diff --git "a/checkpoint-2136960/trainer_state.json" "b/checkpoint-2136960/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2136960/trainer_state.json"
@@ -0,0 +1,25287 @@
+{
+  "best_metric": 3.8959591388702393,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/pp-mod-subj2/transformer/0/checkpoints/checkpoint-915840",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 2136960,
+  "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
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3754749034141944e-05,
+      "loss": 3.595,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3746363086631424e-05,
+      "loss": 3.5607,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3737977139120904e-05,
+      "loss": 3.5675,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372959119161039e-05,
+      "loss": 3.5779,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3721205244099864e-05,
+      "loss": 3.5732,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3712819296589344e-05,
+      "loss": 3.5714,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.370444972788256e-05,
+      "loss": 3.5631,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.369606378037203e-05,
+      "loss": 3.5638,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.368767783286151e-05,
+      "loss": 3.5666,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367929188535099e-05,
+      "loss": 3.5718,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367090593784047e-05,
+      "loss": 3.5657,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.366251999032995e-05,
+      "loss": 3.5705,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.365415042162316e-05,
+      "loss": 3.575,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.364576447411264e-05,
+      "loss": 3.5545,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.363737852660213e-05,
+      "loss": 3.5658,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.362899257909161e-05,
+      "loss": 3.5511,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.362060663158109e-05,
+      "loss": 3.5522,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.361222068407057e-05,
+      "loss": 3.5697,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.360385111536378e-05,
+      "loss": 3.5555,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.359546516785326e-05,
+      "loss": 3.5747,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.358707922034274e-05,
+      "loss": 3.578,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357869327283222e-05,
+      "loss": 3.5669,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.35703073253217e-05,
+      "loss": 3.5687,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.356192137781118e-05,
+      "loss": 3.5671,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.355355180910439e-05,
+      "loss": 3.5528,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.354516586159387e-05,
+      "loss": 3.568,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.353677991408335e-05,
+      "loss": 3.5617,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3528393966572827e-05,
+      "loss": 3.5572,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.352000801906231e-05,
+      "loss": 3.5605,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351162207155179e-05,
+      "loss": 3.5525,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.350323612404127e-05,
+      "loss": 3.561,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.349486655533448e-05,
+      "loss": 3.5654,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.348648060782396e-05,
+      "loss": 3.5725,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347809466031344e-05,
+      "loss": 3.561,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346970871280292e-05,
+      "loss": 3.5577,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346133914409613e-05,
+      "loss": 3.5627,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.345295319658561e-05,
+      "loss": 3.5771,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.344456724907509e-05,
+      "loss": 3.5594,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.343618130156457e-05,
+      "loss": 3.5543,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.342779535405405e-05,
+      "loss": 3.5563,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341942578534727e-05,
+      "loss": 3.5599,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341103983783675e-05,
+      "loss": 3.5639,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.340265389032623e-05,
+      "loss": 3.571,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.33942679428157e-05,
+      "loss": 3.5664,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.338588199530518e-05,
+      "loss": 3.5625,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3377512426598396e-05,
+      "loss": 3.5625,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336912647908787e-05,
+      "loss": 3.5633,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336074053157735e-05,
+      "loss": 3.5644,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.335235458406683e-05,
+      "loss": 3.5438,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.334396863655631e-05,
+      "loss": 3.5678,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.333559906784952e-05,
+      "loss": 3.5601,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3327213120339005e-05,
+      "loss": 3.5687,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3318827172828485e-05,
+      "loss": 3.5591,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3310441225317965e-05,
+      "loss": 3.5589,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3302071656611174e-05,
+      "loss": 3.5549,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3293685709100654e-05,
+      "loss": 3.5531,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3285299761590134e-05,
+      "loss": 3.5752,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3276913814079614e-05,
+      "loss": 3.5567,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.326854424537282e-05,
+      "loss": 3.5821,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.32601582978623e-05,
+      "loss": 3.5576,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325177235035178e-05,
+      "loss": 3.5456,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.324338640284126e-05,
+      "loss": 3.5769,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.323501683413447e-05,
+      "loss": 3.5557,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.322663088662396e-05,
+      "loss": 3.5459,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321824493911344e-05,
+      "loss": 3.5589,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320985899160292e-05,
+      "loss": 3.5662,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.32014730440924e-05,
+      "loss": 3.5474,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.319310347538561e-05,
+      "loss": 3.5546,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.318471752787509e-05,
+      "loss": 3.5483,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.317633158036457e-05,
+      "loss": 3.5485,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.316794563285405e-05,
+      "loss": 3.5704,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315955968534353e-05,
+      "loss": 3.5588,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315117373783301e-05,
+      "loss": 3.5595,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3142804169126216e-05,
+      "loss": 3.5506,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3134418221615696e-05,
+      "loss": 3.5681,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3126032274105176e-05,
+      "loss": 3.5581,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3117646326594656e-05,
+      "loss": 3.5695,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310927675788787e-05,
+      "loss": 3.5568,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310089081037735e-05,
+      "loss": 3.5556,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309250486286683e-05,
+      "loss": 3.5678,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.308411891535631e-05,
+      "loss": 3.5601,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.307573296784579e-05,
+      "loss": 3.5625,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3067363399139e-05,
+      "loss": 3.5633,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.305897745162848e-05,
+      "loss": 3.5582,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.305059150411796e-05,
+      "loss": 3.5482,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304220555660744e-05,
+      "loss": 3.557,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.303383598790065e-05,
+      "loss": 3.5637,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.302545004039013e-05,
+      "loss": 3.5533,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.301706409287961e-05,
+      "loss": 3.556,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300867814536909e-05,
+      "loss": 3.5553,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3000292197858576e-05,
+      "loss": 3.5554,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2991922629151786e-05,
+      "loss": 3.5598,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2983536681641266e-05,
+      "loss": 3.5534,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2975150734130746e-05,
+      "loss": 3.5587,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2966764786620225e-05,
+      "loss": 3.5658,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2958378839109705e-05,
+      "loss": 3.5687,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2950009270402915e-05,
+      "loss": 3.5359,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2941623322892394e-05,
+      "loss": 3.5525,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2933237375381874e-05,
+      "loss": 3.5506,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2924851427871354e-05,
+      "loss": 3.5619,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.291646548036083e-05,
+      "loss": 3.5662,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2908095911654043e-05,
+      "loss": 3.5535,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289970996414353e-05,
+      "loss": 3.5488,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2891324016633e-05,
+      "loss": 3.5575,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.288293806912248e-05,
+      "loss": 3.5536,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.287455212161196e-05,
+      "loss": 3.5562,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.286616617410144e-05,
+      "loss": 3.5477,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.285779660539465e-05,
+      "loss": 3.5567,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284941065788413e-05,
+      "loss": 3.5513,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284102471037361e-05,
+      "loss": 3.5446,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.283263876286309e-05,
+      "loss": 3.5515,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.282425281535257e-05,
+      "loss": 3.5404,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.281586686784205e-05,
+      "loss": 3.558,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.280749729913527e-05,
+      "loss": 3.5531,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279911135162475e-05,
+      "loss": 3.5525,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279072540411423e-05,
+      "loss": 3.5598,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.278233945660371e-05,
+      "loss": 3.557,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.277395350909319e-05,
+      "loss": 3.5501,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.276556756158267e-05,
+      "loss": 3.5547,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.275719799287588e-05,
+      "loss": 3.5589,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.274881204536536e-05,
+      "loss": 3.5474,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.274042609785484e-05,
+      "loss": 3.5419,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.273204015034432e-05,
+      "loss": 3.557,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2723670581637526e-05,
+      "loss": 3.5547,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2715284634127006e-05,
+      "loss": 3.5552,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2706898686616486e-05,
+      "loss": 3.55,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2698512739105966e-05,
+      "loss": 3.5547,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2690126791595446e-05,
+      "loss": 3.5513,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268175722288866e-05,
+      "loss": 3.5433,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.267337127537814e-05,
+      "loss": 3.5465,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.266498532786762e-05,
+      "loss": 3.5614,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.26565993803571e-05,
+      "loss": 3.5652,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264822981165031e-05,
+      "loss": 3.5446,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263984386413979e-05,
+      "loss": 3.5519,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263145791662927e-05,
+      "loss": 3.5634,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.262307196911875e-05,
+      "loss": 3.5503,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.261470240041196e-05,
+      "loss": 3.5462,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.260631645290144e-05,
+      "loss": 3.5498,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.259793050539092e-05,
+      "loss": 3.5667,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.25895445578804e-05,
+      "loss": 3.5621,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2581158610369886e-05,
+      "loss": 3.5501,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2572789041663095e-05,
+      "loss": 3.546,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2564403094152575e-05,
+      "loss": 3.5577,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2556017146642055e-05,
+      "loss": 3.5575,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2547631199131535e-05,
+      "loss": 3.5449,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2539261630424744e-05,
+      "loss": 3.5697,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2530875682914224e-05,
+      "loss": 3.5483,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2522489735403704e-05,
+      "loss": 3.5598,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2514103787893184e-05,
+      "loss": 3.5533,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8981220722198486,
+      "eval_runtime": 308.0333,
+      "eval_samples_per_second": 1238.798,
+      "eval_steps_per_second": 38.713,
+      "step": 1068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.250571784038266e-05,
+      "loss": 3.5604,
+      "step": 1068544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.249734827167587e-05,
+      "loss": 3.541,
+      "step": 1069056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.248896232416535e-05,
+      "loss": 3.5511,
+      "step": 1069568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.248057637665484e-05,
+      "loss": 3.5624,
+      "step": 1070080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.247219042914431e-05,
+      "loss": 3.5593,
+      "step": 1070592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.246382086043753e-05,
+      "loss": 3.5543,
+      "step": 1071104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.245543491292701e-05,
+      "loss": 3.5506,
+      "step": 1071616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.244704896541649e-05,
+      "loss": 3.5485,
+      "step": 1072128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.243866301790596e-05,
+      "loss": 3.5475,
+      "step": 1072640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.243027707039544e-05,
+      "loss": 3.5566,
+      "step": 1073152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.242190750168866e-05,
+      "loss": 3.5473,
+      "step": 1073664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.241352155417813e-05,
+      "loss": 3.554,
+      "step": 1074176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.240513560666761e-05,
+      "loss": 3.5563,
+      "step": 1074688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.239674965915709e-05,
+      "loss": 3.5381,
+      "step": 1075200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.238838009045031e-05,
+      "loss": 3.548,
+      "step": 1075712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2379994142939787e-05,
+      "loss": 3.5342,
+      "step": 1076224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2371608195429267e-05,
+      "loss": 3.5384,
+      "step": 1076736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2363222247918746e-05,
+      "loss": 3.5509,
+      "step": 1077248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2354836300408226e-05,
+      "loss": 3.5385,
+      "step": 1077760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2346466731701436e-05,
+      "loss": 3.5568,
+      "step": 1078272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2338080784190915e-05,
+      "loss": 3.5631,
+      "step": 1078784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2329694836680395e-05,
+      "loss": 3.5454,
+      "step": 1079296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2321308889169875e-05,
+      "loss": 3.5532,
+      "step": 1079808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2312922941659355e-05,
+      "loss": 3.5519,
+      "step": 1080320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2304553372952564e-05,
+      "loss": 3.5379,
+      "step": 1080832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2296167425442044e-05,
+      "loss": 3.554,
+      "step": 1081344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.228778147793153e-05,
+      "loss": 3.5414,
+      "step": 1081856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.227939553042101e-05,
+      "loss": 3.5461,
+      "step": 1082368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.227100958291049e-05,
+      "loss": 3.5426,
+      "step": 1082880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.226262363539997e-05,
+      "loss": 3.5348,
+      "step": 1083392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.225423768788945e-05,
+      "loss": 3.5444,
+      "step": 1083904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.224586811918266e-05,
+      "loss": 3.5519,
+      "step": 1084416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.223748217167214e-05,
+      "loss": 3.5578,
+      "step": 1084928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.222909622416162e-05,
+      "loss": 3.5455,
+      "step": 1085440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.22207102766511e-05,
+      "loss": 3.5399,
+      "step": 1085952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.221234070794431e-05,
+      "loss": 3.5461,
+      "step": 1086464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.220395476043379e-05,
+      "loss": 3.562,
+      "step": 1086976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.219556881292327e-05,
+      "loss": 3.54,
+      "step": 1087488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.218718286541275e-05,
+      "loss": 3.5419,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2178813296705965e-05,
+      "loss": 3.5414,
+      "step": 1088512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2170427349195445e-05,
+      "loss": 3.5402,
+      "step": 1089024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2162041401684925e-05,
+      "loss": 3.5452,
+      "step": 1089536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2153655454174405e-05,
+      "loss": 3.5522,
+      "step": 1090048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2145285885467614e-05,
+      "loss": 3.5547,
+      "step": 1090560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2136899937957094e-05,
+      "loss": 3.5457,
+      "step": 1091072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2128513990446574e-05,
+      "loss": 3.5446,
+      "step": 1091584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2120128042936054e-05,
+      "loss": 3.5453,
+      "step": 1092096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211175847422926e-05,
+      "loss": 3.5526,
+      "step": 1092608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.210337252671874e-05,
+      "loss": 3.5237,
+      "step": 1093120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.209498657920822e-05,
+      "loss": 3.55,
+      "step": 1093632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20866006316977e-05,
+      "loss": 3.5474,
+      "step": 1094144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.207823106299092e-05,
+      "loss": 3.5515,
+      "step": 1094656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20698451154804e-05,
+      "loss": 3.5437,
+      "step": 1095168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.206145916796988e-05,
+      "loss": 3.537,
+      "step": 1095680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.205307322045936e-05,
+      "loss": 3.5443,
+      "step": 1096192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.204468727294884e-05,
+      "loss": 3.5362,
+      "step": 1096704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.203631770424205e-05,
+      "loss": 3.5559,
+      "step": 1097216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.202793175673153e-05,
+      "loss": 3.5442,
+      "step": 1097728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201954580922101e-05,
+      "loss": 3.5625,
+      "step": 1098240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201115986171049e-05,
+      "loss": 3.5417,
+      "step": 1098752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2002790293003696e-05,
+      "loss": 3.5277,
+      "step": 1099264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1994404345493176e-05,
+      "loss": 3.5596,
+      "step": 1099776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1986018397982656e-05,
+      "loss": 3.5374,
+      "step": 1100288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1977632450472136e-05,
+      "loss": 3.5325,
+      "step": 1100800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.196926288176535e-05,
+      "loss": 3.5437,
+      "step": 1101312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.196087693425483e-05,
+      "loss": 3.5461,
+      "step": 1101824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.195249098674431e-05,
+      "loss": 3.5296,
+      "step": 1102336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.194410503923379e-05,
+      "loss": 3.5374,
+      "step": 1102848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1935735470527e-05,
+      "loss": 3.5331,
+      "step": 1103360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.192734952301648e-05,
+      "loss": 3.5331,
+      "step": 1103872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.191896357550596e-05,
+      "loss": 3.5519,
+      "step": 1104384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.191057762799544e-05,
+      "loss": 3.543,
+      "step": 1104896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.190220805928865e-05,
+      "loss": 3.541,
+      "step": 1105408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.189382211177813e-05,
+      "loss": 3.538,
+      "step": 1105920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.188543616426761e-05,
+      "loss": 3.5513,
+      "step": 1106432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.187705021675709e-05,
+      "loss": 3.54,
+      "step": 1106944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1868680648050306e-05,
+      "loss": 3.5513,
+      "step": 1107456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1860294700539786e-05,
+      "loss": 3.5413,
+      "step": 1107968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1851908753029266e-05,
+      "loss": 3.5407,
+      "step": 1108480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.184352280551874e-05,
+      "loss": 3.5478,
+      "step": 1108992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1835153236811955e-05,
+      "loss": 3.5467,
+      "step": 1109504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1826767289301435e-05,
+      "loss": 3.5442,
+      "step": 1110016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1818381341790915e-05,
+      "loss": 3.5481,
+      "step": 1110528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.180999539428039e-05,
+      "loss": 3.5465,
+      "step": 1111040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1801625825573604e-05,
+      "loss": 3.5285,
+      "step": 1111552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1793239878063084e-05,
+      "loss": 3.5383,
+      "step": 1112064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1784853930552564e-05,
+      "loss": 3.5492,
+      "step": 1112576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1776467983042044e-05,
+      "loss": 3.5352,
+      "step": 1113088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.176809841433526e-05,
+      "loss": 3.5393,
+      "step": 1113600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.175971246682474e-05,
+      "loss": 3.5373,
+      "step": 1114112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.175132651931421e-05,
+      "loss": 3.5396,
+      "step": 1114624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.174294057180369e-05,
+      "loss": 3.5404,
+      "step": 1115136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.173457100309691e-05,
+      "loss": 3.5397,
+      "step": 1115648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.172618505558639e-05,
+      "loss": 3.5433,
+      "step": 1116160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.171779910807586e-05,
+      "loss": 3.5446,
+      "step": 1116672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170941316056534e-05,
+      "loss": 3.5553,
+      "step": 1117184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170104359185856e-05,
+      "loss": 3.5208,
+      "step": 1117696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.169265764434804e-05,
+      "loss": 3.5389,
+      "step": 1118208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.168427169683752e-05,
+      "loss": 3.535,
+      "step": 1118720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1675885749327e-05,
+      "loss": 3.5433,
+      "step": 1119232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.166749980181648e-05,
+      "loss": 3.5449,
+      "step": 1119744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1659130233109686e-05,
+      "loss": 3.5418,
+      "step": 1120256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1650744285599166e-05,
+      "loss": 3.5315,
+      "step": 1120768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1642358338088646e-05,
+      "loss": 3.5424,
+      "step": 1121280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1633972390578126e-05,
+      "loss": 3.5408,
+      "step": 1121792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1625602821871335e-05,
+      "loss": 3.5363,
+      "step": 1122304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1617216874360815e-05,
+      "loss": 3.531,
+      "step": 1122816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1608830926850295e-05,
+      "loss": 3.5387,
+      "step": 1123328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1600444979339775e-05,
+      "loss": 3.5404,
+      "step": 1123840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1592075410632984e-05,
+      "loss": 3.5257,
+      "step": 1124352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.158368946312247e-05,
+      "loss": 3.5364,
+      "step": 1124864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.157530351561195e-05,
+      "loss": 3.5249,
+      "step": 1125376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.156691756810143e-05,
+      "loss": 3.5452,
+      "step": 1125888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.155854799939464e-05,
+      "loss": 3.5343,
+      "step": 1126400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.155016205188412e-05,
+      "loss": 3.536,
+      "step": 1126912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.15417761043736e-05,
+      "loss": 3.5429,
+      "step": 1127424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.153339015686308e-05,
+      "loss": 3.5416,
+      "step": 1127936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.152502058815629e-05,
+      "loss": 3.5353,
+      "step": 1128448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.151663464064577e-05,
+      "loss": 3.5406,
+      "step": 1128960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.150824869313525e-05,
+      "loss": 3.5432,
+      "step": 1129472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149986274562473e-05,
+      "loss": 3.5292,
+      "step": 1129984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149147679811421e-05,
+      "loss": 3.5274,
+      "step": 1130496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1483107229407425e-05,
+      "loss": 3.5413,
+      "step": 1131008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1474721281896905e-05,
+      "loss": 3.5431,
+      "step": 1131520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1466335334386384e-05,
+      "loss": 3.5402,
+      "step": 1132032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1457949386875864e-05,
+      "loss": 3.5342,
+      "step": 1132544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1449579818169074e-05,
+      "loss": 3.5378,
+      "step": 1133056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1441193870658553e-05,
+      "loss": 3.5324,
+      "step": 1133568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1432807923148033e-05,
+      "loss": 3.525,
+      "step": 1134080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1424421975637513e-05,
+      "loss": 3.5312,
+      "step": 1134592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.141605240693072e-05,
+      "loss": 3.5487,
+      "step": 1135104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.14076664594202e-05,
+      "loss": 3.5482,
+      "step": 1135616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.139928051190968e-05,
+      "loss": 3.5323,
+      "step": 1136128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.139091094320289e-05,
+      "loss": 3.5358,
+      "step": 1136640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.138252499569238e-05,
+      "loss": 3.5428,
+      "step": 1137152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.137413904818186e-05,
+      "loss": 3.5366,
+      "step": 1137664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.136575310067134e-05,
+      "loss": 3.5301,
+      "step": 1138176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.135738353196455e-05,
+      "loss": 3.5375,
+      "step": 1138688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.134899758445403e-05,
+      "loss": 3.5512,
+      "step": 1139200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.134061163694351e-05,
+      "loss": 3.5443,
+      "step": 1139712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133222568943299e-05,
+      "loss": 3.5361,
+      "step": 1140224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1323856120726196e-05,
+      "loss": 3.5297,
+      "step": 1140736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1315470173215676e-05,
