.61102666927854e-05,
+      "loss": 3.6022,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.610188074527488e-05,
+      "loss": 3.5972,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.609349479776436e-05,
+      "loss": 3.6139,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.608512522905757e-05,
+      "loss": 3.6113,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.607675566035078e-05,
+      "loss": 3.6027,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606836971284026e-05,
+      "loss": 3.6071,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605998376532974e-05,
+      "loss": 3.605,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605159781781922e-05,
+      "loss": 3.589,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.60432118703087e-05,
+      "loss": 3.6067,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6034825922798185e-05,
+      "loss": 3.6028,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6026456354091394e-05,
+      "loss": 3.5875,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6018070406580874e-05,
+      "loss": 3.5956,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6009684459070354e-05,
+      "loss": 3.5946,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6001298511559834e-05,
+      "loss": 3.6023,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5992912564049314e-05,
+      "loss": 3.6005,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.598454299534252e-05,
+      "loss": 3.607,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5976157047832e-05,
+      "loss": 3.6024,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.596777110032148e-05,
+      "loss": 3.5907,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595938515281096e-05,
+      "loss": 3.6084,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595099920530044e-05,
+      "loss": 3.6091,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.594261325778992e-05,
+      "loss": 3.5932,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59342273102794e-05,
+      "loss": 3.5953,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.592584136276888e-05,
+      "loss": 3.5929,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.591748817286583e-05,
+      "loss": 3.5888,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590910222535531e-05,
+      "loss": 3.6082,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590071627784479e-05,
+      "loss": 3.6041,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589233033033427e-05,
+      "loss": 3.603,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.588394438282375e-05,
+      "loss": 3.5984,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.587555843531323e-05,
+      "loss": 3.6017,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.586717248780271e-05,
+      "loss": 3.601,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585878654029219e-05,
+      "loss": 3.5992,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5850433350389126e-05,
+      "loss": 3.5759,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5842047402878606e-05,
+      "loss": 3.6088,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.583366145536809e-05,
+      "loss": 3.5926,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.582527550785757e-05,
+      "loss": 3.6073,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.581688956034705e-05,
+      "loss": 3.5981,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580850361283653e-05,
+      "loss": 3.5943,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580013404412974e-05,
+      "loss": 3.5868,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579174809661922e-05,
+      "loss": 3.5973,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.57833621491087e-05,
+      "loss": 3.6086,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.577497620159818e-05,
+      "loss": 3.5952,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.576659025408766e-05,
+      "loss": 3.615,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.575820430657714e-05,
+      "loss": 3.5967,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5749818359066615e-05,
+      "loss": 3.5797,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5741432411556095e-05,
+      "loss": 3.6093,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.573306284284931e-05,
+      "loss": 3.5905,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.572467689533879e-05,
+      "loss": 3.5864,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5716307326632006e-05,
+      "loss": 3.5988,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5707921379121486e-05,
+      "loss": 3.6004,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5699535431610966e-05,
+      "loss": 3.5828,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569114948410044e-05,
+      "loss": 3.5916,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5682779915393655e-05,
+      "loss": 3.5821,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5674393967883135e-05,
+      "loss": 3.5963,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5666008020372615e-05,
+      "loss": 3.5983,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.565762207286209e-05,
+      "loss": 3.5931,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564923612535157e-05,
+      "loss": 3.5981,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564085017784105e-05,
+      "loss": 3.5866,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563246423033053e-05,
+      "loss": 3.606,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.562407828282001e-05,
+      "loss": 3.5932,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5615708714113224e-05,
+      "loss": 3.6071,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5607322766602704e-05,
+      "loss": 3.592,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5598936819092184e-05,
+      "loss": 3.5933,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559056725038539e-05,
+      "loss": 3.6002,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558218130287487e-05,
+      "loss": 3.6013,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.557379535536435e-05,