+      "loss": 3.5421,
+      "step": 1141248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1307084225705156e-05,
+      "loss": 3.5406,
+      "step": 1141760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1298698278194636e-05,
+      "loss": 3.5284,
+      "step": 1142272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1290312330684116e-05,
+      "loss": 3.5541,
+      "step": 1142784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128194276197733e-05,
+      "loss": 3.5328,
+      "step": 1143296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.127355681446681e-05,
+      "loss": 3.5418,
+      "step": 1143808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.126517086695629e-05,
+      "loss": 3.5345,
+      "step": 1144320
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8996713161468506,
+      "eval_runtime": 302.3946,
+      "eval_samples_per_second": 1261.898,
+      "eval_steps_per_second": 39.435,
+      "step": 1144800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.125678491944577e-05,
+      "loss": 3.5376,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.124839897193525e-05,
+      "loss": 3.5308,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.124001302442473e-05,
+      "loss": 3.5359,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123162707691421e-05,
+      "loss": 3.5444,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.122324112940369e-05,
+      "loss": 3.5448,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.12148715606969e-05,
+      "loss": 3.5399,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.120648561318638e-05,
+      "loss": 3.5305,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.119809966567586e-05,
+      "loss": 3.5339,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118971371816534e-05,
+      "loss": 3.5323,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118134414945855e-05,
+      "loss": 3.5424,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.117295820194803e-05,
+      "loss": 3.5322,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.116457225443751e-05,
+      "loss": 3.5354,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1156186306926996e-05,
+      "loss": 3.5413,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1147816738220205e-05,
+      "loss": 3.5224,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1139430790709685e-05,
+      "loss": 3.5317,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1131044843199165e-05,
+      "loss": 3.5159,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1122658895688645e-05,
+      "loss": 3.5246,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1114289326981854e-05,
+      "loss": 3.5342,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1105903379471334e-05,
+      "loss": 3.5257,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1097517431960814e-05,
+      "loss": 3.5373,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.108913148445029e-05,
+      "loss": 3.5483,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.10807619157435e-05,
+      "loss": 3.5302,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107237596823298e-05,
+      "loss": 3.5362,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.106399002072246e-05,
+      "loss": 3.5338,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.105560407321194e-05,
+      "loss": 3.5297,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.104721812570142e-05,
+      "loss": 3.5339,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.103884855699464e-05,
+      "loss": 3.5219,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.103046260948412e-05,
+      "loss": 3.5322,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.102207666197359e-05,
+      "loss": 3.5287,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.101369071446307e-05,
+      "loss": 3.5201,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.100532114575629e-05,
+      "loss": 3.5346,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.099693519824576e-05,
+      "loss": 3.5272,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.098854925073524e-05,
+      "loss": 3.5411,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.098016330322472e-05,
+      "loss": 3.536,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.097179373451794e-05,
+      "loss": 3.5219,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.096340778700742e-05,
+      "loss": 3.529,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.09550218394969e-05,
+      "loss": 3.5482,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.094663589198638e-05,
+      "loss": 3.5221,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.093826632327959e-05,
+      "loss": 3.5277,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0929880375769066e-05,
+      "loss": 3.5254,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0921494428258546e-05,
+      "loss": 3.5281,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0913108480748026e-05,
+      "loss": 3.5259,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0904738912041235e-05,
+      "loss": 3.5438,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0896352964530715e-05,
+      "loss": 3.5351,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0887967017020195e-05,
+      "loss": 3.528,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0879581069509675e-05,
+      "loss": 3.5334,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087121150080289e-05,
+      "loss": 3.5256,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086282555329237e-05,
+      "loss": 3.5357,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.085443960578185e-05,
+      "loss": 3.5096,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.084605365827133e-05,
+      "loss": 3.5335,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.083768408956454e-05,
+      "loss": 3.5323,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.082929814205402e-05,
+      "loss": 3.5363,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.08209121945435e-05,
+      "loss": 3.5301,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.081252624703298e-05,
+      "loss": 3.5189,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.080415667832619e-05,
+      "loss": 3.5303,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.079577073081567e-05,
+      "loss": 3.5191,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.078738478330515e-05,
+      "loss": 3.5444,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.077899883579463e-05,
+      "loss": 3.5273,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0770629267087844e-05,
+      "loss": 3.5462,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0762243319577324e-05,
+      "loss": 3.526,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0753857372066804e-05,
+      "loss": 3.5188,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0745471424556284e-05,
+      "loss": 3.5372,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.073710185584949e-05,
+      "loss": 3.5224,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072871590833897e-05,
+      "loss": 3.5232,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072032996082845e-05,
+      "loss": 3.5223,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071194401331793e-05,
+      "loss": 3.5359,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.070355806580741e-05,
+      "loss": 3.5151,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.069518849710062e-05,
+      "loss": 3.5231,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.06868025495901e-05,
+      "loss": 3.5168,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.067841660207958e-05,
+      "loss": 3.5191,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.06700470333728e-05,
+      "loss": 3.535,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066166108586228e-05,
+      "loss": 3.5337,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.065327513835176e-05,
+      "loss": 3.5217,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.064488919084124e-05,
+      "loss": 3.5237,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.063650324333072e-05,
+      "loss": 3.5326,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.062813367462393e-05,
+      "loss": 3.5256,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.061974772711341e-05,
+      "loss": 3.5336,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.061136177960289e-05,
+      "loss": 3.5297,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.060297583209237e-05,
+      "loss": 3.5279,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0594622642189305e-05,
+      "loss": 3.5314,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0586236694678785e-05,
+      "loss": 3.5272,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.057785074716827e-05,
+      "loss": 3.5294,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056946479965775e-05,
+      "loss": 3.5363,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056107885214723e-05,
+      "loss": 3.5296,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055269290463671e-05,
+      "loss": 3.5127,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.054430695712619e-05,
+      "loss": 3.5272,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.053592100961567e-05,
+      "loss": 3.5292,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.052755144090888e-05,
+      "loss": 3.5245,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051916549339836e-05,
+      "loss": 3.5259,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051077954588784e-05,
+      "loss": 3.5209,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.050239359837732e-05,
+      "loss": 3.5274,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0494024029670533e-05,
+      "loss": 3.5263,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0485638082160013e-05,
+      "loss": 3.5238,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0477252134649493e-05,
+      "loss": 3.5291,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0468866187138973e-05,
+      "loss": 3.5286,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0460496618432182e-05,
+      "loss": 3.5405,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0452110670921662e-05,
+      "loss": 3.5101,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0443724723411142e-05,
+      "loss": 3.5171,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0435338775900625e-05,
+      "loss": 3.5232,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0426969207193834e-05,
+      "loss": 3.5254,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0418583259683314e-05,
+      "loss": 3.5309,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0410197312172794e-05,
+      "loss": 3.5276,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0401811364662274e-05,
+      "loss": 3.5158,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0393441795955487e-05,
+      "loss": 3.5294,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0385055848444966e-05,
+      "loss": 3.5267,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0376669900934446e-05,
+      "loss": 3.5205,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0368283953423926e-05,
+      "loss": 3.5156,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0359914384717136e-05,
+      "loss": 3.5255,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0351528437206615e-05,
+      "loss": 3.529,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0343142489696095e-05,
+      "loss": 3.5097,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.033475654218558e-05,
+      "loss": 3.5201,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.032637059467506e-05,
+      "loss": 3.5124,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0318001025968268e-05,
+      "loss": 3.5286,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0309615078457748e-05,
+      "loss": 3.5205,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0301229130947228e-05,
+      "loss": 3.52,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0292843183436708e-05,
+      "loss": 3.5262,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.028447361472992e-05,
+      "loss": 3.529,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.02760876672194e-05,
+      "loss": 3.5212,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.026770171970888e-05,
+      "loss": 3.5271,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.025931577219836e-05,
+      "loss": 3.5271,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.025094620349157e-05,
+      "loss": 3.5175,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.024256025598105e-05,
+      "loss": 3.5114,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0234174308470532e-05,
+      "loss": 3.5245,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0225788360960012e-05,
+      "loss": 3.5216,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.021741879225322e-05,
+      "loss": 3.5308,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.02090328447427e-05,
+      "loss": 3.5175,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.020064689723218e-05,
+      "loss": 3.5219,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.019226094972166e-05,
+      "loss": 3.5196,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0183891381014874e-05,
+      "loss": 3.5142,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0175505433504354e-05,
+      "loss": 3.5136,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0167119485993834e-05,
+      "loss": 3.5308,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0158733538483314e-05,
+      "loss": 3.5344,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0150363969776523e-05,
+      "loss": 3.5201,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0141978022266003e-05,
+      "loss": 3.5193,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0133592074755486e-05,
+      "loss": 3.5275,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0125206127244966e-05,
+      "loss": 3.5216,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0116836558538175e-05,
+      "loss": 3.5156,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0108450611027655e-05,
+      "loss": 3.524,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0100064663517135e-05,
+      "loss": 3.5351,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.009167871600661e-05,
+      "loss": 3.5262,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0083309147299827e-05,
+      "loss": 3.5224,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0074923199789307e-05,
+      "loss": 3.5166,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0066537252278787e-05,
+      "loss": 3.526,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.005815130476826e-05,
+      "loss": 3.5238,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0049781736061476e-05,
+      "loss": 3.5117,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0041395788550956e-05,
+      "loss": 3.5405,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.003300984104044e-05,
+      "loss": 3.5195,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0024623893529913e-05,
+      "loss": 3.5279,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.001625432482313e-05,
+      "loss": 3.5245,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.000786837731261e-05,
+      "loss": 3.5217,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.90067195892334,
+      "eval_runtime": 304.9615,
+      "eval_samples_per_second": 1251.276,
+      "eval_steps_per_second": 39.103,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9999482429802085e-05,
+      "loss": 3.5182,
+      "step": 1221632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9991096482291565e-05,
+      "loss": 3.518,
+      "step": 1222144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.998272691358478e-05,
+      "loss": 3.5322,
+      "step": 1222656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.997434096607426e-05,
+      "loss": 3.5272,
+      "step": 1223168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9965955018563734e-05,
+      "loss": 3.5265,
+      "step": 1223680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9957569071053214e-05,
+      "loss": 3.5133,
+      "step": 1224192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.994919950234643e-05,
+      "loss": 3.521,
+      "step": 1224704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.994081355483591e-05,
+      "loss": 3.5167,
+      "step": 1225216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9932427607325387e-05,
+      "loss": 3.5289,
+      "step": 1225728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9924041659814867e-05,
+      "loss": 3.5118,
+      "step": 1226240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9915672091108082e-05,
+      "loss": 3.5278,
+      "step": 1226752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9907286143597556e-05,
+      "loss": 3.5265,
+      "step": 1227264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.989890019608704e-05,
+      "loss": 3.5107,
+      "step": 1227776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.989051424857652e-05,
+      "loss": 3.5142,
+      "step": 1228288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9882144679869735e-05,
+      "loss": 3.506,
+      "step": 1228800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9873758732359208e-05,
+      "loss": 3.5102,
+      "step": 1229312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9865372784848688e-05,
+      "loss": 3.5185,
+      "step": 1229824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9856986837338168e-05,
+      "loss": 3.5107,
+      "step": 1230336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9848617268631384e-05,
+      "loss": 3.5208,
+      "step": 1230848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.984023132112086e-05,
+      "loss": 3.5364,
+      "step": 1231360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.983184537361034e-05,
+      "loss": 3.5164,
+      "step": 1231872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.982345942609982e-05,
+      "loss": 3.5244,
+      "step": 1232384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.98150734785893e-05,
+      "loss": 3.5178,
+      "step": 1232896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.980670390988251e-05,
+      "loss": 3.5165,
+      "step": 1233408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9798317962371993e-05,
+      "loss": 3.5211,
+      "step": 1233920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9789932014861473e-05,
+      "loss": 3.509,
+      "step": 1234432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9781546067350953e-05,
+      "loss": 3.5174,
+      "step": 1234944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.977317649864416e-05,
+      "loss": 3.5147,
+      "step": 1235456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.976479055113364e-05,
+      "loss": 3.5025,
+      "step": 1235968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.975640460362312e-05,
+      "loss": 3.5196,
+      "step": 1236480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.97480186561126e-05,
+      "loss": 3.5151,
+      "step": 1236992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9739649087405814e-05,
+      "loss": 3.5248,
+      "step": 1237504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9731263139895294e-05,
+      "loss": 3.5176,
+      "step": 1238016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9722877192384774e-05,
+      "loss": 3.509,
+      "step": 1238528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9714491244874254e-05,
+      "loss": 3.5157,
+      "step": 1239040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9706121676167463e-05,
+      "loss": 3.5332,
+      "step": 1239552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9697735728656946e-05,
+      "loss": 3.5092,
+      "step": 1240064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9689349781146426e-05,
+      "loss": 3.513,
+      "step": 1240576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9680963833635906e-05,
+      "loss": 3.5121,
+      "step": 1241088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9672594264929115e-05,
+      "loss": 3.5061,
+      "step": 1241600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9664208317418595e-05,
+      "loss": 3.5139,
+      "step": 1242112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9655822369908075e-05,
+      "loss": 3.529,
+      "step": 1242624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9647436422397555e-05,
+      "loss": 3.5228,
+      "step": 1243136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9639066853690768e-05,
+      "loss": 3.5107,
+      "step": 1243648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9630680906180248e-05,
+      "loss": 3.5221,
+      "step": 1244160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9622294958669728e-05,
+      "loss": 3.513,
+      "step": 1244672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9613909011159208e-05,
+      "loss": 3.5194,
+      "step": 1245184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9605539442452417e-05,
+      "loss": 3.4961,
+      "step": 1245696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9597153494941897e-05,
+      "loss": 3.5149,
+      "step": 1246208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.958876754743138e-05,
+      "loss": 3.5206,
+      "step": 1246720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.958038159992086e-05,
+      "loss": 3.5171,
+      "step": 1247232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.957201203121407e-05,
+      "loss": 3.5174,
+      "step": 1247744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.956362608370355e-05,
+      "loss": 3.5032,
+      "step": 1248256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.955524013619303e-05,
+      "loss": 3.5188,
+      "step": 1248768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.954685418868251e-05,
+      "loss": 3.5024,
+      "step": 1249280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.953848461997572e-05,
+      "loss": 3.5287,
+      "step": 1249792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.95300986724652e-05,
+      "loss": 3.5119,
+      "step": 1250304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.952171272495468e-05,
+      "loss": 3.5301,
+      "step": 1250816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.951332677744416e-05,
+      "loss": 3.5113,
+      "step": 1251328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.950495720873737e-05,
+      "loss": 3.5072,
+      "step": 1251840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.949657126122685e-05,
+      "loss": 3.5172,
+      "step": 1252352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9488185313716334e-05,
+      "loss": 3.5107,
+      "step": 1252864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9479799366205814e-05,
+      "loss": 3.5103,
+      "step": 1253376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9471429797499023e-05,
+      "loss": 3.5051,
+      "step": 1253888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9463043849988503e-05,
+      "loss": 3.519,
+      "step": 1254400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9454657902477983e-05,
+      "loss": 3.5033,
+      "step": 1254912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9446271954967463e-05,
+      "loss": 3.5082,
+      "step": 1255424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9437902386260675e-05,
+      "loss": 3.501,
+      "step": 1255936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9429516438750155e-05,
+      "loss": 3.5048,
+      "step": 1256448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9421130491239635e-05,
+      "loss": 3.5234,
+      "step": 1256960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9412744543729115e-05,
+      "loss": 3.5189,
+      "step": 1257472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9404374975022324e-05,
+      "loss": 3.5084,
+      "step": 1257984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9395989027511804e-05,
+      "loss": 3.5079,
+      "step": 1258496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9387603080001287e-05,
+      "loss": 3.5167,
+      "step": 1259008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9379217132490767e-05,
+      "loss": 3.515,
+      "step": 1259520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9370847563783976e-05,
+      "loss": 3.5131,
+      "step": 1260032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9362461616273456e-05,
+      "loss": 3.5185,
+      "step": 1260544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9354075668762936e-05,
+      "loss": 3.5148,
+      "step": 1261056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.934570610005615e-05,
+      "loss": 3.5185,
+      "step": 1261568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.933732015254563e-05,
+      "loss": 3.5137,
+      "step": 1262080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.932893420503511e-05,
+      "loss": 3.5187,
+      "step": 1262592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.932054825752459e-05,
+      "loss": 3.5214,
+      "step": 1263104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.931216231001407e-05,
+      "loss": 3.5137,
+      "step": 1263616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.930377636250355e-05,
+      "loss": 3.4974,
+      "step": 1264128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.929539041499303e-05,
+      "loss": 3.5112,
+      "step": 1264640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.928700446748251e-05,
+      "loss": 3.5139,
+      "step": 1265152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.927863489877572e-05,
+      "loss": 3.5105,
+      "step": 1265664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.92702489512652e-05,
+      "loss": 3.5106,
+      "step": 1266176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.926186300375468e-05,
+      "loss": 3.5063,
+      "step": 1266688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.925347705624416e-05,
+      "loss": 3.5155,
+      "step": 1267200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.924510748753737e-05,
+      "loss": 3.5084,
+      "step": 1267712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.923672154002685e-05,
+      "loss": 3.5096,
+      "step": 1268224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9228335592516333e-05,
+      "loss": 3.5108,
+      "step": 1268736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9219949645005813e-05,
+      "loss": 3.5117,
+      "step": 1269248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9211580076299022e-05,
+      "loss": 3.5268,
+      "step": 1269760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9203194128788502e-05,
+      "loss": 3.4981,
+      "step": 1270272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9194808181277982e-05,
+      "loss": 3.5004,
+      "step": 1270784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9186422233767462e-05,
+      "loss": 3.5123,
+      "step": 1271296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9178052665060675e-05,
+      "loss": 3.5141,
+      "step": 1271808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9169666717550155e-05,
+      "loss": 3.5171,
+      "step": 1272320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9161280770039634e-05,
+      "loss": 3.5119,
+      "step": 1272832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9152894822529108e-05,
+      "loss": 3.5022,
+      "step": 1273344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9144525253822324e-05,
+      "loss": 3.5116,
+      "step": 1273856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9136139306311804e-05,
+      "loss": 3.5168,
+      "step": 1274368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9127753358801287e-05,
+      "loss": 3.5033,
+      "step": 1274880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.911936741129076e-05,
+      "loss": 3.4993,
+      "step": 1275392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9110997842583976e-05,
+      "loss": 3.5145,
+      "step": 1275904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9102611895073456e-05,
+      "loss": 3.5126,
+      "step": 1276416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9094225947562936e-05,
+      "loss": 3.4971,
+      "step": 1276928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9085840000052412e-05,
+      "loss": 3.5019,
+      "step": 1277440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9077470431345628e-05,
+      "loss": 3.5009,
+      "step": 1277952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9069084483835108e-05,
+      "loss": 3.5132,
+      "step": 1278464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.906069853632458e-05,
+      "loss": 3.5063,
+      "step": 1278976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.905231258881406e-05,
+      "loss": 3.5079,
+      "step": 1279488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9043943020107277e-05,
+      "loss": 3.5114,
+      "step": 1280000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9035557072596757e-05,
+      "loss": 3.5175,
+      "step": 1280512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9027171125086234e-05,
+      "loss": 3.5007,
+      "step": 1281024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9018785177575714e-05,
+      "loss": 3.5146,
+      "step": 1281536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.901041560886893e-05,
+      "loss": 3.5127,
+      "step": 1282048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.900202966135841e-05,
+      "loss": 3.5072,
+      "step": 1282560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8993643713847886e-05,
+      "loss": 3.4912,
+      "step": 1283072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8985257766337366e-05,
+      "loss": 3.5104,
+      "step": 1283584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8976888197630582e-05,
+      "loss": 3.5094,
+      "step": 1284096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8968502250120055e-05,
+      "loss": 3.5149,
+      "step": 1284608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8960116302609535e-05,
+      "loss": 3.5008,
+      "step": 1285120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8951730355099015e-05,
+      "loss": 3.5121,
+      "step": 1285632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8943344407588495e-05,
+      "loss": 3.5047,
+      "step": 1286144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8934991217685443e-05,
+      "loss": 3.5,
+      "step": 1286656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8926605270174923e-05,
+      "loss": 3.494,
+      "step": 1287168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8918219322664403e-05,
+      "loss": 3.5205,
+      "step": 1287680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8909833375153883e-05,
+      "loss": 3.5212,
+      "step": 1288192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8901447427643356e-05,
+      "loss": 3.5057,
+      "step": 1288704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.889306148013284e-05,
+      "loss": 3.4995,
+      "step": 1289216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.888467553262232e-05,
+      "loss": 3.519,
+      "step": 1289728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.88762895851118e-05,
+      "loss": 3.5061,
+      "step": 1290240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.886792001640501e-05,
+      "loss": 3.5048,
+      "step": 1290752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.885953406889449e-05,
+      "loss": 3.5073,
+      "step": 1291264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.885114812138397e-05,
+      "loss": 3.518,
+      "step": 1291776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.884276217387345e-05,
+      "loss": 3.5093,
+      "step": 1292288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.883439260516666e-05,
+      "loss": 3.5113,
+      "step": 1292800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.882600665765614e-05,
+      "loss": 3.4993,
+      "step": 1293312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.881762071014562e-05,
+      "loss": 3.5128,
+      "step": 1293824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.88092347626351e-05,
+      "loss": 3.5091,
+      "step": 1294336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.880086519392831e-05,
+      "loss": 3.5009,
+      "step": 1294848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.879247924641779e-05,
+      "loss": 3.526,
+      "step": 1295360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8784093298907273e-05,
+      "loss": 3.5045,
+      "step": 1295872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8775707351396753e-05,
+      "loss": 3.513,
+      "step": 1296384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8767337782689962e-05,
+      "loss": 3.509,
+      "step": 1296896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8758951835179442e-05,
+      "loss": 3.5078,
+      "step": 1297408
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.902683734893799,
+      "eval_runtime": 304.3447,
+      "eval_samples_per_second": 1253.812,
+      "eval_steps_per_second": 39.183,
+      "step": 1297440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8750565887668922e-05,
+      "loss": 3.4981,
+      "step": 1297920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8742179940158402e-05,
+      "loss": 3.5072,
+      "step": 1298432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8733793992647882e-05,
+      "loss": 3.5169,
+      "step": 1298944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8725424423941095e-05,
+      "loss": 3.5131,
+      "step": 1299456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8717038476430575e-05,
+      "loss": 3.5168,
+      "step": 1299968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8708652528920055e-05,
+      "loss": 3.4989,
+      "step": 1300480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8700266581409535e-05,
+      "loss": 3.5041,
+      "step": 1300992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8691897012702744e-05,
+      "loss": 3.503,
+      "step": 1301504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8683511065192227e-05,
+      "loss": 3.5161,
+      "step": 1302016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8675125117681707e-05,
+      "loss": 3.5021,
+      "step": 1302528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8666739170171187e-05,
+      "loss": 3.51,
+      "step": 1303040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8658369601464396e-05,
+      "loss": 3.5147,
+      "step": 1303552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8649983653953876e-05,
+      "loss": 3.4953,
+      "step": 1304064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8641597706443356e-05,
+      "loss": 3.5006,
+      "step": 1304576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8633211758932836e-05,
+      "loss": 3.4918,
+      "step": 1305088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.862484219022605e-05,
+      "loss": 3.4994,
+      "step": 1305600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.861645624271553e-05,
+      "loss": 3.4961,
+      "step": 1306112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.860807029520501e-05,
+      "loss": 3.4984,
+      "step": 1306624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8599684347694488e-05,
+      "loss": 3.5029,
+      "step": 1307136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8591314778987697e-05,
+      "loss": 3.5248,
+      "step": 1307648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.858292883147718e-05,
+      "loss": 3.5022,
+      "step": 1308160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.857454288396666e-05,
+      "loss": 3.5053,
+      "step": 1308672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.856615693645614e-05,
+      "loss": 3.51,
+      "step": 1309184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.855778736774935e-05,
+      "loss": 3.5052,
+      "step": 1309696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.854940142023883e-05,
+      "loss": 3.5006,
+      "step": 1310208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.854101547272831e-05,
+      "loss": 3.4976,
+      "step": 1310720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.853262952521779e-05,
+      "loss": 3.5053,
+      "step": 1311232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8524259956511002e-05,
+      "loss": 3.5002,
+      "step": 1311744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8515874009000482e-05,
+      "loss": 3.4862,
+      "step": 1312256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8507488061489962e-05,
+      "loss": 3.5061,
+      "step": 1312768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8499102113979442e-05,
+      "loss": 3.4965,
+      "step": 1313280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.849073254527265e-05,
+      "loss": 3.5119,
+      "step": 1313792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8482346597762134e-05,
+      "loss": 3.5065,
+      "step": 1314304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8473960650251614e-05,
+      "loss": 3.4907,
+      "step": 1314816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8465574702741094e-05,
+      "loss": 3.5069,
+      "step": 1315328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8457205134034303e-05,
+      "loss": 3.5146,
+      "step": 1315840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8448819186523783e-05,
+      "loss": 3.4917,
+      "step": 1316352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8440433239013263e-05,
+      "loss": 3.504,
+      "step": 1316864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8432047291502743e-05,
+      "loss": 3.4962,
+      "step": 1317376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8423677722795956e-05,
+      "loss": 3.4946,
+      "step": 1317888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8415291775285436e-05,
+      "loss": 3.5006,
+      "step": 1318400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8406905827774916e-05,
+      "loss": 3.514,
+      "step": 1318912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8398519880264396e-05,
+      "loss": 3.5076,
+      "step": 1319424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8390150311557605e-05,
+      "loss": 3.4971,
+      "step": 1319936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8381764364047088e-05,
+      "loss": 3.5086,
+      "step": 1320448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8373378416536568e-05,
+      "loss": 3.498,
+      "step": 1320960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8364992469026048e-05,
+      "loss": 3.5065,
+      "step": 1321472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8356622900319257e-05,
+      "loss": 3.4857,
+      "step": 1321984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8348236952808737e-05,
+      "loss": 3.4965,
+      "step": 1322496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8339851005298217e-05,
+      "loss": 3.5084,
+      "step": 1323008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8331465057787697e-05,
+      "loss": 3.5072,
+      "step": 1323520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.832309548908091e-05,
+      "loss": 3.4987,
+      "step": 1324032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.831470954157039e-05,
+      "loss": 3.4891,
+      "step": 1324544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.830632359405987e-05,
+      "loss": 3.5054,
+      "step": 1325056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.829793764654935e-05,
+      "loss": 3.488,
+      "step": 1325568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.828956807784256e-05,
+      "loss": 3.5142,
+      "step": 1326080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8281182130332042e-05,
+      "loss": 3.5026,
+      "step": 1326592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.827279618282152e-05,
+      "loss": 3.5166,
+      "step": 1327104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8264410235311e-05,
+      "loss": 3.4993,
+      "step": 1327616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.825604066660421e-05,
+      "loss": 3.4961,
+      "step": 1328128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.824765471909369e-05,
+      "loss": 3.4983,
+      "step": 1328640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.823926877158317e-05,
+      "loss": 3.5037,
+      "step": 1329152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.823088282407265e-05,
+      "loss": 3.4917,
+      "step": 1329664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8222513255365863e-05,
+      "loss": 3.4914,
+      "step": 1330176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8214127307855343e-05,
+      "loss": 3.5095,
+      "step": 1330688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8205741360344823e-05,
+      "loss": 3.4917,
+      "step": 1331200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8197355412834303e-05,
+      "loss": 3.4893,
+      "step": 1331712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8188985844127512e-05,
+      "loss": 3.4914,
+      "step": 1332224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8180599896616995e-05,
+      "loss": 3.4931,
+      "step": 1332736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8172213949106475e-05,
+      "loss": 3.5084,
+      "step": 1333248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8163828001595955e-05,
+      "loss": 3.5062,
+      "step": 1333760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8155458432889164e-05,
+      "loss": 3.4941,
+      "step": 1334272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8147072485378644e-05,
+      "loss": 3.4971,
+      "step": 1334784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8138686537868124e-05,
+      "loss": 3.5016,
+      "step": 1335296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8130300590357604e-05,
+      "loss": 3.4986,
+      "step": 1335808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8121931021650817e-05,
+      "loss": 3.5001,
+      "step": 1336320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8113545074140297e-05,
+      "loss": 3.5061,
+      "step": 1336832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8105159126629777e-05,
+      "loss": 3.5002,
+      "step": 1337344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8096789557922986e-05,
+      "loss": 3.5083,
+      "step": 1337856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8088403610412466e-05,
+      "loss": 3.4991,
+      "step": 1338368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8080017662901946e-05,
+      "loss": 3.5025,
+      "step": 1338880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.807163171539143e-05,
+      "loss": 3.503,
+      "step": 1339392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8063245767880902e-05,
+      "loss": 3.5027,
+      "step": 1339904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8054859820370382e-05,
+      "loss": 3.4875,
+      "step": 1340416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8046473872859862e-05,
+      "loss": 3.497,
+      "step": 1340928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8038087925349342e-05,
+      "loss": 3.4997,
+      "step": 1341440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8029718356642555e-05,
+      "loss": 3.4981,
+      "step": 1341952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8021332409132034e-05,
+      "loss": 3.498,
+      "step": 1342464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8012946461621514e-05,
+      "loss": 3.4909,
+      "step": 1342976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8004560514110994e-05,
+      "loss": 3.499,
+      "step": 1343488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7996190945404203e-05,
+      "loss": 3.4952,
+      "step": 1344000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7987804997893687e-05,
+      "loss": 3.4994,
+      "step": 1344512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7979419050383167e-05,
+      "loss": 3.4908,
+      "step": 1345024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7971033102872647e-05,
+      "loss": 3.4992,
+      "step": 1345536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7962663534165856e-05,
+      "loss": 3.5083,
+      "step": 1346048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7954277586655336e-05,
+      "loss": 3.4985,
+      "step": 1346560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7945891639144816e-05,
+      "loss": 3.4798,
+      "step": 1347072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7937505691634296e-05,
+      "loss": 3.4967,
+      "step": 1347584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7929136122927508e-05,
+      "loss": 3.5009,
+      "step": 1348096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7920750175416988e-05,
+      "loss": 3.5013,
+      "step": 1348608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7912364227906468e-05,
+      "loss": 3.4995,
+      "step": 1349120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7903978280395948e-05,
+      "loss": 3.491,
+      "step": 1349632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7895608711689157e-05,
+      "loss": 3.4935,
+      "step": 1350144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7887222764178637e-05,
+      "loss": 3.504,
+      "step": 1350656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.787883681666812e-05,
+      "loss": 3.4861,
+      "step": 1351168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.78704508691576e-05,
+      "loss": 3.4907,
+      "step": 1351680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.786208130045081e-05,
+      "loss": 3.4994,
+      "step": 1352192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.785369535294029e-05,
+      "loss": 3.5019,
+      "step": 1352704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.784530940542977e-05,
+      "loss": 3.4785,
+      "step": 1353216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.783692345791925e-05,
+      "loss": 3.4944,
+      "step": 1353728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7828553889212462e-05,
+      "loss": 3.4849,
+      "step": 1354240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7820167941701942e-05,
+      "loss": 3.5019,
+      "step": 1354752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7811781994191422e-05,
+      "loss": 3.4935,
+      "step": 1355264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7803396046680902e-05,
+      "loss": 3.4864,
+      "step": 1355776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.779502647797411e-05,
+      "loss": 3.5056,
+      "step": 1356288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.778664053046359e-05,
+      "loss": 3.5016,
+      "step": 1356800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7778270961756803e-05,
+      "loss": 3.4887,
+      "step": 1357312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7769885014246283e-05,
+      "loss": 3.5008,
+      "step": 1357824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7761499066735763e-05,
+      "loss": 3.4962,
+      "step": 1358336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7753113119225243e-05,
+      "loss": 3.4954,
+      "step": 1358848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7744727171714723e-05,
+      "loss": 3.4777,
+      "step": 1359360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7736341224204203e-05,
+      "loss": 3.5019,
+      "step": 1359872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7727955276693683e-05,
+      "loss": 3.4921,
+      "step": 1360384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7719569329183166e-05,
+      "loss": 3.5006,
+      "step": 1360896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7711199760476375e-05,
+      "loss": 3.49,
+      "step": 1361408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7702813812965855e-05,
+      "loss": 3.4962,
+      "step": 1361920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7694427865455335e-05,
+      "loss": 3.4947,
+      "step": 1362432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7686041917944815e-05,
+      "loss": 3.4871,
+      "step": 1362944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7677672349238028e-05,
+      "loss": 3.4789,
+      "step": 1363456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7669286401727508e-05,
+      "loss": 3.5098,
+      "step": 1363968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7660900454216988e-05,
+      "loss": 3.5015,
+      "step": 1364480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7652514506706468e-05,
+      "loss": 3.493,
+      "step": 1364992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7644128559195948e-05,
+      "loss": 3.4892,
+      "step": 1365504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7635742611685428e-05,
+      "loss": 3.5024,
+      "step": 1366016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7627356664174908e-05,
+      "loss": 3.496,
+      "step": 1366528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.761898709546812e-05,
+      "loss": 3.4891,
+      "step": 1367040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.76106011479576e-05,
+      "loss": 3.4888,
+      "step": 1367552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.760221520044708e-05,
+      "loss": 3.5035,
+      "step": 1368064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.759382925293656e-05,
+      "loss": 3.4962,
+      "step": 1368576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.758545968422977e-05,
+      "loss": 3.503,
+      "step": 1369088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.757707373671925e-05,
+      "loss": 3.4855,
+      "step": 1369600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.756868778920873e-05,
+      "loss": 3.4998,
+      "step": 1370112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7560301841698212e-05,
+      "loss": 3.4958,
+      "step": 1370624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.755193227299142e-05,
+      "loss": 3.4895,
+      "step": 1371136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.75435463254809e-05,
+      "loss": 3.5098,
+      "step": 1371648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.753516037797038e-05,
+      "loss": 3.4944,
+      "step": 1372160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.752677443045986e-05,
+      "loss": 3.498,
+      "step": 1372672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7518404861753074e-05,
+      "loss": 3.4914,
+      "step": 1373184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7510018914242554e-05,
+      "loss": 3.4974,
+      "step": 1373696
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9054596424102783,
+      "eval_runtime": 306.8879,
+      "eval_samples_per_second": 1243.422,
+      "eval_steps_per_second": 38.858,
+      "step": 1373760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7501632966732034e-05,
+      "loss": 3.4837,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7493247019221507e-05,
+      "loss": 3.4975,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7484861071710987e-05,
+      "loss": 3.5012,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7476475124200467e-05,
+      "loss": 3.4992,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7468105555493683e-05,
+      "loss": 3.5059,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.745971960798316e-05,
+      "loss": 3.4818,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.745133366047264e-05,
+      "loss": 3.49,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.744294771296212e-05,
+      "loss": 3.4884,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7434578144255335e-05,
+      "loss": 3.5023,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.742619219674481e-05,
+      "loss": 3.4901,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.741780624923429e-05,
+      "loss": 3.496,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7409436680527507e-05,
+      "loss": 3.4994,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.740105073301698e-05,
+      "loss": 3.4835,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.739266478550646e-05,
+      "loss": 3.4897,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.738427883799594e-05,
+      "loss": 3.4768,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7375909269289156e-05,
+      "loss": 3.4876,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7367523321778633e-05,
+      "loss": 3.4861,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7359137374268113e-05,
+      "loss": 3.4852,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7350751426757593e-05,
+      "loss": 3.4878,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.734238185805081e-05,
+      "loss": 3.5107,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7333995910540282e-05,
+      "loss": 3.4921,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7325609963029765e-05,
+      "loss": 3.4919,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7317224015519245e-05,
+      "loss": 3.4958,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7308854446812454e-05,
+      "loss": 3.4935,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7300468499301934e-05,
+      "loss": 3.4855,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7292082551791414e-05,
+      "loss": 3.4816,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7283696604280894e-05,
+      "loss": 3.4936,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7275327035574107e-05,
+      "loss": 3.4827,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7266941088063587e-05,
+      "loss": 3.477,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7258555140553066e-05,
+      "loss": 3.4928,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7250169193042546e-05,
+      "loss": 3.4856,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7241799624335756e-05,
+      "loss": 3.4951,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7233413676825235e-05,
+      "loss": 3.4911,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.722502772931472e-05,
+      "loss": 3.4792,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.72166417818042e-05,
+      "loss": 3.4923,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7208272213097408e-05,
+      "loss": 3.506,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7199886265586888e-05,
+      "loss": 3.4769,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7191500318076368e-05,
+      "loss": 3.4933,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7183114370565848e-05,
+      "loss": 3.4786,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.717474480185906e-05,
+      "loss": 3.4852,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7166375233152276e-05,
+      "loss": 3.4846,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7157989285641756e-05,
+      "loss": 3.5011,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.714960333813123e-05,
+      "loss": 3.4945,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.714121739062071e-05,
+      "loss": 3.4851,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.713283144311019e-05,
+      "loss": 3.4938,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.712444549559967e-05,
+      "loss": 3.4826,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7116059548089152e-05,
+      "loss": 3.495,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7107673600578632e-05,
+      "loss": 3.4766,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.709930403187184e-05,
+      "loss": 3.4805,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.709091808436132e-05,
+      "loss": 3.4942,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.70825321368508e-05,
+      "loss": 3.493,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.707414618934028e-05,
+      "loss": 3.4858,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7065776620633494e-05,
+      "loss": 3.4811,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7057390673122974e-05,
+      "loss": 3.4922,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7049004725612454e-05,
+      "loss": 3.4739,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7040618778101934e-05,
+      "loss": 3.4953,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7032249209395143e-05,
+      "loss": 3.4911,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7023863261884623e-05,
+      "loss": 3.5008,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7015477314374106e-05,
+      "loss": 3.4914,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7007091366863586e-05,
+      "loss": 3.4814,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6998721798156795e-05,
+      "loss": 3.4849,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6990335850646275e-05,
+      "loss": 3.4889,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6981949903135755e-05,
+      "loss": 3.477,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6973563955625235e-05,
+      "loss": 3.4801,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6965194386918448e-05,
+      "loss": 3.4914,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6956808439407927e-05,
+      "loss": 3.4833,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6948422491897407e-05,
+      "loss": 3.4752,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6940036544386887e-05,
+      "loss": 3.4783,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6931666975680096e-05,
+      "loss": 3.4763,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6923281028169576e-05,
+      "loss": 3.494,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.691489508065906e-05,
+      "loss": 3.491,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.690650913314854e-05,
+      "loss": 3.4821,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.689813956444175e-05,
+      "loss": 3.4863,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6889769995734958e-05,
+      "loss": 3.4842,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6881384048224438e-05,
+      "loss": 3.4899,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.687299810071392e-05,
+      "loss": 3.4846,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.68646121532034e-05,
+      "loss": 3.4938,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.685622620569288e-05,
+      "loss": 3.4911,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.684785663698609e-05,
+      "loss": 3.4891,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.683947068947557e-05,
+      "loss": 3.4866,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.683108474196505e-05,
+      "loss": 3.49,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.682269879445453e-05,
+      "loss": 3.4912,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6814312846944013e-05,
+      "loss": 3.4889,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6805926899433493e-05,
+      "loss": 3.4778,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6797540951922973e-05,
+      "loss": 3.4828,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6789155004412453e-05,
+      "loss": 3.4868,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6780785435705662e-05,
+      "loss": 3.4826,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6772399488195142e-05,
+      "loss": 3.4876,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6764013540684622e-05,
+      "loss": 3.4749,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6755627593174106e-05,
+      "loss": 3.4844,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6747258024467315e-05,
+      "loss": 3.4845,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6738872076956795e-05,
+      "loss": 3.4884,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6730502508250004e-05,
+      "loss": 3.4736,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6722116560739484e-05,
+      "loss": 3.4881,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6713730613228967e-05,
+      "loss": 3.4988,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6705344665718447e-05,
+      "loss": 3.4845,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6696975097011656e-05,
+      "loss": 3.4682,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6688589149501136e-05,
+      "loss": 3.4833,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6680203201990616e-05,
+      "loss": 3.4869,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6671817254480096e-05,
+      "loss": 3.4882,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6663431306969576e-05,
+      "loss": 3.4858,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.665504535945906e-05,
+      "loss": 3.4756,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6646659411948532e-05,
+      "loss": 3.4818,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6638273464438012e-05,
+      "loss": 3.4912,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.662990389573123e-05,
+      "loss": 3.4776,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6621517948220708e-05,
+      "loss": 3.4797,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6613132000710185e-05,
+      "loss": 3.4854,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.66047624320034e-05,
+      "loss": 3.4881,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.659637648449288e-05,
+      "loss": 3.4654,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.658799053698236e-05,
+      "loss": 3.482,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6579604589471834e-05,
+      "loss": 3.4716,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.657123502076505e-05,
+      "loss": 3.4857,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.656284907325453e-05,
+      "loss": 3.4829,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6554463125744006e-05,
+      "loss": 3.4773,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6546077178233486e-05,
+      "loss": 3.4911,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6537707609526702e-05,
+      "loss": 3.4913,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.652933804081991e-05,
+      "loss": 3.4699,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.652095209330939e-05,
+      "loss": 3.4869,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6512566145798874e-05,
+      "loss": 3.4849,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6504180198288354e-05,
+      "loss": 3.482,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6495794250777834e-05,
+      "loss": 3.4665,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6487424682071043e-05,
+      "loss": 3.4877,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6479038734560523e-05,
+      "loss": 3.481,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6470652787050003e-05,
+      "loss": 3.485,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.646226683953948e-05,
+      "loss": 3.4799,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.645388089202896e-05,
+      "loss": 3.4861,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.644549494451844e-05,
+      "loss": 3.4755,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.643710899700792e-05,