+      "loss": 3.5976,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.556540940785383e-05,
+      "loss": 3.6001,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.555702346034331e-05,
+      "loss": 3.5866,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554863751283279e-05,
+      "loss": 3.5883,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554025156532227e-05,
+      "loss": 3.5939,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553186561781175e-05,
+      "loss": 3.6007,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.552349604910496e-05,
+      "loss": 3.5871,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.551511010159445e-05,
+      "loss": 3.5901,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.550672415408393e-05,
+      "loss": 3.5914,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.549833820657341e-05,
+      "loss": 3.5942,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548996863786662e-05,
+      "loss": 3.5973,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.54815826903561e-05,
+      "loss": 3.5846,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.547319674284558e-05,
+      "loss": 3.5936,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.546481079533506e-05,
+      "loss": 3.6026,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5456441226628266e-05,
+      "loss": 3.6006,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5448071657921476e-05,
+      "loss": 3.5745,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5439685710410956e-05,
+      "loss": 3.588,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5431299762900436e-05,
+      "loss": 3.5861,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5422913815389915e-05,
+      "loss": 3.5998,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.54145278678794e-05,
+      "loss": 3.5986,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.540614192036888e-05,
+      "loss": 3.5921,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.539775597285836e-05,
+      "loss": 3.588,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538937002534784e-05,
+      "loss": 3.5913,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538100045664105e-05,
+      "loss": 3.5859,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.537261450913053e-05,
+      "loss": 3.5937,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.536422856162001e-05,
+      "loss": 3.5865,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.535584261410949e-05,
+      "loss": 3.5929,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.534748942420643e-05,
+      "loss": 3.5874,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.533910347669591e-05,
+      "loss": 3.5762,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.533071752918539e-05,
+      "loss": 3.5845,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532233158167487e-05,
+      "loss": 3.5853,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5313945634164356e-05,
+      "loss": 3.5941,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5305559686653836e-05,
+      "loss": 3.5862,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5297173739143316e-05,
+      "loss": 3.5931,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5288787791632796e-05,
+      "loss": 3.5933,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5280418222926005e-05,
+      "loss": 3.5907,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5272032275415485e-05,
+      "loss": 3.5932,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5263646327904965e-05,
+      "loss": 3.5898,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5255260380394445e-05,
+      "loss": 3.5913,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5246874432883925e-05,
+      "loss": 3.5868,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52384884853734e-05,
+      "loss": 3.5825,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523010253786288e-05,
+      "loss": 3.5889,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5221732969156094e-05,
+      "loss": 3.5881,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5213347021645574e-05,
+      "loss": 3.5925,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5204961074135054e-05,
+      "loss": 3.5896,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5196575126624534e-05,
+      "loss": 3.5839,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5188189179114013e-05,
+      "loss": 3.5845,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517981961040722e-05,
+      "loss": 3.5766,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.51714336628967e-05,
+      "loss": 3.5965,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.516304771538618e-05,
+      "loss": 3.588,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.515466176787566e-05,
+      "loss": 3.5994,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.514627582036514e-05,
+      "loss": 3.5799,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.513790625165835e-05,
+      "loss": 3.5961,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512952030414783e-05,
+      "loss": 3.5935,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512113435663731e-05,
+      "loss": 3.5836,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.511274840912679e-05,
+      "loss": 3.5844,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.510437884042001e-05,
+      "loss": 3.5851,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.509599289290949e-05,
+      "loss": 3.5996,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.508760694539897e-05,
+      "loss": 3.5972,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.507922099788845e-05,
+      "loss": 3.5851,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5070851429181656e-05,
+      "loss": 3.5815,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5062465481671136e-05,
+      "loss": 3.5909,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5054079534160616e-05,
+      "loss": 3.594,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5045693586650096e-05,
+      "loss": 3.5829,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5037324017943305e-05,