+      "loss": 3.4721,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.642873942830113e-05,
+      "loss": 3.4688,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6420353480790612e-05,
+      "loss": 3.4941,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6411967533280092e-05,
+      "loss": 3.487,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6403581585769572e-05,
+      "loss": 3.4835,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.639521201706278e-05,
+      "loss": 3.4776,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.638682606955226e-05,
+      "loss": 3.4899,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.637844012204174e-05,
+      "loss": 3.485,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.637005417453122e-05,
+      "loss": 3.4747,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.63616682270207e-05,
+      "loss": 3.4782,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6353282279510184e-05,
+      "loss": 3.4878,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6344896331999664e-05,
+      "loss": 3.4833,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6336526763292873e-05,
+      "loss": 3.4911,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6328140815782353e-05,
+      "loss": 3.4759,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6319754868271833e-05,
+      "loss": 3.4853,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6311368920761313e-05,
+      "loss": 3.4812,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6302982973250793e-05,
+      "loss": 3.4749,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6294597025740277e-05,
+      "loss": 3.4974,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6286211078229757e-05,
+      "loss": 3.4792,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6277825130719237e-05,
+      "loss": 3.4858,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6269455562012446e-05,
+      "loss": 3.4802,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6261069614501926e-05,
+      "loss": 3.4829,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9071192741394043,
+      "eval_runtime": 304.1252,
+      "eval_samples_per_second": 1254.717,
+      "eval_steps_per_second": 39.211,
+      "step": 1450080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6252683666991406e-05,
+      "loss": 3.479,
+      "step": 1450496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6244297719480885e-05,
+      "loss": 3.4807,
+      "step": 1451008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6235928150774098e-05,
+      "loss": 3.4885,
+      "step": 1451520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6227542203263578e-05,
+      "loss": 3.4867,
+      "step": 1452032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6219156255753058e-05,
+      "loss": 3.4945,
+      "step": 1452544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6210770308242538e-05,
+      "loss": 3.4735,
+      "step": 1453056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6202400739535747e-05,
+      "loss": 3.4773,
+      "step": 1453568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.619401479202523e-05,
+      "loss": 3.4697,
+      "step": 1454080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.618562884451471e-05,
+      "loss": 3.4943,
+      "step": 1454592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.617724289700419e-05,
+      "loss": 3.4744,
+      "step": 1455104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.61688733282974e-05,
+      "loss": 3.4843,
+      "step": 1455616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.616048738078688e-05,
+      "loss": 3.4857,
+      "step": 1456128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.615210143327636e-05,
+      "loss": 3.4763,
+      "step": 1456640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.614371548576584e-05,
+      "loss": 3.469,
+      "step": 1457152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.613534591705905e-05,
+      "loss": 3.4702,
+      "step": 1457664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.612695996954853e-05,
+      "loss": 3.4695,
+      "step": 1458176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.611857402203801e-05,
+      "loss": 3.4756,
+      "step": 1458688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.611018807452749e-05,
+      "loss": 3.4758,
+      "step": 1459200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.61018185058207e-05,
+      "loss": 3.4746,
+      "step": 1459712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6093432558310184e-05,
+      "loss": 3.4954,
+      "step": 1460224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6085046610799664e-05,
+      "loss": 3.4802,
+      "step": 1460736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6076660663289137e-05,
+      "loss": 3.4788,
+      "step": 1461248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6068291094582353e-05,
+      "loss": 3.4847,
+      "step": 1461760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6059905147071833e-05,
+      "loss": 3.4806,
+      "step": 1462272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6051519199561313e-05,
+      "loss": 3.4738,
+      "step": 1462784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6043149630854525e-05,
+      "loss": 3.4695,
+      "step": 1463296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6034763683344005e-05,
+      "loss": 3.4849,
+      "step": 1463808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6026377735833485e-05,
+      "loss": 3.4636,
+      "step": 1464320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6017991788322965e-05,
+      "loss": 3.4687,
+      "step": 1464832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6009622219616174e-05,
+      "loss": 3.4748,
+      "step": 1465344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6001236272105654e-05,
+      "loss": 3.4717,
+      "step": 1465856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5992850324595138e-05,
+      "loss": 3.4845,
+      "step": 1466368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.598446437708461e-05,
+      "loss": 3.4772,
+      "step": 1466880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5976094808377827e-05,
+      "loss": 3.4702,
+      "step": 1467392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5967708860867307e-05,
+      "loss": 3.4794,
+      "step": 1467904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5959322913356787e-05,
+      "loss": 3.4911,
+      "step": 1468416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5950936965846263e-05,
+      "loss": 3.4681,
+      "step": 1468928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.594256739713948e-05,
+      "loss": 3.48,
+      "step": 1469440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.593418144962896e-05,
+      "loss": 3.4626,
+      "step": 1469952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.592579550211844e-05,
+      "loss": 3.4821,
+      "step": 1470464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5917409554607912e-05,
+      "loss": 3.4659,
+      "step": 1470976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5909039985901128e-05,
+      "loss": 3.4905,
+      "step": 1471488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5900654038390608e-05,
+      "loss": 3.4778,
+      "step": 1472000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5892268090880085e-05,
+      "loss": 3.4749,
+      "step": 1472512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5883882143369564e-05,
+      "loss": 3.4819,
+      "step": 1473024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5875496195859044e-05,
+      "loss": 3.4702,
+      "step": 1473536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5867110248348524e-05,
+      "loss": 3.4791,
+      "step": 1474048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5858724300838004e-05,
+      "loss": 3.4656,
+      "step": 1474560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5850354732131217e-05,
+      "loss": 3.4665,
+      "step": 1475072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5841985163424433e-05,
+      "loss": 3.4822,
+      "step": 1475584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5833599215913913e-05,
+      "loss": 3.4754,
+      "step": 1476096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5825213268403386e-05,
+      "loss": 3.4786,
+      "step": 1476608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5816827320892866e-05,
+      "loss": 3.4662,
+      "step": 1477120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5808441373382346e-05,
+      "loss": 3.477,
+      "step": 1477632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.580005542587183e-05,
+      "loss": 3.4612,
+      "step": 1478144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.579166947836131e-05,
+      "loss": 3.4824,
+      "step": 1478656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.578328353085079e-05,
+      "loss": 3.4772,
+      "step": 1479168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5774913962143998e-05,
+      "loss": 3.489,
+      "step": 1479680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5766528014633478e-05,
+      "loss": 3.48,
+      "step": 1480192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5758142067122958e-05,
+      "loss": 3.4713,
+      "step": 1480704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5749756119612438e-05,
+      "loss": 3.4712,
+      "step": 1481216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5741370172101918e-05,
+      "loss": 3.4782,
+      "step": 1481728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.57329842245914e-05,
+      "loss": 3.4676,
+      "step": 1482240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.572459827708088e-05,
+      "loss": 3.4659,
+      "step": 1482752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.571621232957036e-05,
+      "loss": 3.4759,
+      "step": 1483264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.570784276086357e-05,
+      "loss": 3.4747,
+      "step": 1483776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.569945681335305e-05,
+      "loss": 3.4599,
+      "step": 1484288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.569107086584253e-05,
+      "loss": 3.4644,
+      "step": 1484800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.568268491833201e-05,
+      "loss": 3.463,
+      "step": 1485312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5674315349625223e-05,
+      "loss": 3.4852,
+      "step": 1485824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5665929402114703e-05,
+      "loss": 3.4779,
+      "step": 1486336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5657543454604183e-05,
+      "loss": 3.4687,
+      "step": 1486848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5649157507093662e-05,
+      "loss": 3.4764,
+      "step": 1487360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5640804317190604e-05,
+      "loss": 3.4694,
+      "step": 1487872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5632418369680084e-05,
+      "loss": 3.4821,
+      "step": 1488384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5624032422169564e-05,
+      "loss": 3.4662,
+      "step": 1488896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5615646474659044e-05,
+      "loss": 3.4868,
+      "step": 1489408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5607260527148524e-05,
+      "loss": 3.4712,
+      "step": 1489920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5598890958441733e-05,
+      "loss": 3.4776,
+      "step": 1490432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5590505010931216e-05,
+      "loss": 3.4777,
+      "step": 1490944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5582119063420696e-05,
+      "loss": 3.474,
+      "step": 1491456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5573733115910176e-05,
+      "loss": 3.4827,
+      "step": 1491968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5565347168399656e-05,
+      "loss": 3.4736,
+      "step": 1492480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5556961220889136e-05,
+      "loss": 3.4653,
+      "step": 1492992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5548575273378616e-05,
+      "loss": 3.4724,
+      "step": 1493504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5540189325868096e-05,
+      "loss": 3.468,
+      "step": 1494016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.553181975716131e-05,
+      "loss": 3.4778,
+      "step": 1494528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.552343380965079e-05,
+      "loss": 3.4729,
+      "step": 1495040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.551504786214027e-05,
+      "loss": 3.4621,
+      "step": 1495552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.550666191462974e-05,
+      "loss": 3.4757,
+      "step": 1496064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5498308724726687e-05,
+      "loss": 3.4682,
+      "step": 1496576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.548992277721617e-05,
+      "loss": 3.4753,
+      "step": 1497088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.548153682970565e-05,
+      "loss": 3.4626,
+      "step": 1497600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.547315088219513e-05,
+      "loss": 3.4778,
+      "step": 1498112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.546476493468461e-05,
+      "loss": 3.4837,
+      "step": 1498624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.545637898717409e-05,
+      "loss": 3.4699,
+      "step": 1499136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.54480094184673e-05,
+      "loss": 3.4576,
+      "step": 1499648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.543962347095678e-05,
+      "loss": 3.4721,
+      "step": 1500160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5431237523446262e-05,
+      "loss": 3.4676,
+      "step": 1500672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5422851575935742e-05,
+      "loss": 3.4799,
+      "step": 1501184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5414465628425215e-05,
+      "loss": 3.4733,
+      "step": 1501696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5406079680914695e-05,
+      "loss": 3.467,
+      "step": 1502208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5397693733404175e-05,
+      "loss": 3.4706,
+      "step": 1502720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.538932416469739e-05,
+      "loss": 3.479,
+      "step": 1503232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5380938217186868e-05,
+      "loss": 3.4611,
+      "step": 1503744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5372552269676348e-05,
+      "loss": 3.4683,
+      "step": 1504256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5364166322165828e-05,
+      "loss": 3.4707,
+      "step": 1504768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5355796753459044e-05,
+      "loss": 3.4772,
+      "step": 1505280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5347410805948517e-05,
+      "loss": 3.4513,
+      "step": 1505792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5339024858438e-05,
+      "loss": 3.4691,
+      "step": 1506304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.533063891092748e-05,
+      "loss": 3.4671,
+      "step": 1506816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.532226934222069e-05,
+      "loss": 3.4688,
+      "step": 1507328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.531388339471017e-05,
+      "loss": 3.4735,
+      "step": 1507840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.530549744719965e-05,
+      "loss": 3.4601,
+      "step": 1508352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.529711149968913e-05,
+      "loss": 3.48,
+      "step": 1508864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.528874193098234e-05,
+      "loss": 3.477,
+      "step": 1509376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5280372362275557e-05,
+      "loss": 3.4605,
+      "step": 1509888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5271986414765037e-05,
+      "loss": 3.4742,
+      "step": 1510400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5263600467254517e-05,
+      "loss": 3.4712,
+      "step": 1510912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.525521451974399e-05,
+      "loss": 3.4702,
+      "step": 1511424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.524682857223347e-05,
+      "loss": 3.4505,
+      "step": 1511936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5238459003526686e-05,
+      "loss": 3.4774,
+      "step": 1512448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5230073056016163e-05,
+      "loss": 3.4665,
+      "step": 1512960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5221687108505643e-05,
+      "loss": 3.4719,
+      "step": 1513472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5213301160995123e-05,
+      "loss": 3.47,
+      "step": 1513984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5204915213484603e-05,
+      "loss": 3.4752,
+      "step": 1514496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5196529265974083e-05,
+      "loss": 3.4672,
+      "step": 1515008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5188143318463563e-05,
+      "loss": 3.458,
+      "step": 1515520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5179773749756775e-05,
+      "loss": 3.4602,
+      "step": 1516032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5171387802246255e-05,
+      "loss": 3.4753,
+      "step": 1516544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5163001854735735e-05,
+      "loss": 3.475,
+      "step": 1517056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5154615907225215e-05,
+      "loss": 3.4717,
+      "step": 1517568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5146246338518424e-05,
+      "loss": 3.4677,
+      "step": 1518080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5137860391007907e-05,
+      "loss": 3.4759,
+      "step": 1518592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5129474443497387e-05,
+      "loss": 3.4708,
+      "step": 1519104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5121088495986867e-05,
+      "loss": 3.4622,
+      "step": 1519616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5112702548476347e-05,
+      "loss": 3.4674,
+      "step": 1520128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5104316600965827e-05,
+      "loss": 3.4741,
+      "step": 1520640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5095930653455307e-05,
+      "loss": 3.4737,
+      "step": 1521152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5087561084748516e-05,
+      "loss": 3.4784,
+      "step": 1521664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5079175137238e-05,
+      "loss": 3.4618,
+      "step": 1522176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.507078918972748e-05,
+      "loss": 3.4744,
+      "step": 1522688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.506240324221696e-05,
+      "loss": 3.4668,
+      "step": 1523200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.505401729470644e-05,
+      "loss": 3.467,
+      "step": 1523712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.504563134719592e-05,
+      "loss": 3.4796,
+      "step": 1524224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.50372453996854e-05,
+      "loss": 3.4706,
+      "step": 1524736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5028859452174876e-05,
+      "loss": 3.4735,
+      "step": 1525248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.502048988346809e-05,
+      "loss": 3.4644,
+      "step": 1525760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5012103935957572e-05,
+      "loss": 3.4733,
+      "step": 1526272
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9077465534210205,
+      "eval_runtime": 304.2404,
+      "eval_samples_per_second": 1254.242,
+      "eval_steps_per_second": 39.196,
+      "step": 1526400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5003717988447052e-05,
+      "loss": 3.4603,
+      "step": 1526784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4995332040936528e-05,
+      "loss": 3.4703,
+      "step": 1527296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4986962472229737e-05,
+      "loss": 3.4751,
+      "step": 1527808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4978576524719217e-05,
+      "loss": 3.4742,
+      "step": 1528320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.49701905772087e-05,
+      "loss": 3.4837,
+      "step": 1528832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.496180462969818e-05,
+      "loss": 3.4593,
+      "step": 1529344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.495341868218766e-05,
+      "loss": 3.4664,
+      "step": 1529856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.494503273467714e-05,
+      "loss": 3.4581,
+      "step": 1530368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4936646787166617e-05,
+      "loss": 3.4784,
+      "step": 1530880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4928260839656097e-05,
+      "loss": 3.4673,
+      "step": 1531392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.491989127094931e-05,
+      "loss": 3.4691,
+      "step": 1531904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4911505323438793e-05,
+      "loss": 3.4681,
+      "step": 1532416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.490311937592827e-05,
+      "loss": 3.4667,
+      "step": 1532928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.489473342841775e-05,
+      "loss": 3.457,
+      "step": 1533440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4886363859710962e-05,
+      "loss": 3.462,
+      "step": 1533952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.487797791220044e-05,
+      "loss": 3.4517,
+      "step": 1534464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.486959196468992e-05,
+      "loss": 3.4644,
+      "step": 1534976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4861206017179402e-05,
+      "loss": 3.4626,
+      "step": 1535488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4852836448472614e-05,
+      "loss": 3.461,
+      "step": 1536000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.484445050096209e-05,
+      "loss": 3.4874,
+      "step": 1536512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.483606455345157e-05,
+      "loss": 3.4655,
+      "step": 1537024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.482767860594105e-05,
+      "loss": 3.4686,
+      "step": 1537536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4819309037234263e-05,
+      "loss": 3.4725,
+      "step": 1538048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4810923089723743e-05,
+      "loss": 3.4659,
+      "step": 1538560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4802553521016956e-05,
+      "loss": 3.4631,
+      "step": 1539072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4794167573506436e-05,
+      "loss": 3.4575,
+      "step": 1539584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4785781625995912e-05,
+      "loss": 3.4727,
+      "step": 1540096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4777395678485392e-05,
+      "loss": 3.4474,
+      "step": 1540608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4769009730974872e-05,
+      "loss": 3.4587,
+      "step": 1541120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4760623783464355e-05,
+      "loss": 3.4598,
+      "step": 1541632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4752237835953835e-05,
+      "loss": 3.461,
+      "step": 1542144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4743868267247045e-05,
+      "loss": 3.4733,
+      "step": 1542656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4735482319736524e-05,
+      "loss": 3.466,
+      "step": 1543168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4727096372226004e-05,
+      "loss": 3.4617,
+      "step": 1543680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4718710424715484e-05,
+      "loss": 3.4669,
+      "step": 1544192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4710324477204964e-05,
+      "loss": 3.4746,
+      "step": 1544704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4701938529694448e-05,
+      "loss": 3.4582,
+      "step": 1545216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4693552582183924e-05,
+      "loss": 3.4691,
+      "step": 1545728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4685166634673404e-05,
+      "loss": 3.4507,
+      "step": 1546240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4676797065966617e-05,
+      "loss": 3.4693,
+      "step": 1546752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4668411118456093e-05,
+      "loss": 3.4532,
+      "step": 1547264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.466004154974931e-05,
+      "loss": 3.4798,
+      "step": 1547776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.465165560223879e-05,
+      "loss": 3.4657,
+      "step": 1548288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.464326965472827e-05,
+      "loss": 3.4596,
+      "step": 1548800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4634883707217746e-05,
+      "loss": 3.4696,
+      "step": 1549312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4626497759707226e-05,
+      "loss": 3.4586,
+      "step": 1549824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4618111812196706e-05,
+      "loss": 3.4671,
+      "step": 1550336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4609725864686185e-05,
+      "loss": 3.4563,
+      "step": 1550848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4601356295979398e-05,
+      "loss": 3.4521,
+      "step": 1551360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4592970348468878e-05,
+      "loss": 3.4734,
+      "step": 1551872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4584584400958358e-05,
+      "loss": 3.4626,
+      "step": 1552384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4576198453447838e-05,
+      "loss": 3.4664,
+      "step": 1552896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4567812505937318e-05,
+      "loss": 3.4574,
+      "step": 1553408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4559426558426798e-05,
+      "loss": 3.4648,
+      "step": 1553920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4551040610916278e-05,
+      "loss": 3.4517,
+      "step": 1554432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4542654663405758e-05,
+      "loss": 3.4662,
+      "step": 1554944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.453428509469897e-05,
+      "loss": 3.4632,
+      "step": 1555456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.452591552599218e-05,
+      "loss": 3.4788,
+      "step": 1555968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.451752957848166e-05,
+      "loss": 3.4665,
+      "step": 1556480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.450914363097114e-05,
+      "loss": 3.4553,
+      "step": 1556992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.450075768346062e-05,
+      "loss": 3.4608,
+      "step": 1557504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4492371735950102e-05,
+      "loss": 3.4651,
+      "step": 1558016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.448398578843958e-05,
+      "loss": 3.4527,
+      "step": 1558528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.447559984092906e-05,
+      "loss": 3.4575,
+      "step": 1559040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.446721389341854e-05,
+      "loss": 3.4603,
+      "step": 1559552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4458844324711748e-05,
+      "loss": 3.4646,
+      "step": 1560064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.445045837720123e-05,
+      "loss": 3.4466,
+      "step": 1560576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.444207242969071e-05,
+      "loss": 3.4536,
+      "step": 1561088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.443368648218019e-05,
+      "loss": 3.4499,
+      "step": 1561600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.44253169134734e-05,
+      "loss": 3.4689,
+      "step": 1562112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.441693096596288e-05,
+      "loss": 3.4647,
+      "step": 1562624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.440854501845236e-05,
+      "loss": 3.4588,
+      "step": 1563136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.440015907094184e-05,
+      "loss": 3.4615,
+      "step": 1563648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4391805881038785e-05,
+      "loss": 3.4596,
+      "step": 1564160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4383419933528265e-05,
+      "loss": 3.4728,
+      "step": 1564672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4375033986017745e-05,
+      "loss": 3.452,
+      "step": 1565184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4366648038507222e-05,
+      "loss": 3.4714,
+      "step": 1565696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.43582620909967e-05,
+      "loss": 3.4621,
+      "step": 1566208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4349892522289918e-05,
+      "loss": 3.4635,
+      "step": 1566720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4341506574779398e-05,
+      "loss": 3.4676,
+      "step": 1567232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4333120627268874e-05,
+      "loss": 3.4637,
+      "step": 1567744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4324734679758354e-05,
+      "loss": 3.4694,
+      "step": 1568256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4316348732247834e-05,
+      "loss": 3.4629,
+      "step": 1568768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4307962784737314e-05,
+      "loss": 3.4508,
+      "step": 1569280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4299576837226794e-05,
+      "loss": 3.4619,
+      "step": 1569792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4291190889716274e-05,
+      "loss": 3.4568,
+      "step": 1570304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4282821321009486e-05,
+      "loss": 3.4634,
+      "step": 1570816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4274435373498966e-05,
+      "loss": 3.4595,
+      "step": 1571328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4266049425988446e-05,
+      "loss": 3.4524,
+      "step": 1571840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4257663478477926e-05,
+      "loss": 3.4656,
+      "step": 1572352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.424931028857487e-05,
+      "loss": 3.4568,
+      "step": 1572864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4240924341064348e-05,
+      "loss": 3.4657,
+      "step": 1573376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4232538393553828e-05,
+      "loss": 3.4488,
+      "step": 1573888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4224152446043308e-05,
+      "loss": 3.4649,
+      "step": 1574400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4215766498532788e-05,
+      "loss": 3.4779,
+      "step": 1574912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4207380551022268e-05,
+      "loss": 3.4571,
+      "step": 1575424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.419901098231548e-05,
+      "loss": 3.4418,
+      "step": 1575936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.419062503480496e-05,
+      "loss": 3.4594,
+      "step": 1576448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.418223908729444e-05,