+      "loss": 3.599,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5028938070432785e-05,
+      "loss": 3.5841,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5020552122922265e-05,
+      "loss": 3.5952,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5012166175411745e-05,
+      "loss": 3.5866,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8959591388702393,
+      "eval_runtime": 304.3358,
+      "eval_samples_per_second": 1253.849,
+      "eval_steps_per_second": 39.184,
+      "step": 915840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5003780227901225e-05,
+      "loss": 3.5898,
+      "step": 915968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.499539428039071e-05,
+      "loss": 3.5781,
+      "step": 916480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.498700833288019e-05,
+      "loss": 3.5885,
+      "step": 916992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.49786387641734e-05,
+      "loss": 3.5934,
+      "step": 917504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.497025281666288e-05,
+      "loss": 3.5979,
+      "step": 918016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.496186686915236e-05,
+      "loss": 3.5862,
+      "step": 918528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.495348092164184e-05,
+      "loss": 3.5818,
+      "step": 919040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.494509497413132e-05,
+      "loss": 3.5843,
+      "step": 919552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.49367090266208e-05,
+      "loss": 3.5839,
+      "step": 920064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.492832307911028e-05,
+      "loss": 3.5874,
+      "step": 920576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.491995351040349e-05,
+      "loss": 3.5835,
+      "step": 921088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.491156756289297e-05,
+      "loss": 3.584,
+      "step": 921600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.490319799418618e-05,
+      "loss": 3.5942,
+      "step": 922112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4894812046675665e-05,
+      "loss": 3.5737,
+      "step": 922624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4886426099165145e-05,
+      "loss": 3.5808,
+      "step": 923136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4878040151654625e-05,
+      "loss": 3.5741,
+      "step": 923648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4869670582947834e-05,
+      "loss": 3.5707,
+      "step": 924160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4861284635437314e-05,
+      "loss": 3.5837,
+      "step": 924672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4852898687926794e-05,
+      "loss": 3.577,
+      "step": 925184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4844512740416274e-05,
+      "loss": 3.5933,
+      "step": 925696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4836126792905754e-05,
+      "loss": 3.5945,
+      "step": 926208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.482775722419896e-05,
+      "loss": 3.5859,
+      "step": 926720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.481937127668844e-05,
+      "loss": 3.5863,
+      "step": 927232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.481098532917792e-05,
+      "loss": 3.5862,
+      "step": 927744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.48025993816674e-05,
+      "loss": 3.5677,
+      "step": 928256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.479421343415688e-05,
+      "loss": 3.5913,
+      "step": 928768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.478582748664636e-05,
+      "loss": 3.5786,
+      "step": 929280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.477745791793958e-05,
+      "loss": 3.5729,
+      "step": 929792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.476907197042906e-05,
+      "loss": 3.579,
+      "step": 930304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.476068602291853e-05,
+      "loss": 3.5729,
+      "step": 930816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.475230007540801e-05,
+      "loss": 3.5809,
+      "step": 931328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.474391412789749e-05,
+      "loss": 3.5787,
+      "step": 931840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.473554455919071e-05,
+      "loss": 3.5931,
+      "step": 932352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.472715861168018e-05,
+      "loss": 3.5812,
+      "step": 932864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.471877266416966e-05,
+      "loss": 3.5757,
+      "step": 933376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.471038671665914e-05,
+      "loss": 3.5868,
+      "step": 933888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.470200076914862e-05,
+      "loss": 3.5895,
+      "step": 934400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.46936148216381e-05,
+      "loss": 3.576,
+      "step": 934912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.468522887412759e-05,
+      "loss": 3.5732,
+      "step": 935424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.46768593054208e-05,
+      "loss": 3.5743,
+      "step": 935936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4668489736714006e-05,
+      "loss": 3.5748,
+      "step": 936448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4660103789203486e-05,
+      "loss": 3.5847,
+      "step": 936960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4651717841692966e-05,
+      "loss": 3.5835,
+      "step": 937472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4643331894182446e-05,
+      "loss": 3.5866,
+      "step": 937984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4634945946671926e-05,
+      "loss": 3.5817,
+      "step": 938496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4626559999161406e-05,
+      "loss": 3.5789,
+      "step": 939008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4618174051650886e-05,
+      "loss": 3.5817,
+      "step": 939520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4609804482944095e-05,
+      "loss": 3.579,
+      "step": 940032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4601418535433575e-05,
+      "loss": 3.5567,
+      "step": 940544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4593032587923055e-05,
+      "loss": 3.5883,
+      "step": 941056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.458464664041254e-05,
+      "loss": 3.5765,
+      "step": 941568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.457627707170575e-05,
+      "loss": 3.5881,
+      "step": 942080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.456789112419523e-05,
+      "loss": 3.5779,
+      "step": 942592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.455950517668471e-05,
+      "loss": 3.578,
+      "step": 943104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.455111922917419e-05,
+      "loss": 3.565,
+      "step": 943616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.45427496604674e-05,
+      "loss": 3.5763,
+      "step": 944128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.453436371295688e-05,
+      "loss": 3.5943,
+      "step": 944640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.452597776544636e-05,
+      "loss": 3.5761,
+      "step": 945152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.451759181793584e-05,
+      "loss": 3.5944,
+      "step": 945664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.450920587042532e-05,
+      "loss": 3.579,
+      "step": 946176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.45008199229148e-05,
+      "loss": 3.5588,
+      "step": 946688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.449243397540428e-05,
+      "loss": 3.595,
+      "step": 947200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.448404802789376e-05,
+      "loss": 3.5711,
+      "step": 947712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4475678459186975e-05,
+      "loss": 3.5647,
+      "step": 948224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4467308890480184e-05,
+      "loss": 3.582,
+      "step": 948736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4458922942969664e-05,
+      "loss": 3.5806,
+      "step": 949248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4450536995459144e-05,
+      "loss": 3.5661,
+      "step": 949760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4442151047948624e-05,
+      "loss": 3.5705,
+      "step": 950272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4433765100438104e-05,
+      "loss": 3.5662,
+      "step": 950784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4425379152927584e-05,
+      "loss": 3.5702,
+      "step": 951296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4416993205417064e-05,
+      "loss": 3.5846,
+      "step": 951808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.440860725790654e-05,
+      "loss": 3.5747,
+      "step": 952320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.440023768919975e-05,
+      "loss": 3.5776,
+      "step": 952832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.439186812049296e-05,
+      "loss": 3.5672,
+      "step": 953344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.438348217298245e-05,
+      "loss": 3.589,
+      "step": 953856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.437509622547193e-05,
+      "loss": 3.5741,
+      "step": 954368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.436671027796141e-05,
+      "loss": 3.5875,
+      "step": 954880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.435834070925462e-05,
+      "loss": 3.5718,
+      "step": 955392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.43499547617441e-05,
+      "loss": 3.5755,
+      "step": 955904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.434156881423358e-05,
+      "loss": 3.5837,
+      "step": 956416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.433318286672306e-05,
+      "loss": 3.5797,
+      "step": 956928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.432479691921254e-05,
+      "loss": 3.5773,
+      "step": 957440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4316427350505747e-05,
+      "loss": 3.582,
+      "step": 957952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4308041402995226e-05,
+      "loss": 3.57,
+      "step": 958464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4299655455484706e-05,
+      "loss": 3.5677,
+      "step": 958976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4291269507974186e-05,
+      "loss": 3.5767,
+      "step": 959488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4282883560463666e-05,
+      "loss": 3.5816,
+      "step": 960000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4274497612953146e-05,
+      "loss": 3.5701,
+      "step": 960512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4266111665442626e-05,
+      "loss": 3.5758,
+      "step": 961024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.425774209673584e-05,
+      "loss": 3.5712,
+      "step": 961536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4249356149225315e-05,
+      "loss": 3.5728,
+      "step": 962048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4240970201714795e-05,
+      "loss": 3.5793,
+      "step": 962560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4232584254204275e-05,
+      "loss": 3.5679,
+      "step": 963072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4224198306693755e-05,
+      "loss": 3.5734,
+      "step": 963584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4215812359183235e-05,
+      "loss": 3.5853,
+      "step": 964096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4207426411672715e-05,
+      "loss": 3.5871,
+      "step": 964608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4199056842965924e-05,
+      "loss": 3.5527,
+      "step": 965120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4190670895455404e-05,
+      "loss": 3.571,
+      "step": 965632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4182284947944884e-05,
+      "loss": 3.5652,
+      "step": 966144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4173899000434364e-05,
+      "loss": 3.5795,