+      "loss": 3.4614,
+      "step": 1576960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.417385313978392e-05,
+      "loss": 3.4645,
+      "step": 1577472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.41654671922734e-05,
+      "loss": 3.464,
+      "step": 1577984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4157081244762876e-05,
+      "loss": 3.4548,
+      "step": 1578496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4148695297252356e-05,
+      "loss": 3.4592,
+      "step": 1579008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4140325728545572e-05,
+      "loss": 3.4643,
+      "step": 1579520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4131939781035052e-05,
+      "loss": 3.4502,
+      "step": 1580032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.412355383352453e-05,
+      "loss": 3.4561,
+      "step": 1580544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.411516788601401e-05,
+      "loss": 3.4583,
+      "step": 1581056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.410679831730722e-05,
+      "loss": 3.4654,
+      "step": 1581568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.40984123697967e-05,
+      "loss": 3.4391,
+      "step": 1582080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.409002642228618e-05,
+      "loss": 3.4569,
+      "step": 1582592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.408164047477566e-05,
+      "loss": 3.4518,
+      "step": 1583104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4073270906068874e-05,
+      "loss": 3.4607,
+      "step": 1583616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.406488495855835e-05,
+      "loss": 3.4565,
+      "step": 1584128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.405649901104783e-05,
+      "loss": 3.453,
+      "step": 1584640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.404811306353731e-05,
+      "loss": 3.4623,
+      "step": 1585152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4039743494830526e-05,
+      "loss": 3.4669,
+      "step": 1585664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4031373926123735e-05,
+      "loss": 3.449,
+      "step": 1586176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4022987978613215e-05,
+      "loss": 3.4648,
+      "step": 1586688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4014602031102695e-05,
+      "loss": 3.4559,
+      "step": 1587200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.400621608359217e-05,
+      "loss": 3.4608,
+      "step": 1587712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3997830136081655e-05,
+      "loss": 3.4421,
+      "step": 1588224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3989460567374867e-05,
+      "loss": 3.4646,
+      "step": 1588736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3981074619864347e-05,
+      "loss": 3.4483,
+      "step": 1589248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3972688672353824e-05,
+      "loss": 3.4634,
+      "step": 1589760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3964302724843304e-05,
+      "loss": 3.4505,
+      "step": 1590272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3955916777332784e-05,
+      "loss": 3.4648,
+      "step": 1590784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3947530829822264e-05,
+      "loss": 3.4587,
+      "step": 1591296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3939144882311744e-05,
+      "loss": 3.4466,
+      "step": 1591808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3930775313604956e-05,
+      "loss": 3.4454,
+      "step": 1592320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3922389366094436e-05,
+      "loss": 3.4695,
+      "step": 1592832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3914003418583916e-05,
+      "loss": 3.4582,
+      "step": 1593344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3905617471073396e-05,
+      "loss": 3.4635,
+      "step": 1593856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3897247902366605e-05,
+      "loss": 3.4538,
+      "step": 1594368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.388886195485609e-05,
+      "loss": 3.463,
+      "step": 1594880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.388047600734557e-05,
+      "loss": 3.4621,
+      "step": 1595392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.387209005983505e-05,
+      "loss": 3.4495,
+      "step": 1595904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.386370411232453e-05,
+      "loss": 3.4537,
+      "step": 1596416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3855318164814005e-05,
+      "loss": 3.4629,
+      "step": 1596928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3846932217303485e-05,
+      "loss": 3.4633,
+      "step": 1597440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3838562648596697e-05,
+      "loss": 3.4677,
+      "step": 1597952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3830176701086177e-05,
+      "loss": 3.448,
+      "step": 1598464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3821790753575657e-05,
+      "loss": 3.4617,
+      "step": 1598976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3813404806065137e-05,
+      "loss": 3.455,
+      "step": 1599488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3805018858554617e-05,
+      "loss": 3.4575,
+      "step": 1600000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3796632911044097e-05,
+      "loss": 3.4644,
+      "step": 1600512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3788246963533577e-05,
+      "loss": 3.4551,
+      "step": 1601024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3779861016023057e-05,
+      "loss": 3.4667,
+      "step": 1601536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.377149144731627e-05,
+      "loss": 3.4487,
+      "step": 1602048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.376310549980575e-05,
+      "loss": 3.4673,
+      "step": 1602560
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9099957942962646,
+      "eval_runtime": 304.3572,
+      "eval_samples_per_second": 1253.76,
+      "eval_steps_per_second": 39.181,
+      "step": 1602720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.375471955229523e-05,
+      "loss": 3.45,
+      "step": 1603072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.374633360478471e-05,
+      "loss": 3.4608,
+      "step": 1603584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.373796403607792e-05,
+      "loss": 3.4616,
+      "step": 1604096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.37295780885674e-05,
+      "loss": 3.4622,
+      "step": 1604608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3721192141056882e-05,
+      "loss": 3.4689,
+      "step": 1605120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.371280619354636e-05,
+      "loss": 3.45,
+      "step": 1605632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.370442024603584e-05,
+      "loss": 3.4531,
+      "step": 1606144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3696034298525318e-05,
+      "loss": 3.4459,
+      "step": 1606656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3687648351014798e-05,
+      "loss": 3.4659,
+      "step": 1607168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3679262403504278e-05,
+      "loss": 3.4616,
+      "step": 1607680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.367089283479749e-05,
+      "loss": 3.4544,
+      "step": 1608192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.366250688728697e-05,
+      "loss": 3.4609,
+      "step": 1608704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.365412093977645e-05,
+      "loss": 3.4477,
+      "step": 1609216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.364573499226593e-05,
+      "loss": 3.4475,
+      "step": 1609728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.363736542355914e-05,
+      "loss": 3.4512,
+      "step": 1610240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.362897947604862e-05,
+      "loss": 3.4392,
+      "step": 1610752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3620593528538103e-05,
+      "loss": 3.4508,
+      "step": 1611264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3612207581027583e-05,
+      "loss": 3.4486,
+      "step": 1611776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3603838012320792e-05,
+      "loss": 3.4485,
+      "step": 1612288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3595468443614005e-05,
+      "loss": 3.4802,
+      "step": 1612800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.358708249610348e-05,
+      "loss": 3.45,
+      "step": 1613312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3578696548592964e-05,
+      "loss": 3.4572,
+      "step": 1613824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3570310601082444e-05,
+      "loss": 3.4589,
+      "step": 1614336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3561924653571924e-05,
+      "loss": 3.4539,
+      "step": 1614848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3553538706061404e-05,
+      "loss": 3.4476,
+      "step": 1615360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3545169137354613e-05,
+      "loss": 3.4519,
+      "step": 1615872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3536783189844093e-05,
+      "loss": 3.4569,
+      "step": 1616384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3528397242333573e-05,
+      "loss": 3.4409,
+      "step": 1616896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3520011294823057e-05,
+      "loss": 3.4454,
+      "step": 1617408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3511625347312537e-05,
+      "loss": 3.4452,
+      "step": 1617920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3503239399802013e-05,
+      "loss": 3.455,
+      "step": 1618432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3494869831095226e-05,
+      "loss": 3.4563,
+      "step": 1618944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3486483883584706e-05,
+      "loss": 3.4546,
+      "step": 1619456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3478097936074186e-05,
+      "loss": 3.4523,
+      "step": 1619968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3469711988563666e-05,
+      "loss": 3.4538,
+      "step": 1620480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3461326041053145e-05,
+      "loss": 3.4593,
+      "step": 1620992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3452940093542625e-05,
+      "loss": 3.4469,
+      "step": 1621504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3444554146032105e-05,
+      "loss": 3.4582,
+      "step": 1622016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3436168198521585e-05,
+      "loss": 3.4383,
+      "step": 1622528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3427798629814794e-05,
+      "loss": 3.4598,
+      "step": 1623040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3419412682304274e-05,
+      "loss": 3.4361,
+      "step": 1623552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3411043113597487e-05,
+      "loss": 3.4717,
+      "step": 1624064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3402657166086967e-05,
+      "loss": 3.4529,
+      "step": 1624576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3394271218576447e-05,
+      "loss": 3.4504,
+      "step": 1625088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3385885271065927e-05,
+      "loss": 3.4572,
+      "step": 1625600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3377499323555407e-05,
+      "loss": 3.4522,
+      "step": 1626112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3369113376044887e-05,
+      "loss": 3.4526,
+      "step": 1626624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3360727428534367e-05,
+      "loss": 3.4418,
+      "step": 1627136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.335235785982758e-05,
+      "loss": 3.4366,
+      "step": 1627648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.334397191231706e-05,
+      "loss": 3.4638,
+      "step": 1628160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.333558596480654e-05,
+      "loss": 3.4513,
+      "step": 1628672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.332720001729602e-05,
+      "loss": 3.4577,
+      "step": 1629184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3318814069785496e-05,
+      "loss": 3.4459,
+      "step": 1629696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3310428122274975e-05,
+      "loss": 3.4517,
+      "step": 1630208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.330204217476446e-05,
+      "loss": 3.4391,
+      "step": 1630720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.329365622725394e-05,
+      "loss": 3.456,
+      "step": 1631232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3285286658547148e-05,
+      "loss": 3.4498,
+      "step": 1631744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.327691708984036e-05,
+      "loss": 3.4646,
+      "step": 1632256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.326853114232984e-05,
+      "loss": 3.4596,
+      "step": 1632768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.326014519481932e-05,
+      "loss": 3.4436,
+      "step": 1633280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.32517592473088e-05,
+      "loss": 3.449,
+      "step": 1633792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3243389678602013e-05,
+      "loss": 3.4583,
+      "step": 1634304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3235003731091493e-05,
+      "loss": 3.4411,
+      "step": 1634816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.322661778358097e-05,
+      "loss": 3.4453,
+      "step": 1635328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.321823183607045e-05,
+      "loss": 3.4511,
+      "step": 1635840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.320984588855993e-05,
+      "loss": 3.4509,
+      "step": 1636352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3201459941049412e-05,
+      "loss": 3.4362,
+      "step": 1636864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3193073993538892e-05,
+      "loss": 3.4433,
+      "step": 1637376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3184688046028372e-05,
+      "loss": 3.4379,
+      "step": 1637888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.317631847732158e-05,
+      "loss": 3.4569,
+      "step": 1638400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.316794890861479e-05,
+      "loss": 3.4521,
+      "step": 1638912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3159562961104274e-05,
+      "loss": 3.446,
+      "step": 1639424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3151177013593754e-05,
+      "loss": 3.4542,
+      "step": 1639936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3142791066083234e-05,
+      "loss": 3.4428,
+      "step": 1640448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3134405118572714e-05,
+      "loss": 3.4626,
+      "step": 1640960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3126019171062194e-05,
+      "loss": 3.4396,
+      "step": 1641472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.311763322355167e-05,
+      "loss": 3.4585,
+      "step": 1641984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3109263654844883e-05,
+      "loss": 3.4512,
+      "step": 1642496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3100894086138095e-05,
+      "loss": 3.4504,
+      "step": 1643008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3092508138627575e-05,
+      "loss": 3.4568,
+      "step": 1643520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3084122191117055e-05,
+      "loss": 3.4525,
+      "step": 1644032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3075736243606535e-05,
+      "loss": 3.4536,
+      "step": 1644544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3067350296096015e-05,
+      "loss": 3.4512,
+      "step": 1645056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3058964348585495e-05,
+      "loss": 3.4395,
+      "step": 1645568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3050578401074975e-05,
+      "loss": 3.4513,
+      "step": 1646080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3042192453564455e-05,
+      "loss": 3.4473,
+      "step": 1646592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3033822884857667e-05,
+      "loss": 3.452,
+      "step": 1647104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3025436937347144e-05,
+      "loss": 3.4451,
+      "step": 1647616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3017050989836624e-05,
+      "loss": 3.4407,
+      "step": 1648128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3008665042326104e-05,
+      "loss": 3.4511,
+      "step": 1648640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.300031185242305e-05,
+      "loss": 3.4503,
+      "step": 1649152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.299192590491253e-05,
+      "loss": 3.4502,
+      "step": 1649664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.298353995740201e-05,
+      "loss": 3.4382,
+      "step": 1650176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.297515400989149e-05,
+      "loss": 3.4548,
+      "step": 1650688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.296676806238097e-05,
+      "loss": 3.4623,
+      "step": 1651200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2958382114870445e-05,
+      "loss": 3.4491,
+      "step": 1651712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.295001254616366e-05,
+      "loss": 3.4294,
+      "step": 1652224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.294162659865314e-05,
+      "loss": 3.4491,
+      "step": 1652736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2933240651142618e-05,
+      "loss": 3.4485,
+      "step": 1653248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2924854703632098e-05,
+      "loss": 3.4508,
+      "step": 1653760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2916468756121578e-05,
+      "loss": 3.4543,
+      "step": 1654272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2908082808611058e-05,
+      "loss": 3.4439,
+      "step": 1654784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2899696861100538e-05,
+      "loss": 3.4462,
+      "step": 1655296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.289131091359002e-05,
+      "loss": 3.4521,
+      "step": 1655808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.288294134488323e-05,
+      "loss": 3.4378,
+      "step": 1656320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.287455539737271e-05,
+      "loss": 3.4454,
+      "step": 1656832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.286616944986219e-05,
+      "loss": 3.4448,
+      "step": 1657344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.28577998811554e-05,
+      "loss": 3.4537,
+      "step": 1657856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2849413933644882e-05,
+      "loss": 3.4302,
+      "step": 1658368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2841027986134362e-05,
+      "loss": 3.4421,
+      "step": 1658880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2832642038623842e-05,
+      "loss": 3.4427,
+      "step": 1659392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.282427246991705e-05,
+      "loss": 3.4484,
+      "step": 1659904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.281588652240653e-05,
+      "loss": 3.4493,
+      "step": 1660416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.280750057489601e-05,
+      "loss": 3.4375,
+      "step": 1660928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.279911462738549e-05,
+      "loss": 3.4527,
+      "step": 1661440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2790745058678704e-05,
+      "loss": 3.4544,
+      "step": 1661952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2782359111168184e-05,
+      "loss": 3.4356,
+      "step": 1662464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2773989542461393e-05,
+      "loss": 3.4551,
+      "step": 1662976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2765603594950873e-05,
+      "loss": 3.4402,
+      "step": 1663488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2757217647440353e-05,
+      "loss": 3.4529,
+      "step": 1664000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2748831699929836e-05,
+      "loss": 3.4364,
+      "step": 1664512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2740462131223045e-05,
+      "loss": 3.4491,
+      "step": 1665024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2732076183712525e-05,
+      "loss": 3.4392,
+      "step": 1665536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2723690236202005e-05,
+      "loss": 3.4486,
+      "step": 1666048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2715304288691485e-05,
+      "loss": 3.4412,
+      "step": 1666560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2706934719984697e-05,
+      "loss": 3.4505,
+      "step": 1667072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2698548772474177e-05,
+      "loss": 3.4457,
+      "step": 1667584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2690162824963657e-05,
+      "loss": 3.4399,
+      "step": 1668096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2681776877453137e-05,
+      "loss": 3.4289,
+      "step": 1668608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2673390929942617e-05,
+      "loss": 3.4603,
+      "step": 1669120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2665004982432097e-05,
+      "loss": 3.4436,
+      "step": 1669632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2656619034921574e-05,
+      "loss": 3.4521,
+      "step": 1670144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.264824946621479e-05,
+      "loss": 3.4392,
+      "step": 1670656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.263986351870427e-05,
+      "loss": 3.4538,
+      "step": 1671168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2631477571193746e-05,
+      "loss": 3.4494,
+      "step": 1671680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2623091623683226e-05,
+      "loss": 3.4395,
+      "step": 1672192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2614705676172706e-05,
+      "loss": 3.4425,
+      "step": 1672704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2606319728662186e-05,
+      "loss": 3.4499,
+      "step": 1673216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2597933781151666e-05,
+      "loss": 3.4502,
+      "step": 1673728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.258956421244488e-05,
+      "loss": 3.4626,
+      "step": 1674240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.258117826493436e-05,
+      "loss": 3.4331,
+      "step": 1674752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.257279231742384e-05,
+      "loss": 3.4519,
+      "step": 1675264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.256440636991332e-05,
+      "loss": 3.4399,
+      "step": 1675776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.25560204224028e-05,
+      "loss": 3.4496,
+      "step": 1676288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2547634474892275e-05,
+      "loss": 3.4493,
+      "step": 1676800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2539248527381758e-05,
+      "loss": 3.4393,
+      "step": 1677312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2530862579871238e-05,
+      "loss": 3.4562,
+      "step": 1677824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.252249301116445e-05,
+      "loss": 3.4414,
+      "step": 1678336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2514107063653927e-05,
+      "loss": 3.4523,
+      "step": 1678848
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9129364490509033,
+      "eval_runtime": 304.133,
+      "eval_samples_per_second": 1254.685,
+      "eval_steps_per_second": 39.21,
+      "step": 1679040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2505721116143407e-05,
+      "loss": 3.4406,
+      "step": 1679360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2497335168632887e-05,
+      "loss": 3.4447,
+      "step": 1679872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.24889655999261e-05,
+      "loss": 3.4515,
+      "step": 1680384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.248057965241558e-05,
+      "loss": 3.4508,
+      "step": 1680896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.247219370490506e-05,
+      "loss": 3.4556,
+      "step": 1681408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.246380775739454e-05,
+      "loss": 3.4432,
+      "step": 1681920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.245542180988402e-05,
+      "loss": 3.4358,
+      "step": 1682432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.24470358623735e-05,
+      "loss": 3.4402,
+      "step": 1682944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.243864991486298e-05,
+      "loss": 3.4505,
+      "step": 1683456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.243026396735246e-05,
+      "loss": 3.4492,
+      "step": 1683968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2421894398645672e-05,
+      "loss": 3.4416,
+      "step": 1684480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2413508451135152e-05,
+      "loss": 3.4495,
+      "step": 1684992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2405122503624632e-05,
+      "loss": 3.4318,
+      "step": 1685504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.239673655611411e-05,
+      "loss": 3.4422,
+      "step": 1686016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.238836698740732e-05,
+      "loss": 3.4411,
+      "step": 1686528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.23799810398968e-05,
+      "loss": 3.4294,
+      "step": 1687040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2371595092386284e-05,
+      "loss": 3.4419,
+      "step": 1687552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.236320914487576e-05,
+      "loss": 3.4332,
+      "step": 1688064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2354839576168973e-05,
+      "loss": 3.4369,
+      "step": 1688576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2346453628658453e-05,
+      "loss": 3.4703,
+      "step": 1689088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.233806768114793e-05,
+      "loss": 3.441,
+      "step": 1689600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2329698112441146e-05,
+      "loss": 3.4408,
+      "step": 1690112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2321312164930625e-05,
+      "loss": 3.4469,
+      "step": 1690624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2312926217420105e-05,
+      "loss": 3.4481,
+      "step": 1691136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2304540269909582e-05,
+      "loss": 3.4301,
+      "step": 1691648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2296154322399062e-05,
+      "loss": 3.4419,
+      "step": 1692160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2287768374888542e-05,
+      "loss": 3.4437,
+      "step": 1692672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2279398806181754e-05,
+      "loss": 3.4329,
+      "step": 1693184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2271012858671234e-05,
+      "loss": 3.4361,
+      "step": 1693696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2262626911160714e-05,
+      "loss": 3.4321,
+      "step": 1694208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2254240963650194e-05,
+      "loss": 3.4436,
+      "step": 1694720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2245871394943403e-05,
+      "loss": 3.4482,
+      "step": 1695232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2237485447432883e-05,
+      "loss": 3.4429,
+      "step": 1695744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2229099499922367e-05,
+      "loss": 3.4418,
+      "step": 1696256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2220713552411847e-05,
+      "loss": 3.4408,
+      "step": 1696768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2212327604901327e-05,
+      "loss": 3.4493,
+      "step": 1697280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2203941657390807e-05,
+      "loss": 3.436,
+      "step": 1697792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2195555709880287e-05,
+      "loss": 3.4521,
+      "step": 1698304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2187169762369763e-05,
+      "loss": 3.4247,
+      "step": 1698816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2178800193662976e-05,
+      "loss": 3.446,
+      "step": 1699328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2170430624956188e-05,
+      "loss": 3.428,
+      "step": 1699840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2162044677445668e-05,
+      "loss": 3.4596,
+      "step": 1700352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2153658729935148e-05,
+      "loss": 3.4414,
+      "step": 1700864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2145272782424628e-05,
+      "loss": 3.4427,
+      "step": 1701376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2136886834914108e-05,
+      "loss": 3.4442,
+      "step": 1701888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2128500887403584e-05,
+      "loss": 3.4414,
+      "step": 1702400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2120114939893068e-05,
+      "loss": 3.4438,
+      "step": 1702912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2111728992382548e-05,
+      "loss": 3.4323,
+      "step": 1703424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.210335942367576e-05,
+      "loss": 3.4219,
+      "step": 1703936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2094973476165237e-05,
+      "loss": 3.4547,
+      "step": 1704448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2086587528654717e-05,
+      "loss": 3.4349,
+      "step": 1704960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2078201581144197e-05,
+      "loss": 3.4508,
+      "step": 1705472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2069815633633677e-05,
+      "loss": 3.4314,
+      "step": 1705984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.206142968612316e-05,
+      "loss": 3.4406,
+      "step": 1706496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.205304373861264e-05,
+      "loss": 3.4257,
+      "step": 1707008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.204465779110212e-05,
+      "loss": 3.4454,
+      "step": 1707520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.203628822239533e-05,
+      "loss": 3.441,
+      "step": 1708032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2027918653688538e-05,
+      "loss": 3.4498,
+      "step": 1708544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.201953270617802e-05,
+      "loss": 3.4486,
+      "step": 1709056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.20111467586675e-05,
+      "loss": 3.4312,
+      "step": 1709568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.200276081115698e-05,
+      "loss": 3.4352,
+      "step": 1710080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.199439124245019e-05,
+      "loss": 3.4518,
+      "step": 1710592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.198600529493967e-05,
+      "loss": 3.4281,
+      "step": 1711104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.197761934742915e-05,
+      "loss": 3.433,
+      "step": 1711616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.196923339991863e-05,
+      "loss": 3.4376,
+      "step": 1712128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1960847452408114e-05,
+      "loss": 3.4466,
+      "step": 1712640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.195246150489759e-05,
+      "loss": 3.423,
+      "step": 1713152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.194407555738707e-05,
+      "loss": 3.4311,
+      "step": 1713664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.193568960987655e-05,
+      "loss": 3.4288,
+      "step": 1714176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1927320041169763e-05,
+      "loss": 3.436,
+      "step": 1714688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1918950472462975e-05,
+      "loss": 3.4447,
+      "step": 1715200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1910564524952455e-05,
+      "loss": 3.4362,
+      "step": 1715712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1902178577441935e-05,
+      "loss": 3.4426,
+      "step": 1716224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1893792629931415e-05,
+      "loss": 3.4301,
+      "step": 1716736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.188540668242089e-05,
+      "loss": 3.4551,
+      "step": 1717248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.187702073491037e-05,
+      "loss": 3.4292,
+      "step": 1717760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1868651166203584e-05,
+      "loss": 3.4482,
+      "step": 1718272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1860265218693064e-05,
+      "loss": 3.4405,
+      "step": 1718784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1851879271182544e-05,
+      "loss": 3.438,
+      "step": 1719296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1843493323672024e-05,
+      "loss": 3.4482,
+      "step": 1719808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1835107376161504e-05,
+      "loss": 3.4424,
+      "step": 1720320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1826737807454713e-05,
+      "loss": 3.4379,
+      "step": 1720832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1818351859944193e-05,
+      "loss": 3.4425,
+      "step": 1721344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1809965912433676e-05,
+      "loss": 3.4302,
+      "step": 1721856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1801579964923156e-05,
+      "loss": 3.4363,
+      "step": 1722368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1793194017412636e-05,
+      "loss": 3.4384,
+      "step": 1722880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1784808069902116e-05,
+      "loss": 3.4411,
+      "step": 1723392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1776422122391596e-05,
+      "loss": 3.434,
+      "step": 1723904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1768036174881073e-05,
+      "loss": 3.4292,
+      "step": 1724416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1759666606174285e-05,
+      "loss": 3.4365,
+      "step": 1724928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1751297037467498e-05,
+      "loss": 3.4422,
+      "step": 1725440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1742911089956978e-05,
+      "loss": 3.44,
+      "step": 1725952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1734525142446457e-05,
+      "loss": 3.4244,
+      "step": 1726464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1726139194935937e-05,
+      "loss": 3.4432,
+      "step": 1726976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1717753247425417e-05,
+      "loss": 3.4507,
+      "step": 1727488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1709367299914894e-05,
+      "loss": 3.4377,
+      "step": 1728000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.170099773120811e-05,
+      "loss": 3.4201,
+      "step": 1728512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.169261178369759e-05,
+      "loss": 3.4378,
+      "step": 1729024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.168422583618707e-05,
+      "loss": 3.4319,
+      "step": 1729536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1675839888676546e-05,
+      "loss": 3.4422,
+      "step": 1730048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1667453941166026e-05,
+      "loss": 3.4437,
+      "step": 1730560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1659067993655506e-05,
+      "loss": 3.4319,
+      "step": 1731072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1650682046144986e-05,
+      "loss": 3.435,
+      "step": 1731584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.164229609863447e-05,
+      "loss": 3.4427,
+      "step": 1732096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.163392652992768e-05,
+      "loss": 3.4284,
+      "step": 1732608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.162554058241716e-05,
+      "loss": 3.4355,
+      "step": 1733120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.161715463490664e-05,
+      "loss": 3.4311,
+      "step": 1733632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1608785066199848e-05,
+      "loss": 3.4459,
+      "step": 1734144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.160039911868933e-05,
+      "loss": 3.4181,
+      "step": 1734656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.159201317117881e-05,
+      "loss": 3.4279,
+      "step": 1735168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.158362722366829e-05,
+      "loss": 3.4318,
+      "step": 1735680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.15752576549615e-05,
+      "loss": 3.4403,
+      "step": 1736192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.156687170745098e-05,
+      "loss": 3.434,
+      "step": 1736704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.155848575994046e-05,
+      "loss": 3.43,
+      "step": 1737216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.155009981242994e-05,
+      "loss": 3.4362,
+      "step": 1737728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1541730243723152e-05,
+      "loss": 3.4427,
+      "step": 1738240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1533344296212632e-05,
+      "loss": 3.425,
+      "step": 1738752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.152497472750584e-05,
+      "loss": 3.4468,
+      "step": 1739264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.151658877999532e-05,
+      "loss": 3.4296,
+      "step": 1739776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.15082028324848e-05,
+      "loss": 3.4372,
+      "step": 1740288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1499816884974285e-05,
+      "loss": 3.4265,
+      "step": 1740800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1491447316267494e-05,
+      "loss": 3.4352,
+      "step": 1741312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1483061368756974e-05,
+      "loss": 3.4284,
+      "step": 1741824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1474675421246454e-05,
+      "loss": 3.4359,
+      "step": 1742336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1466289473735934e-05,
+      "loss": 3.4373,
+      "step": 1742848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1457903526225414e-05,
+      "loss": 3.4344,
+      "step": 1743360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1449533957518626e-05,
+      "loss": 3.4337,
+      "step": 1743872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1441148010008106e-05,
+      "loss": 3.4317,
+      "step": 1744384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1432762062497586e-05,
+      "loss": 3.42,
+      "step": 1744896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1424376114987066e-05,
+      "loss": 3.4466,
+      "step": 1745408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1415990167476546e-05,
+      "loss": 3.4341,
+      "step": 1745920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1407604219966022e-05,
+      "loss": 3.439,
+      "step": 1746432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1399234651259238e-05,
+      "loss": 3.4289,
+      "step": 1746944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1390848703748718e-05,
+      "loss": 3.4476,
+      "step": 1747456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1382462756238195e-05,
+      "loss": 3.434,
+      "step": 1747968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1374076808727675e-05,
+      "loss": 3.4281,
+      "step": 1748480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1365690861217155e-05,
+      "loss": 3.4324,
+      "step": 1748992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1357321292510367e-05,
+      "loss": 3.4361,
+      "step": 1749504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1348935344999847e-05,
+      "loss": 3.4386,
+      "step": 1750016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1340549397489327e-05,
+      "loss": 3.4519,
+      "step": 1750528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1332163449978807e-05,
+      "loss": 3.4219,
+      "step": 1751040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1323777502468287e-05,
+      "loss": 3.4401,
+      "step": 1751552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1315391554957767e-05,
+      "loss": 3.431,
+      "step": 1752064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1307021986250976e-05,
+      "loss": 3.4415,
+      "step": 1752576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1298636038740456e-05,
+      "loss": 3.4344,
+      "step": 1753088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.129025009122994e-05,
+      "loss": 3.4326,
+      "step": 1753600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.128186414371942e-05,
+      "loss": 3.4428,
+      "step": 1754112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.127349457501263e-05,
+      "loss": 3.4308,
+      "step": 1754624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.126510862750211e-05,
+      "loss": 3.4403,
+      "step": 1755136
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.915910482406616,
+      "eval_runtime": 302.2896,
+      "eval_samples_per_second": 1262.336,
+      "eval_steps_per_second": 39.449,
+      "step": 1755360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.125672267999159e-05,
+      "loss": 3.434,
+      "step": 1755648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1248336732481068e-05,
+      "loss": 3.4309,
+      "step": 1756160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.123996716377428e-05,
+      "loss": 3.4404,
+      "step": 1756672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.123158121626376e-05,
+      "loss": 3.44,
+      "step": 1757184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.122319526875324e-05,
+      "loss": 3.4486,
+      "step": 1757696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.121480932124272e-05,
+      "loss": 3.43,
+      "step": 1758208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.12064233737322e-05,
+      "loss": 3.4257,
+      "step": 1758720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1198037426221677e-05,
+      "loss": 3.4267,
+      "step": 1759232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1189667857514893e-05,
+      "loss": 3.4433,
+      "step": 1759744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1181281910004373e-05,
+      "loss": 3.4406,
+      "step": 1760256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.117289596249385e-05,
+      "loss": 3.428,
+      "step": 1760768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.116451001498333e-05,
+      "loss": 3.4377,
+      "step": 1761280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.115612406747281e-05,
+      "loss": 3.4261,
+      "step": 1761792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.114773811996229e-05,
+      "loss": 3.427,
+      "step": 1762304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1139368551255502e-05,
+      "loss": 3.4269,
+      "step": 1762816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1130982603744982e-05,
+      "loss": 3.4208,
+      "step": 1763328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1122596656234462e-05,
+      "loss": 3.4276,
+      "step": 1763840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1114210708723942e-05,
+      "loss": 3.4247,
+      "step": 1764352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1105824761213422e-05,
+      "loss": 3.4239,
+      "step": 1764864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.109745519250663e-05,
+      "loss": 3.4577,
+      "step": 1765376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.108906924499611e-05,
+      "loss": 3.4327,
+      "step": 1765888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1080683297485594e-05,
+      "loss": 3.4296,
+      "step": 1766400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1072297349975074e-05,
+      "loss": 3.4341,
+      "step": 1766912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1063911402464554e-05,
+      "loss": 3.4359,
+      "step": 1767424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1055541833757763e-05,
+      "loss": 3.4205,
+      "step": 1767936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1047155886247243e-05,
+      "loss": 3.4314,
+      "step": 1768448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1038769938736723e-05,
+      "loss": 3.4317,
+      "step": 1768960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1030383991226203e-05,
+      "loss": 3.4234,
+      "step": 1769472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1021998043715683e-05,
+      "loss": 3.4219,
+      "step": 1769984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1013612096205163e-05,
+      "loss": 3.422,
+      "step": 1770496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1005226148694643e-05,
+      "loss": 3.4339,
+      "step": 1771008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0996856579987855e-05,
+      "loss": 3.4359,
+      "step": 1771520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0988470632477332e-05,
+      "loss": 3.4304,
+      "step": 1772032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0980084684966815e-05,
+      "loss": 3.4261,
+      "step": 1772544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0971698737456295e-05,
+      "loss": 3.4271,
+      "step": 1773056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0963312789945775e-05,
+      "loss": 3.4353,
+      "step": 1773568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0954926842435255e-05,
+      "loss": 3.4299,
+      "step": 1774080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0946540894924735e-05,
+      "loss": 3.4369,
+      "step": 1774592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.093815494741421e-05,
+      "loss": 3.4174,
+      "step": 1775104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0929801757511157e-05,
+      "loss": 3.4302,
+      "step": 1775616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0921415810000637e-05,
+      "loss": 3.4199,
+      "step": 1776128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0913029862490117e-05,
+      "loss": 3.4443,
+      "step": 1776640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0904643914979597e-05,
+      "loss": 3.4316,
+      "step": 1777152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0896257967469077e-05,
+      "loss": 3.431,
+      "step": 1777664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0887888398762286e-05,
+      "loss": 3.4361,
+      "step": 1778176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.087950245125177e-05,
+      "loss": 3.4288,
+      "step": 1778688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.087111650374125e-05,
+      "loss": 3.4333,
+      "step": 1779200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.086273055623073e-05,
+      "loss": 3.4235,
+      "step": 1779712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0854360987523938e-05,
+      "loss": 3.4045,
+      "step": 1780224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0845975040013418e-05,
+      "loss": 3.4438,
+      "step": 1780736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0837589092502898e-05,
+      "loss": 3.4269,
+      "step": 1781248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0829203144992378e-05,
+      "loss": 3.4402,
+      "step": 1781760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0820817197481858e-05,
+      "loss": 3.4253,
+      "step": 1782272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0812431249971338e-05,
+      "loss": 3.4282,
+      "step": 1782784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.080406168126455e-05,
+      "loss": 3.4136,
+      "step": 1783296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.079567573375403e-05,
+      "loss": 3.4336,
+      "step": 1783808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.078728978624351e-05,
+      "loss": 3.4298,
+      "step": 1784320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0778903838732987e-05,
+      "loss": 3.4358,
+      "step": 1784832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.077051789122247e-05,
+      "loss": 3.4426,
+      "step": 1785344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.076213194371195e-05,
+      "loss": 3.4211,
+      "step": 1785856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.075374599620143e-05,
+      "loss": 3.424,
+      "step": 1786368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.074537642749464e-05,
+      "loss": 3.4396,
+      "step": 1786880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.073699047998412e-05,
+      "loss": 3.4167,
+      "step": 1787392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.07286045324736e-05,
+      "loss": 3.4189,
+      "step": 1787904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.072021858496308e-05,
+      "loss": 3.4288,
+      "step": 1788416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0711832637452562e-05,
+      "loss": 3.4316,
+      "step": 1788928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.070346306874577e-05,
+      "loss": 3.4143,
+      "step": 1789440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.069507712123525e-05,
+      "loss": 3.4203,
+      "step": 1789952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.068669117372473e-05,
+      "loss": 3.4136,
+      "step": 1790464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.067830522621421e-05,
+      "loss": 3.4264,
+      "step": 1790976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0669919278703688e-05,
+      "loss": 3.434,
+      "step": 1791488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.066153333119317e-05,
+      "loss": 3.4272,
+      "step": 1792000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.065314738368265e-05,
+      "loss": 3.4271,
+      "step": 1792512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0644777814975864e-05,
+      "loss": 3.4238,
+      "step": 1793024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.063639186746534e-05,
+      "loss": 3.4398,
+      "step": 1793536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.062800591995482e-05,
+      "loss": 3.4164,
+      "step": 1794048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.06196199724443e-05,
+      "loss": 3.4396,
+      "step": 1794560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0611250403737516e-05,
+      "loss": 3.425,
+      "step": 1795072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0602864456226992e-05,
+      "loss": 3.422,
+      "step": 1795584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0594478508716472e-05,
+      "loss": 3.4385,
+      "step": 1796096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0586092561205952e-05,
+      "loss": 3.4308,
+      "step": 1796608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.057772299249916e-05,
+      "loss": 3.4291,
+      "step": 1797120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.056933704498864e-05,
+      "loss": 3.4335,
+      "step": 1797632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0560951097478125e-05,
+      "loss": 3.4203,
+      "step": 1798144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0552565149967605e-05,
+      "loss": 3.425,
+      "step": 1798656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0544179202457085e-05,
+      "loss": 3.4263,
+      "step": 1799168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0535793254946565e-05,
+      "loss": 3.4337,
+      "step": 1799680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0527407307436045e-05,
+      "loss": 3.4207,
+      "step": 1800192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.051902135992552e-05,
+      "loss": 3.4215,
+      "step": 1800704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0510651791218734e-05,
+      "loss": 3.4207,
+      "step": 1801216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0502282222511946e-05,
+      "loss": 3.4309,
+      "step": 1801728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0493896275001426e-05,
+      "loss": 3.4312,
+      "step": 1802240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0485510327490906e-05,
+      "loss": 3.4162,
+      "step": 1802752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0477124379980386e-05,
+      "loss": 3.4328,
+      "step": 1803264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0468738432469866e-05,
+      "loss": 3.4355,
+      "step": 1803776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0460352484959343e-05,
+      "loss": 3.4287,
+      "step": 1804288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.045198291625256e-05,
+      "loss": 3.4117,
+      "step": 1804800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.044359696874204e-05,
+      "loss": 3.425,
+      "step": 1805312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.043521102123152e-05,
+      "loss": 3.4173,
+      "step": 1805824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0426825073720995e-05,
+      "loss": 3.432,
+      "step": 1806336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0418439126210475e-05,
+      "loss": 3.4323,
+      "step": 1806848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0410053178699955e-05,
+      "loss": 3.4213,
+      "step": 1807360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0401667231189435e-05,
+      "loss": 3.4245,
+      "step": 1807872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0393281283678918e-05,
+      "loss": 3.4254,
+      "step": 1808384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0384911714972127e-05,
+      "loss": 3.4222,
+      "step": 1808896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0376525767461607e-05,
+      "loss": 3.4259,
+      "step": 1809408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0368139819951087e-05,
+      "loss": 3.4187,
+      "step": 1809920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0359753872440567e-05,
+      "loss": 3.4379,
+      "step": 1810432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.035138430373378e-05,
+      "loss": 3.4068,
+      "step": 1810944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0343014735026992e-05,
+      "loss": 3.4165,
+      "step": 1811456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.033462878751647e-05,
+      "loss": 3.4169,
+      "step": 1811968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.032624284000595e-05,
+      "loss": 3.4299,
+      "step": 1812480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.031785689249543e-05,
+      "loss": 3.4187,
+      "step": 1812992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.030947094498491e-05,
+      "loss": 3.4193,
+      "step": 1813504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.030108499747439e-05,
+      "loss": 3.4276,
+      "step": 1814016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0292699049963872e-05,
+      "loss": 3.4345,
+      "step": 1814528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0284313102453352e-05,
+      "loss": 3.4141,
+      "step": 1815040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.027594353374656e-05,
+      "loss": 3.4304,
+      "step": 1815552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.026755758623604e-05,
+      "loss": 3.4252,
+      "step": 1816064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.025917163872552e-05,
+      "loss": 3.4246,
+      "step": 1816576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0250785691215e-05,
+      "loss": 3.4169,
+      "step": 1817088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0242416122508213e-05,
+      "loss": 3.4221,
+      "step": 1817600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0234030174997693e-05,
+      "loss": 3.4204,
+      "step": 1818112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0225644227487173e-05,
+      "loss": 3.4263,
+      "step": 1818624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.021725827997665e-05,
+      "loss": 3.4273,
+      "step": 1819136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0208888711269862e-05,
+      "loss": 3.4224,
+      "step": 1819648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0200502763759342e-05,
+      "loss": 3.4228,
+      "step": 1820160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0192116816248825e-05,
+      "loss": 3.4192,
+      "step": 1820672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0183730868738302e-05,
+      "loss": 3.4077,
+      "step": 1821184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0175344921227782e-05,
+      "loss": 3.4383,
+      "step": 1821696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0166958973717262e-05,
+      "loss": 3.4224,
+      "step": 1822208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0158573026206742e-05,
+      "loss": 3.4279,
+      "step": 1822720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.015020345749995e-05,
+      "loss": 3.4176,
+      "step": 1823232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0141817509989434e-05,
+      "loss": 3.4346,
+      "step": 1823744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0133431562478914e-05,
+      "loss": 3.4217,
+      "step": 1824256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0125045614968394e-05,
+      "loss": 3.4179,
+      "step": 1824768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0116659667457874e-05,
+      "loss": 3.4221,
+      "step": 1825280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0108290098751083e-05,
+      "loss": 3.4217,
+      "step": 1825792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0099904151240563e-05,
+      "loss": 3.4275,
+      "step": 1826304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0091518203730043e-05,
+      "loss": 3.4464,
+      "step": 1826816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0083132256219527e-05,
+      "loss": 3.4069,
+      "step": 1827328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0074746308709003e-05,
+      "loss": 3.4308,
+      "step": 1827840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0066376740002216e-05,
+      "loss": 3.4191,
+      "step": 1828352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0057990792491696e-05,
+      "loss": 3.4299,
+      "step": 1828864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0049604844981176e-05,
+      "loss": 3.4216,
+      "step": 1829376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0041218897470655e-05,
+      "loss": 3.4246,
+      "step": 1829888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0032832949960135e-05,
+      "loss": 3.4295,
+      "step": 1830400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0024463381253348e-05,
+      "loss": 3.4221,
+      "step": 1830912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0016077433742828e-05,
+      "loss": 3.429,
+      "step": 1831424
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9174063205718994,
+      "eval_runtime": 301.9317,
+      "eval_samples_per_second": 1263.832,
+      "eval_steps_per_second": 39.496,
+      "step": 1831680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0007691486232304e-05,
+      "loss": 3.4261,
+      "step": 1831936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9999321917525517e-05,
+      "loss": 3.4197,
+      "step": 1832448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9990935970014997e-05,
+      "loss": 3.4284,
+      "step": 1832960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9982550022504477e-05,
+      "loss": 3.4268,
+      "step": 1833472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9974164074993957e-05,
+      "loss": 3.4375,
+      "step": 1833984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.996579450628717e-05,
+      "loss": 3.4192,
+      "step": 1834496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.995740855877665e-05,
+      "loss": 3.4157,
+      "step": 1835008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9949022611266126e-05,
+      "loss": 3.4193,
+      "step": 1835520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.994063666375561e-05,
+      "loss": 3.4284,
+      "step": 1836032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.993225071624509e-05,
+      "loss": 3.4278,
+      "step": 1836544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.992386476873457e-05,
+      "loss": 3.4191,
+      "step": 1837056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.991547882122405e-05,
+      "loss": 3.4258,
+      "step": 1837568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.990709287371353e-05,
+      "loss": 3.4196,
+      "step": 1838080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.989873968381047e-05,
+      "loss": 3.4142,
+      "step": 1838592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.989035373629995e-05,
+      "loss": 3.419,
+      "step": 1839104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.988196778878943e-05,
+      "loss": 3.4096,
+      "step": 1839616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.987358184127891e-05,
+      "loss": 3.4169,
+      "step": 1840128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9865212272572123e-05,
+      "loss": 3.4122,
+      "step": 1840640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.98568263250616e-05,
+      "loss": 3.4128,
+      "step": 1841152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.984844037755108e-05,
+      "loss": 3.4429,
+      "step": 1841664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.984005443004056e-05,
+      "loss": 3.4259,
+      "step": 1842176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9831668482530043e-05,
+      "loss": 3.4202,
+      "step": 1842688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9823282535019523e-05,
+      "loss": 3.4235,
+      "step": 1843200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9814896587509003e-05,
+      "loss": 3.4251,
+      "step": 1843712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9806527018802212e-05,
+      "loss": 3.4066,
+      "step": 1844224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9798141071291692e-05,
+      "loss": 3.4208,
+      "step": 1844736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.978975512378117e-05,
+      "loss": 3.4225,
+      "step": 1845248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.978136917627065e-05,
+      "loss": 3.4102,
+      "step": 1845760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.977298322876013e-05,
+      "loss": 3.409,
+      "step": 1846272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.976459728124961e-05,
+      "loss": 3.4118,
+      "step": 1846784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.975621133373909e-05,
+      "loss": 3.4213,