+      "step": 966656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.416551305292385e-05,
+      "loss": 3.5818,
+      "step": 967168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.415714348421706e-05,
+      "loss": 3.5722,
+      "step": 967680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.414875753670654e-05,
+      "loss": 3.5694,
+      "step": 968192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.414037158919602e-05,
+      "loss": 3.5749,
+      "step": 968704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.41319856416855e-05,
+      "loss": 3.5681,
+      "step": 969216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.412359969417498e-05,
+      "loss": 3.5739,
+      "step": 969728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.411521374666446e-05,
+      "loss": 3.5653,
+      "step": 970240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.410682779915394e-05,
+      "loss": 3.5762,
+      "step": 970752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.409844185164342e-05,
+      "loss": 3.5664,
+      "step": 971264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.409007228293663e-05,
+      "loss": 3.56,
+      "step": 971776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.408168633542611e-05,
+      "loss": 3.5656,
+      "step": 972288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.407330038791559e-05,
+      "loss": 3.5636,
+      "step": 972800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.406491444040507e-05,
+      "loss": 3.5779,
+      "step": 973312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.405652849289455e-05,
+      "loss": 3.5647,
+      "step": 973824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.404814254538403e-05,
+      "loss": 3.5738,
+      "step": 974336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4039772976677244e-05,
+      "loss": 3.5763,
+      "step": 974848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4031387029166724e-05,
+      "loss": 3.5733,
+      "step": 975360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4023001081656204e-05,
+      "loss": 3.5698,
+      "step": 975872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.401461513414568e-05,
+      "loss": 3.5734,
+      "step": 976384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.400624556543889e-05,
+      "loss": 3.5743,
+      "step": 976896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.399785961792837e-05,
+      "loss": 3.5656,
+      "step": 977408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3989473670417846e-05,
+      "loss": 3.5599,
+      "step": 977920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3981087722907326e-05,
+      "loss": 3.5723,
+      "step": 978432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.397271815420054e-05,
+      "loss": 3.5686,
+      "step": 978944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.396433220669002e-05,
+      "loss": 3.5733,
+      "step": 979456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.39559462591795e-05,
+      "loss": 3.5698,
+      "step": 979968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.394756031166898e-05,
+      "loss": 3.5673,
+      "step": 980480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393917436415846e-05,
+      "loss": 3.5681,
+      "step": 980992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393080479545168e-05,
+      "loss": 3.5599,
+      "step": 981504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.392241884794115e-05,
+      "loss": 3.5739,
+      "step": 982016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.391403290043063e-05,
+      "loss": 3.5702,
+      "step": 982528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.390564695292011e-05,
+      "loss": 3.5832,
+      "step": 983040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.389726100540959e-05,
+      "loss": 3.5633,
+      "step": 983552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.38888914367028e-05,
+      "loss": 3.5732,
+      "step": 984064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.388050548919228e-05,
+      "loss": 3.5768,
+      "step": 984576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.387211954168176e-05,
+      "loss": 3.5673,
+      "step": 985088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.386373359417124e-05,
+      "loss": 3.5634,
+      "step": 985600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.385534764666073e-05,
+      "loss": 3.5684,
+      "step": 986112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.384696169915021e-05,
+      "loss": 3.5808,
+      "step": 986624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3838575751639687e-05,
+      "loss": 3.5781,
+      "step": 987136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3830206182932896e-05,
+      "loss": 3.5685,
+      "step": 987648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3821820235422376e-05,
+      "loss": 3.5621,
+      "step": 988160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3813434287911856e-05,
+      "loss": 3.5734,
+      "step": 988672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3805048340401336e-05,
+      "loss": 3.574,
+      "step": 989184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3796678771694545e-05,
+      "loss": 3.561,
+      "step": 989696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3788292824184025e-05,
+      "loss": 3.5874,
+      "step": 990208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3779906876673505e-05,
+      "loss": 3.5643,
+      "step": 990720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3771520929162985e-05,
+      "loss": 3.5759,
+      "step": 991232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3763134981652464e-05,
+      "loss": 3.5697,
+      "step": 991744
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8969345092773438,
+      "eval_runtime": 304.8845,
+      "eval_samples_per_second": 1251.592,
+      "eval_steps_per_second": 39.113,
+      "step": 992160
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 6.437366363528755e+17,
+  "trial_name": null,
+  "trial_params": null
+}