+      "step": 1847296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9747841765032304e-05,
+      "loss": 3.426,
+      "step": 1847808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.973945581752178e-05,
+      "loss": 3.4206,
+      "step": 1848320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9731069870011264e-05,
+      "loss": 3.4186,
+      "step": 1848832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9722683922500744e-05,
+      "loss": 3.4185,
+      "step": 1849344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9714297974990224e-05,
+      "loss": 3.4235,
+      "step": 1849856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9705912027479704e-05,
+      "loss": 3.4205,
+      "step": 1850368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9697526079969184e-05,
+      "loss": 3.4212,
+      "step": 1850880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9689156511262393e-05,
+      "loss": 3.4062,
+      "step": 1851392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9680770563751873e-05,
+      "loss": 3.418,
+      "step": 1851904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9672384616241356e-05,
+      "loss": 3.4115,
+      "step": 1852416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9663998668730836e-05,
+      "loss": 3.4316,
+      "step": 1852928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9655612721220313e-05,
+      "loss": 3.4262,
+      "step": 1853440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9647243152513525e-05,
+      "loss": 3.4168,
+      "step": 1853952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9638857205003005e-05,
+      "loss": 3.4251,
+      "step": 1854464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9630471257492485e-05,
+      "loss": 3.4171,
+      "step": 1854976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9622085309981965e-05,
+      "loss": 3.4244,
+      "step": 1855488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9613699362471445e-05,
+      "loss": 3.4144,
+      "step": 1856000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9605329793764657e-05,
+      "loss": 3.3903,
+      "step": 1856512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9596943846254137e-05,
+      "loss": 3.4323,
+      "step": 1857024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9588557898743614e-05,
+      "loss": 3.4139,
+      "step": 1857536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9580171951233094e-05,
+      "loss": 3.4298,
+      "step": 1858048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9571786003722574e-05,
+      "loss": 3.4144,
+      "step": 1858560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9563400056212057e-05,
+      "loss": 3.416,
+      "step": 1859072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9555030487505266e-05,
+      "loss": 3.4035,
+      "step": 1859584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9546644539994746e-05,
+      "loss": 3.4206,
+      "step": 1860096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9538258592484226e-05,
+      "loss": 3.4229,
+      "step": 1860608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9529872644973706e-05,
+      "loss": 3.4214,
+      "step": 1861120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9521486697463186e-05,
+      "loss": 3.4319,
+      "step": 1861632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9513100749952666e-05,
+      "loss": 3.4118,
+      "step": 1862144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9504714802442146e-05,
+      "loss": 3.4096,
+      "step": 1862656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.949634523373536e-05,
+      "loss": 3.4301,
+      "step": 1863168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.948795928622484e-05,
+      "loss": 3.4067,
+      "step": 1863680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9479573338714315e-05,
+      "loss": 3.4084,
+      "step": 1864192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9471187391203795e-05,
+      "loss": 3.4152,
+      "step": 1864704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9462801443693275e-05,
+      "loss": 3.423,
+      "step": 1865216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9454415496182758e-05,
+      "loss": 3.4046,
+      "step": 1865728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9446045927475967e-05,
+      "loss": 3.4123,
+      "step": 1866240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9437659979965447e-05,
+      "loss": 3.4039,
+      "step": 1866752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9429274032454927e-05,
+      "loss": 3.4126,
+      "step": 1867264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9420888084944407e-05,
+      "loss": 3.4256,
+      "step": 1867776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9412502137433887e-05,
+      "loss": 3.414,
+      "step": 1868288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9404116189923367e-05,
+      "loss": 3.4178,
+      "step": 1868800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.939574662121658e-05,
+      "loss": 3.4141,
+      "step": 1869312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.938736067370606e-05,
+      "loss": 3.4323,
+      "step": 1869824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.937897472619554e-05,
+      "loss": 3.4072,
+      "step": 1870336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.937058877868502e-05,
+      "loss": 3.426,
+      "step": 1870848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.936221920997823e-05,
+      "loss": 3.4173,
+      "step": 1871360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9353833262467712e-05,
+      "loss": 3.4118,
+      "step": 1871872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9345447314957192e-05,
+      "loss": 3.4265,
+      "step": 1872384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9337061367446672e-05,
+      "loss": 3.4217,
+      "step": 1872896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.932869179873988e-05,
+      "loss": 3.4203,
+      "step": 1873408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.932030585122936e-05,
+      "loss": 3.4199,
+      "step": 1873920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.931191990371884e-05,
+      "loss": 3.413,
+      "step": 1874432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.930353395620832e-05,
+      "loss": 3.4116,
+      "step": 1874944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.92951480086978e-05,
+      "loss": 3.4164,
+      "step": 1875456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.928676206118728e-05,
+      "loss": 3.4218,
+      "step": 1875968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.927837611367676e-05,
+      "loss": 3.4087,
+      "step": 1876480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.926999016616624e-05,
+      "loss": 3.4127,
+      "step": 1876992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.926162059745945e-05,
+      "loss": 3.4108,
+      "step": 1877504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9253251028752666e-05,
+      "loss": 3.4225,
+      "step": 1878016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9244865081242146e-05,
+      "loss": 3.4184,
+      "step": 1878528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9236479133731622e-05,
+      "loss": 3.4068,
+      "step": 1879040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9228093186221102e-05,
+      "loss": 3.4222,
+      "step": 1879552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9219707238710582e-05,
+      "loss": 3.4242,
+      "step": 1880064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9211321291200062e-05,
+      "loss": 3.4178,
+      "step": 1880576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9202951722493274e-05,
+      "loss": 3.4016,
+      "step": 1881088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9194565774982754e-05,
+      "loss": 3.4135,
+      "step": 1881600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9186179827472234e-05,
+      "loss": 3.4095,
+      "step": 1882112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9177793879961714e-05,
+      "loss": 3.4231,
+      "step": 1882624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9169407932451194e-05,
+      "loss": 3.4197,
+      "step": 1883136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9161021984940674e-05,
+      "loss": 3.4171,
+      "step": 1883648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.915263603743015e-05,
+      "loss": 3.4127,
+      "step": 1884160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9144250089919634e-05,
+      "loss": 3.4127,
+      "step": 1884672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9135880521212847e-05,
+      "loss": 3.4116,
+      "step": 1885184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9127494573702327e-05,
+      "loss": 3.4138,
+      "step": 1885696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9119108626191803e-05,
+      "loss": 3.4062,
+      "step": 1886208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9110722678681283e-05,
+      "loss": 3.4249,
+      "step": 1886720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9102353109974496e-05,
+      "loss": 3.4002,
+      "step": 1887232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9093967162463976e-05,
+      "loss": 3.4013,
+      "step": 1887744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9085581214953456e-05,
+      "loss": 3.4093,
+      "step": 1888256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9077195267442935e-05,
+      "loss": 3.4116,
+      "step": 1888768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9068825698736148e-05,
+      "loss": 3.4129,
+      "step": 1889280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9060439751225625e-05,
+      "loss": 3.4093,
+      "step": 1889792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9052053803715105e-05,
+      "loss": 3.4158,
+      "step": 1890304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.904368423500832e-05,
+      "loss": 3.4265,
+      "step": 1890816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.90352982874978e-05,
+      "loss": 3.4045,
+      "step": 1891328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.902692871879101e-05,
+      "loss": 3.4168,
+      "step": 1891840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.901854277128049e-05,
+      "loss": 3.4139,
+      "step": 1892352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.901015682376997e-05,
+      "loss": 3.4154,
+      "step": 1892864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.900177087625945e-05,
+      "loss": 3.4084,
+      "step": 1893376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8993401307552662e-05,
+      "loss": 3.4082,
+      "step": 1893888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8985015360042142e-05,
+      "loss": 3.4083,
+      "step": 1894400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8976629412531622e-05,
+      "loss": 3.4152,
+      "step": 1894912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8968243465021098e-05,
+      "loss": 3.4181,
+      "step": 1895424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.895987389631431e-05,
+      "loss": 3.4119,
+      "step": 1895936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.895148794880379e-05,
+      "loss": 3.4127,
+      "step": 1896448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8943102001293274e-05,
+      "loss": 3.4087,
+      "step": 1896960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.893471605378275e-05,
+      "loss": 3.3976,
+      "step": 1897472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.892633010627223e-05,
+      "loss": 3.4258,
+      "step": 1897984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8917960537565443e-05,
+      "loss": 3.4063,
+      "step": 1898496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.890957459005492e-05,
+      "loss": 3.4212,
+      "step": 1899008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.89011886425444e-05,
+      "loss": 3.4028,
+      "step": 1899520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8892802695033883e-05,
+      "loss": 3.4245,
+      "step": 1900032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8884416747523363e-05,
+      "loss": 3.4127,
+      "step": 1900544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8876030800012843e-05,
+      "loss": 3.408,
+      "step": 1901056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8867644852502323e-05,
+      "loss": 3.4117,
+      "step": 1901568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8859275283795532e-05,
+      "loss": 3.4113,
+      "step": 1902080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8850889336285012e-05,
+      "loss": 3.4178,
+      "step": 1902592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8842503388774492e-05,
+      "loss": 3.4325,
+      "step": 1903104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8834117441263975e-05,
+      "loss": 3.3976,
+      "step": 1903616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8825747872557184e-05,
+      "loss": 3.4246,
+      "step": 1904128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8817361925046664e-05,
+      "loss": 3.4068,
+      "step": 1904640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8808975977536144e-05,
+      "loss": 3.4175,
+      "step": 1905152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8800590030025624e-05,
+      "loss": 3.4127,
+      "step": 1905664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8792204082515104e-05,
+      "loss": 3.411,
+      "step": 1906176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8783818135004584e-05,
+      "loss": 3.4182,
+      "step": 1906688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8775448566297796e-05,
+      "loss": 3.4147,
+      "step": 1907200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8767062618787276e-05,
+      "loss": 3.414,
+      "step": 1907712
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9203341007232666,
+      "eval_runtime": 312.041,
+      "eval_samples_per_second": 1222.887,
+      "eval_steps_per_second": 38.216,
+      "step": 1908000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8758676671276753e-05,
+      "loss": 3.4115,
+      "step": 1908224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8750290723766233e-05,
+      "loss": 3.4125,
+      "step": 1908736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8741921155059445e-05,
+      "loss": 3.4171,
+      "step": 1909248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.873353520754893e-05,
+      "loss": 3.4151,
+      "step": 1909760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8725149260038405e-05,
+      "loss": 3.4254,
+      "step": 1910272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8716763312527885e-05,
+      "loss": 3.4107,
+      "step": 1910784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8708377365017365e-05,
+      "loss": 3.4062,
+      "step": 1911296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8699991417506845e-05,
+      "loss": 3.4097,
+      "step": 1911808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8691605469996325e-05,
+      "loss": 3.4142,
+      "step": 1912320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8683219522485805e-05,
+      "loss": 3.4182,
+      "step": 1912832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8674849953779018e-05,
+      "loss": 3.4102,
+      "step": 1913344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8666464006268498e-05,
+      "loss": 3.4109,
+      "step": 1913856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8658078058757978e-05,
+      "loss": 3.4104,
+      "step": 1914368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8649692111247458e-05,
+      "loss": 3.4013,
+      "step": 1914880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8641322542540667e-05,
+      "loss": 3.4085,
+      "step": 1915392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8632936595030147e-05,
+      "loss": 3.3978,
+      "step": 1915904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.862455064751963e-05,
+      "loss": 3.402,
+      "step": 1916416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.861616470000911e-05,
+      "loss": 3.4038,
+      "step": 1916928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8607778752498586e-05,
+      "loss": 3.4039,
+      "step": 1917440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8599392804988066e-05,
+      "loss": 3.4273,
+      "step": 1917952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8591006857477546e-05,
+      "loss": 3.4165,
+      "step": 1918464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.858263728877076e-05,
+      "loss": 3.4099,
+      "step": 1918976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.857425134126024e-05,
+      "loss": 3.4121,
+      "step": 1919488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.856586539374972e-05,
+      "loss": 3.413,
+      "step": 1920000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.85574794462392e-05,
+      "loss": 3.3961,
+      "step": 1920512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8549109877532408e-05,
+      "loss": 3.4126,
+      "step": 1921024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8540723930021888e-05,
+      "loss": 3.4139,
+      "step": 1921536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8532337982511368e-05,
+      "loss": 3.3941,
+      "step": 1922048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.852395203500085e-05,
+      "loss": 3.4052,
+      "step": 1922560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.851556608749033e-05,
+      "loss": 3.3989,
+      "step": 1923072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.850718013997981e-05,
+      "loss": 3.4098,
+      "step": 1923584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.849881057127302e-05,
+      "loss": 3.41,
+      "step": 1924096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.84904246237625e-05,
+      "loss": 3.4123,
+      "step": 1924608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.848203867625198e-05,
+      "loss": 3.4148,
+      "step": 1925120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.847365272874146e-05,
+      "loss": 3.4014,
+      "step": 1925632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.846526678123094e-05,
+      "loss": 3.4139,
+      "step": 1926144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.845688083372042e-05,
+      "loss": 3.4142,
+      "step": 1926656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.84484948862099e-05,
+      "loss": 3.4069,
+      "step": 1927168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8440125317503112e-05,
+      "loss": 3.3979,
+      "step": 1927680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.843173936999259e-05,
+      "loss": 3.4065,
+      "step": 1928192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.842335342248207e-05,
+      "loss": 3.4005,
+      "step": 1928704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8414967474971552e-05,
+      "loss": 3.4194,
+      "step": 1929216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8406581527461032e-05,
+      "loss": 3.4138,
+      "step": 1929728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8398195579950512e-05,
+      "loss": 3.4032,
+      "step": 1930240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.838982601124372e-05,
+      "loss": 3.4158,
+      "step": 1930752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.83814400637332e-05,
+      "loss": 3.4048,
+      "step": 1931264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.837305411622268e-05,
+      "loss": 3.4119,
+      "step": 1931776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.836466816871216e-05,
+      "loss": 3.4105,
+      "step": 1932288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8356298600005373e-05,
+      "loss": 3.3758,
+      "step": 1932800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8347912652494853e-05,
+      "loss": 3.4217,
+      "step": 1933312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8339526704984333e-05,
+      "loss": 3.4075,
+      "step": 1933824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8331140757473813e-05,
+      "loss": 3.4157,
+      "step": 1934336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8322754809963293e-05,
+      "loss": 3.4042,
+      "step": 1934848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8314385241256506e-05,
+      "loss": 3.4056,
+      "step": 1935360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8305999293745986e-05,
+      "loss": 3.3936,
+      "step": 1935872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8297613346235466e-05,
+      "loss": 3.4099,
+      "step": 1936384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8289227398724942e-05,
+      "loss": 3.4115,
+      "step": 1936896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8280857830018155e-05,
+      "loss": 3.4107,
+      "step": 1937408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8272471882507635e-05,
+      "loss": 3.4255,
+      "step": 1937920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8264085934997115e-05,
+      "loss": 3.4033,
+      "step": 1938432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8255699987486595e-05,
+      "loss": 3.3951,
+      "step": 1938944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8247330418779807e-05,
+      "loss": 3.4218,
+      "step": 1939456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8238944471269287e-05,
+      "loss": 3.3979,
+      "step": 1939968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8230558523758767e-05,
+      "loss": 3.3921,
+      "step": 1940480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8222172576248244e-05,
+      "loss": 3.4103,
+      "step": 1940992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8213786628737724e-05,
+      "loss": 3.4095,
+      "step": 1941504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8205400681227207e-05,
+      "loss": 3.3956,
+      "step": 1942016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8197014733716687e-05,
+      "loss": 3.4001,
+      "step": 1942528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8188645165009896e-05,
+      "loss": 3.3926,
+      "step": 1943040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8180259217499376e-05,
+      "loss": 3.4024,
+      "step": 1943552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8171873269988856e-05,
+      "loss": 3.4158,
+      "step": 1944064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8163487322478336e-05,
+      "loss": 3.4032,
+      "step": 1944576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8155101374967816e-05,
+      "loss": 3.4055,
+      "step": 1945088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8146731806261028e-05,
+      "loss": 3.4012,
+      "step": 1945600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8138345858750508e-05,
+      "loss": 3.4225,
+      "step": 1946112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8129959911239988e-05,
+      "loss": 3.3982,
+      "step": 1946624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8121573963729468e-05,
+      "loss": 3.4182,
+      "step": 1947136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8113204395022677e-05,
+      "loss": 3.4032,
+      "step": 1947648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.810481844751216e-05,
+      "loss": 3.4041,
+      "step": 1948160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.809643250000164e-05,
+      "loss": 3.4125,
+      "step": 1948672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.808804655249112e-05,
+      "loss": 3.4152,
+      "step": 1949184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.807967698378433e-05,
+      "loss": 3.4066,
+      "step": 1949696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.807129103627381e-05,
+      "loss": 3.4088,
+      "step": 1950208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.806290508876329e-05,
+      "loss": 3.3986,
+      "step": 1950720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.805451914125277e-05,
+      "loss": 3.3994,
+      "step": 1951232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8046149572545982e-05,
+      "loss": 3.4043,
+      "step": 1951744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8037763625035462e-05,
+      "loss": 3.409,
+      "step": 1952256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8029377677524942e-05,
+      "loss": 3.3998,
+      "step": 1952768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8020991730014422e-05,
+      "loss": 3.4021,
+      "step": 1953280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.80126057825039e-05,
+      "loss": 3.404,
+      "step": 1953792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8004219834993378e-05,
+      "loss": 3.4058,
+      "step": 1954304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7995850266286594e-05,
+      "loss": 3.4103,
+      "step": 1954816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.798746431877607e-05,
+      "loss": 3.3952,
+      "step": 1955328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.797907837126555e-05,
+      "loss": 3.4123,
+      "step": 1955840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.797069242375503e-05,
+      "loss": 3.4087,
+      "step": 1956352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.796230647624451e-05,
+      "loss": 3.4122,
+      "step": 1956864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7953936907537723e-05,
+      "loss": 3.3886,
+      "step": 1957376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7945550960027203e-05,
+      "loss": 3.4004,
+      "step": 1957888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7937165012516683e-05,
+      "loss": 3.4006,
+      "step": 1958400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7928779065006163e-05,
+      "loss": 3.4106,
+      "step": 1958912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7920393117495643e-05,
+      "loss": 3.4069,
+      "step": 1959424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7912007169985123e-05,
+      "loss": 3.4081,
+      "step": 1959936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7903621222474603e-05,
+      "loss": 3.4035,
+      "step": 1960448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7895235274964083e-05,
+      "loss": 3.4,
+      "step": 1960960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7886865706257295e-05,
+      "loss": 3.3989,
+      "step": 1961472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7878479758746775e-05,
+      "loss": 3.4066,
+      "step": 1961984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7870093811236252e-05,
+      "loss": 3.3976,
+      "step": 1962496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7861707863725732e-05,
+      "loss": 3.4069,
+      "step": 1963008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7853338295018944e-05,
+      "loss": 3.3969,
+      "step": 1963520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7844952347508424e-05,
+      "loss": 3.3903,
+      "step": 1964032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7836566399997904e-05,
+      "loss": 3.3994,
+      "step": 1964544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7828180452487384e-05,
+      "loss": 3.4015,
+      "step": 1965056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7819810883780597e-05,
+      "loss": 3.4015,
+      "step": 1965568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7811424936270077e-05,
+      "loss": 3.3997,
+      "step": 1966080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7803038988759553e-05,
+      "loss": 3.4075,
+      "step": 1966592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7794653041249036e-05,
+      "loss": 3.4137,
+      "step": 1967104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.778628347254225e-05,
+      "loss": 3.3957,
+      "step": 1967616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7777913903835458e-05,
+      "loss": 3.4052,
+      "step": 1968128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7769527956324938e-05,
+      "loss": 3.4072,
+      "step": 1968640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7761142008814418e-05,
+      "loss": 3.4019,
+      "step": 1969152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7752756061303898e-05,
+      "loss": 3.3994,
+      "step": 1969664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.774438649259711e-05,
+      "loss": 3.3948,
+      "step": 1970176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.773600054508659e-05,
+      "loss": 3.4018,
+      "step": 1970688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.772761459757607e-05,
+      "loss": 3.4024,
+      "step": 1971200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7719228650065547e-05,
+      "loss": 3.4049,
+      "step": 1971712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.771085908135876e-05,
+      "loss": 3.4024,
+      "step": 1972224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.770247313384824e-05,
+      "loss": 3.3977,
+      "step": 1972736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7694087186337723e-05,
+      "loss": 3.4006,
+      "step": 1973248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.76857012388272e-05,
+      "loss": 3.3845,
+      "step": 1973760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.767731529131668e-05,
+      "loss": 3.4138,
+      "step": 1974272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.766894572260989e-05,
+      "loss": 3.3971,
+      "step": 1974784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.766055977509937e-05,
+      "loss": 3.4127,
+      "step": 1975296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7652173827588848e-05,
+      "loss": 3.3928,
+      "step": 1975808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.764378788007833e-05,
+      "loss": 3.4139,
+      "step": 1976320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.763540193256781e-05,
+      "loss": 3.4073,
+      "step": 1976832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.762701598505729e-05,
+      "loss": 3.3948,
+      "step": 1977344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.76186464163505e-05,
+      "loss": 3.402,
+      "step": 1977856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.761026046883998e-05,
+      "loss": 3.3975,
+      "step": 1978368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.760187452132946e-05,
+      "loss": 3.4114,
+      "step": 1978880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.759348857381894e-05,
+      "loss": 3.4173,
+      "step": 1979392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7585102626308424e-05,
+      "loss": 3.3902,
+      "step": 1979904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7576733057601633e-05,
+      "loss": 3.4076,
+      "step": 1980416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7568347110091113e-05,
+      "loss": 3.3997,
+      "step": 1980928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7559961162580593e-05,
+      "loss": 3.4061,
+      "step": 1981440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7551575215070073e-05,
+      "loss": 3.4015,
+      "step": 1981952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7543189267559553e-05,
+      "loss": 3.4076,
+      "step": 1982464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7534803320049033e-05,
+      "loss": 3.4035,
+      "step": 1982976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7526433751342245e-05,
+      "loss": 3.4019,
+      "step": 1983488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7518047803831725e-05,
+      "loss": 3.4064,
+      "step": 1984000
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.921309232711792,
+      "eval_runtime": 311.1808,
+      "eval_samples_per_second": 1226.268,
+      "eval_steps_per_second": 38.322,
+      "step": 1984320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.75096618563212e-05,
+      "loss": 3.3962,
+      "step": 1984512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.750127590881068e-05,
+      "loss": 3.3995,
+      "step": 1985024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.749288996130016e-05,
+      "loss": 3.4065,
+      "step": 1985536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7484520392593377e-05,
+      "loss": 3.4077,
+      "step": 1986048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7476134445082854e-05,
+      "loss": 3.4146,
+      "step": 1986560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7467748497572334e-05,
+      "loss": 3.3996,
+      "step": 1987072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7459362550061814e-05,
+      "loss": 3.3954,
+      "step": 1987584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7450976602551294e-05,
+      "loss": 3.3986,
+      "step": 1988096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7442607033844506e-05,
+      "loss": 3.4011,
+      "step": 1988608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7434221086333986e-05,
+      "loss": 3.4115,
+      "step": 1989120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7425835138823466e-05,
+      "loss": 3.3981,
+      "step": 1989632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7417449191312946e-05,
+      "loss": 3.3998,
+      "step": 1990144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7409063243802426e-05,
+      "loss": 3.4017,
+      "step": 1990656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7400677296291906e-05,
+      "loss": 3.3908,
+      "step": 1991168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7392307727585115e-05,
+      "loss": 3.4,
+      "step": 1991680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7383921780074595e-05,
+      "loss": 3.3872,
+      "step": 1992192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.737553583256408e-05,
+      "loss": 3.3908,
+      "step": 1992704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.736714988505356e-05,
+      "loss": 3.3961,
+      "step": 1993216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7358763937543035e-05,
+      "loss": 3.39,
+      "step": 1993728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7350394368836248e-05,
+      "loss": 3.4135,
+      "step": 1994240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7342008421325727e-05,
+      "loss": 3.4093,
+      "step": 1994752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7333622473815207e-05,
+      "loss": 3.3997,
+      "step": 1995264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7325236526304687e-05,
+      "loss": 3.402,
+      "step": 1995776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7316850578794167e-05,
+      "loss": 3.4004,
+      "step": 1996288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.730848101008738e-05,
+      "loss": 3.3898,
+      "step": 1996800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7300095062576856e-05,
+      "loss": 3.4027,
+      "step": 1997312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7291709115066336e-05,
+      "loss": 3.3997,
+      "step": 1997824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7283323167555816e-05,
+      "loss": 3.3851,
+      "step": 1998336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.72749372200453e-05,
+      "loss": 3.3919,
+      "step": 1998848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.726655127253478e-05,
+      "loss": 3.3914,
+      "step": 1999360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.725816532502426e-05,
+      "loss": 3.4039,
+      "step": 1999872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.724977937751374e-05,
+      "loss": 3.398,
+      "step": 2000384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.724140980880695e-05,
+      "loss": 3.4001,
+      "step": 2000896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.723302386129643e-05,
+      "loss": 3.402,
+      "step": 2001408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.722463791378591e-05,
+      "loss": 3.3887,
+      "step": 2001920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7216251966275392e-05,
+      "loss": 3.4042,
+      "step": 2002432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.720786601876487e-05,
+      "loss": 3.4064,
+      "step": 2002944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.719949645005808e-05,
+      "loss": 3.3917,
+      "step": 2003456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.719111050254756e-05,
+      "loss": 3.393,
+      "step": 2003968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7182724555037037e-05,
+      "loss": 3.3935,
+      "step": 2004480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7174338607526517e-05,
+      "loss": 3.3929,
+      "step": 2004992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7165952660016e-05,
+      "loss": 3.4078,
+      "step": 2005504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7157583091309213e-05,
+      "loss": 3.4028,
+      "step": 2006016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.714919714379869e-05,
+      "loss": 3.3938,
+      "step": 2006528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.714081119628817e-05,
+      "loss": 3.4039,
+      "step": 2007040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.713242524877765e-05,
+      "loss": 3.3968,
+      "step": 2007552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.712403930126713e-05,
+      "loss": 3.3986,
+      "step": 2008064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.711565335375661e-05,
+      "loss": 3.3994,
+      "step": 2008576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7107283785049822e-05,
+      "loss": 3.3703,
+      "step": 2009088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7098897837539302e-05,
+      "loss": 3.4076,
+      "step": 2009600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7090511890028782e-05,
+      "loss": 3.3914,
+      "step": 2010112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7082125942518262e-05,
+      "loss": 3.4036,
+      "step": 2010624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7073739995007742e-05,
+      "loss": 3.3976,
+      "step": 2011136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.706535404749722e-05,
+      "loss": 3.3946,
+      "step": 2011648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7056984478790434e-05,
+      "loss": 3.3848,
+      "step": 2012160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7048598531279914e-05,
+      "loss": 3.3998,
+      "step": 2012672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7040212583769394e-05,
+      "loss": 3.4004,
+      "step": 2013184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.703182663625887e-05,
+      "loss": 3.398,
+      "step": 2013696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.702344068874835e-05,
+      "loss": 3.4136,
+      "step": 2014208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.701505474123783e-05,
+      "loss": 3.3983,
+      "step": 2014720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.700666879372731e-05,
+      "loss": 3.3794,
+      "step": 2015232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6998299225020523e-05,
+      "loss": 3.4088,
+      "step": 2015744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6989913277510003e-05,
+      "loss": 3.3878,
+      "step": 2016256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6981527329999483e-05,
+      "loss": 3.3856,
+      "step": 2016768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6973141382488963e-05,
+      "loss": 3.399,
+      "step": 2017280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6964755434978443e-05,
+      "loss": 3.3974,
+      "step": 2017792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6956385866271655e-05,
+      "loss": 3.3849,
+      "step": 2018304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6947999918761135e-05,
+      "loss": 3.3899,
+      "step": 2018816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6939613971250615e-05,
+      "loss": 3.3842,
+      "step": 2019328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6931228023740095e-05,
+      "loss": 3.3902,
+      "step": 2019840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6922842076229575e-05,
+      "loss": 3.4042,
+      "step": 2020352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6914456128719052e-05,
+      "loss": 3.3955,
+      "step": 2020864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6906070181208532e-05,
+      "loss": 3.3955,
+      "step": 2021376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6897700612501748e-05,
+      "loss": 3.3897,
+      "step": 2021888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6889314664991224e-05,
+      "loss": 3.4076,
+      "step": 2022400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6880928717480704e-05,
+      "loss": 3.3899,
+      "step": 2022912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6872542769970184e-05,
+      "loss": 3.4079,
+      "step": 2023424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6864173201263397e-05,
+      "loss": 3.3905,
+      "step": 2023936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6855787253752873e-05,
+      "loss": 3.3969,
+      "step": 2024448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6847401306242357e-05,
+      "loss": 3.4027,
+      "step": 2024960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6839015358731837e-05,
+      "loss": 3.4063,
+      "step": 2025472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.683064579002505e-05,
+      "loss": 3.3937,
+      "step": 2025984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6822259842514526e-05,
+      "loss": 3.4022,
+      "step": 2026496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6813873895004006e-05,
+      "loss": 3.3854,
+      "step": 2027008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6805487947493486e-05,
+      "loss": 3.3908,
+      "step": 2027520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6797101999982965e-05,
+      "loss": 3.3948,
+      "step": 2028032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6788732431276178e-05,
+      "loss": 3.4001,
+      "step": 2028544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6780346483765658e-05,
+      "loss": 3.3876,
+      "step": 2029056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6771960536255138e-05,
+      "loss": 3.3902,
+      "step": 2029568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6763574588744618e-05,
+      "loss": 3.3939,
+      "step": 2030080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6755188641234098e-05,
+      "loss": 3.392,
+      "step": 2030592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.674681907252731e-05,
+      "loss": 3.4017,
+      "step": 2031104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.673843312501679e-05,
+      "loss": 3.3831,
+      "step": 2031616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.673004717750627e-05,
+      "loss": 3.3978,
+      "step": 2032128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.672166122999575e-05,
+      "loss": 3.402,
+      "step": 2032640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.671327528248523e-05,
+      "loss": 3.4036,
+      "step": 2033152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.670490571377844e-05,
+      "loss": 3.3768,
+      "step": 2033664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.669651976626792e-05,
+      "loss": 3.3919,
+      "step": 2034176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6688133818757402e-05,
+      "loss": 3.3929,
+      "step": 2034688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.667974787124688e-05,
+      "loss": 3.3978,
+      "step": 2035200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.667136192373636e-05,
+      "loss": 3.3985,
+      "step": 2035712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.666297597622584e-05,
+      "loss": 3.394,
+      "step": 2036224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.665459002871532e-05,
+      "loss": 3.39,
+      "step": 2036736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.66462040812048e-05,
+      "loss": 3.3955,
+      "step": 2037248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.663783451249801e-05,
+      "loss": 3.3885,
+      "step": 2037760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.662944856498749e-05,
+      "loss": 3.3965,
+      "step": 2038272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.662106261747697e-05,
+      "loss": 3.3881,
+      "step": 2038784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.661267666996645e-05,
+      "loss": 3.3975,
+      "step": 2039296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.660430710125966e-05,
+      "loss": 3.3865,
+      "step": 2039808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6595937532552873e-05,
+      "loss": 3.3797,
+      "step": 2040320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6587551585042353e-05,
+      "loss": 3.3857,
+      "step": 2040832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6579165637531833e-05,
+      "loss": 3.3908,
+      "step": 2041344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6570779690021313e-05,
+      "loss": 3.3943,
+      "step": 2041856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6562393742510793e-05,
+      "loss": 3.3885,
+      "step": 2042368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6554007795000273e-05,
+      "loss": 3.3943,
+      "step": 2042880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6545638226293485e-05,
+      "loss": 3.4008,
+      "step": 2043392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6537252278782965e-05,
+      "loss": 3.3907,
+      "step": 2043904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6528866331272445e-05,
+      "loss": 3.397,
+      "step": 2044416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6520480383761925e-05,
+      "loss": 3.3944,
+      "step": 2044928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6512110815055134e-05,
+      "loss": 3.3914,
+      "step": 2045440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6503724867544614e-05,
+      "loss": 3.3878,
+      "step": 2045952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6495338920034094e-05,
+      "loss": 3.3838,
+      "step": 2046464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6486952972523574e-05,
+      "loss": 3.3943,
+      "step": 2046976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6478583403816786e-05,
+      "loss": 3.39,
+      "step": 2047488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6470197456306266e-05,
+      "loss": 3.3969,
+      "step": 2048000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6461811508795746e-05,
+      "loss": 3.3962,
+      "step": 2048512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6453425561285226e-05,
+      "loss": 3.3867,
+      "step": 2049024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6445039613774706e-05,
+      "loss": 3.387,
+      "step": 2049536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.643667004506792e-05,
+      "loss": 3.376,
+      "step": 2050048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.64282840975574e-05,
+      "loss": 3.4017,
+      "step": 2050560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.641989815004688e-05,
+      "loss": 3.3855,
+      "step": 2051072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6411512202536355e-05,
+      "loss": 3.4022,
+      "step": 2051584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6403142633829568e-05,
+      "loss": 3.3855,
+      "step": 2052096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6394756686319048e-05,
+      "loss": 3.4015,
+      "step": 2052608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6386370738808528e-05,
+      "loss": 3.3943,
+      "step": 2053120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6377984791298008e-05,
+      "loss": 3.381,
+      "step": 2053632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.636961522259122e-05,
+      "loss": 3.3953,
+      "step": 2054144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.63612292750807e-05,
+      "loss": 3.3854,
+      "step": 2054656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.635284332757018e-05,
+      "loss": 3.4028,
+      "step": 2055168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6344457380059656e-05,
+      "loss": 3.402,
+      "step": 2055680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.633607143254914e-05,
+      "loss": 3.3858,
+      "step": 2056192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6327701863842352e-05,
+      "loss": 3.3923,
+      "step": 2056704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.631931591633183e-05,
+      "loss": 3.3891,
+      "step": 2057216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.631092996882131e-05,
+      "loss": 3.4006,
+      "step": 2057728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.630254402131079e-05,
+      "loss": 3.386,
+      "step": 2058240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.629415807380027e-05,
+      "loss": 3.4007,
+      "step": 2058752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.628577212628975e-05,
+      "loss": 3.3887,
+      "step": 2059264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.627740255758296e-05,
+      "loss": 3.3973,
+      "step": 2059776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.626901661007244e-05,
+      "loss": 3.3913,
+      "step": 2060288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9234910011291504,
+      "eval_runtime": 304.4771,
+      "eval_samples_per_second": 1253.267,
+      "eval_steps_per_second": 39.166,
+      "step": 2060640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.626063066256192e-05,
+      "loss": 3.3885,
+      "step": 2060800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.62522447150514e-05,
+      "loss": 3.3873,
+      "step": 2061312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.624385876754088e-05,
+      "loss": 3.3933,
+      "step": 2061824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6235489198834093e-05,
+      "loss": 3.3979,
+      "step": 2062336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6227103251323573e-05,
+      "loss": 3.4072,
+      "step": 2062848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6218717303813053e-05,
+      "loss": 3.388,
+      "step": 2063360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6210331356302533e-05,
+      "loss": 3.3875,
+      "step": 2063872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.620194540879201e-05,
+      "loss": 3.3907,
+      "step": 2064384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.619355946128149e-05,
+      "loss": 3.3938,
+      "step": 2064896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6185189892574702e-05,
+      "loss": 3.3957,
+      "step": 2065408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6176803945064182e-05,
+      "loss": 3.385,
+      "step": 2065920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6168417997553662e-05,
+      "loss": 3.3906,
+      "step": 2066432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6160032050043142e-05,
+      "loss": 3.3945,
+      "step": 2066944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6151646102532622e-05,
+      "loss": 3.3824,
+      "step": 2067456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6143276533825835e-05,
+      "loss": 3.3874,
+      "step": 2067968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.613489058631531e-05,
+      "loss": 3.377,
+      "step": 2068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6126504638804795e-05,
+      "loss": 3.3755,
+      "step": 2068992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6118118691294275e-05,
+      "loss": 3.3876,
+      "step": 2069504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6109732743783754e-05,
+      "loss": 3.3845,
+      "step": 2070016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6101363175076964e-05,
+      "loss": 3.4015,
+      "step": 2070528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6092977227566444e-05,
+      "loss": 3.3976,
+      "step": 2071040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6084591280055923e-05,
+      "loss": 3.3903,
+      "step": 2071552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6076205332545403e-05,
+      "loss": 3.3934,
+      "step": 2072064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6067819385034887e-05,
+      "loss": 3.3897,
+      "step": 2072576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6059449816328096e-05,
+      "loss": 3.3763,
+      "step": 2073088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6051063868817576e-05,
+      "loss": 3.3973,
+      "step": 2073600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6042677921307056e-05,
+      "loss": 3.3875,
+      "step": 2074112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6034291973796536e-05,
+      "loss": 3.3761,
+      "step": 2074624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6025906026286016e-05,
+      "loss": 3.3838,
+      "step": 2075136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6017520078775496e-05,
+      "loss": 3.3822,
+      "step": 2075648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6009134131264976e-05,
+      "loss": 3.3895,
+      "step": 2076160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6000764562558188e-05,
+      "loss": 3.3862,
+      "step": 2076672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5992378615047665e-05,
+      "loss": 3.3968,
+      "step": 2077184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5983992667537145e-05,
+      "loss": 3.3882,
+      "step": 2077696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5975606720026625e-05,
+      "loss": 3.3783,
+      "step": 2078208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5967220772516105e-05,
+      "loss": 3.3944,
+      "step": 2078720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5958834825005588e-05,
+      "loss": 3.3954,
+      "step": 2079232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5950465256298797e-05,
+      "loss": 3.3821,
+      "step": 2079744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5942079308788277e-05,
+      "loss": 3.3821,
+      "step": 2080256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5933693361277757e-05,
+      "loss": 3.382,
+      "step": 2080768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5925307413767237e-05,
+      "loss": 3.3798,
+      "step": 2081280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5916921466256717e-05,
+      "loss": 3.3955,
+      "step": 2081792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5908535518746197e-05,
+      "loss": 3.3926,
+      "step": 2082304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5900149571235677e-05,
+      "loss": 3.3899,
+      "step": 2082816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.589178000252889e-05,
+      "loss": 3.3884,
+      "step": 2083328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.588339405501837e-05,
+      "loss": 3.3896,
+      "step": 2083840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5875008107507846e-05,
+      "loss": 3.3895,
+      "step": 2084352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5866622159997326e-05,
+      "loss": 3.3878,
+      "step": 2084864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.585825259129054e-05,
+      "loss": 3.365,
+      "step": 2085376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.584986664378002e-05,
+      "loss": 3.3968,
+      "step": 2085888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5841480696269498e-05,
+      "loss": 3.3812,
+      "step": 2086400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5833094748758978e-05,
+      "loss": 3.3952,
+      "step": 2086912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5824708801248458e-05,
+      "loss": 3.3878,
+      "step": 2087424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5816339232541667e-05,
+      "loss": 3.3835,
+      "step": 2087936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.580795328503115e-05,
+      "loss": 3.374,
+      "step": 2088448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.579956733752063e-05,
+      "loss": 3.3862,
+      "step": 2088960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.579118139001011e-05,
+      "loss": 3.3953,
+      "step": 2089472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.578281182130332e-05,
+      "loss": 3.3842,
+      "step": 2089984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.57744258737928e-05,
+      "loss": 3.403,
+      "step": 2090496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.576603992628228e-05,
+      "loss": 3.3854,
+      "step": 2091008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.575765397877176e-05,
+      "loss": 3.3713,
+      "step": 2091520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5749268031261243e-05,
+      "loss": 3.3971,
+      "step": 2092032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5740898462554452e-05,
+      "loss": 3.3839,
+      "step": 2092544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5732512515043932e-05,
+      "loss": 3.3769,
+      "step": 2093056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.572412656753341e-05,
+      "loss": 3.3851,
+      "step": 2093568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.571574062002289e-05,
+      "loss": 3.3891,
+      "step": 2094080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5707371051316104e-05,
+      "loss": 3.3719,
+      "step": 2094592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5698985103805584e-05,
+      "loss": 3.3793,
+      "step": 2095104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5690599156295064e-05,
+      "loss": 3.3717,
+      "step": 2095616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5682213208784544e-05,
+      "loss": 3.3843,
+      "step": 2096128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5673843640077753e-05,
+      "loss": 3.3896,
+      "step": 2096640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5665457692567233e-05,
+      "loss": 3.3827,
+      "step": 2097152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5657071745056713e-05,
+      "loss": 3.3871,
+      "step": 2097664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5648685797546196e-05,
+      "loss": 3.379,
+      "step": 2098176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5640316228839405e-05,
+      "loss": 3.3946,
+      "step": 2098688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5631930281328885e-05,
+      "loss": 3.3837,
+      "step": 2099200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5623544333818365e-05,
+      "loss": 3.399,
+      "step": 2099712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5615158386307845e-05,
+      "loss": 3.3811,
+      "step": 2100224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5606788817601058e-05,
+      "loss": 3.3841,
+      "step": 2100736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5598402870090538e-05,
+      "loss": 3.3911,
+      "step": 2101248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5590016922580018e-05,
+      "loss": 3.3934,
+      "step": 2101760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5581630975069498e-05,
+      "loss": 3.3863,
+      "step": 2102272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5573245027558974e-05,
+      "loss": 3.391,
+      "step": 2102784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5564875458852187e-05,
+      "loss": 3.3761,
+      "step": 2103296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5556489511341667e-05,
+      "loss": 3.3806,
+      "step": 2103808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.554810356383115e-05,
+      "loss": 3.3813,
+      "step": 2104320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5539717616320627e-05,
+      "loss": 3.3922,
+      "step": 2104832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5531331668810106e-05,
+      "loss": 3.3782,
+      "step": 2105344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5522945721299586e-05,
+      "loss": 3.3801,
+      "step": 2105856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5514559773789066e-05,
+      "loss": 3.3846,
+      "step": 2106368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5506173826278546e-05,
+      "loss": 3.382,
+      "step": 2106880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.549780425757176e-05,
+      "loss": 3.39,
+      "step": 2107392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.548941831006124e-05,
+      "loss": 3.3753,
+      "step": 2107904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.548103236255072e-05,
+      "loss": 3.3864,
+      "step": 2108416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.54726464150402e-05,
+      "loss": 3.3933,
+      "step": 2108928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.546426046752968e-05,
+      "loss": 3.391,
+      "step": 2109440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5455890898822888e-05,
+      "loss": 3.3683,
+      "step": 2109952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5447504951312368e-05,
+      "loss": 3.3806,
+      "step": 2110464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.543911900380185e-05,
+      "loss": 3.3807,
+      "step": 2110976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.543073305629133e-05,
+      "loss": 3.3887,
+      "step": 2111488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5422347108780808e-05,
+      "loss": 3.3869,
+      "step": 2112000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.541397754007402e-05,
+      "loss": 3.384,
+      "step": 2112512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.54055915925635e-05,
+      "loss": 3.3783,
+      "step": 2113024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.539720564505298e-05,
+      "loss": 3.3846,
+      "step": 2113536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.538881969754246e-05,
+      "loss": 3.3769,
+      "step": 2114048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.538043375003194e-05,
+      "loss": 3.3861,
+      "step": 2114560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.537204780252142e-05,
+      "loss": 3.3792,
+      "step": 2115072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.53636618550109e-05,
+      "loss": 3.3848,
+      "step": 2115584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.535527590750038e-05,
+      "loss": 3.3793,
+      "step": 2116096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.534690633879359e-05,
+      "loss": 3.3674,
+      "step": 2116608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.53385367700868e-05,
+      "loss": 3.3755,
+      "step": 2117120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.533015082257628e-05,
+      "loss": 3.3762,
+      "step": 2117632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.532176487506576e-05,
+      "loss": 3.3872,
+      "step": 2118144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.531337892755524e-05,
+      "loss": 3.3788,
+      "step": 2118656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.530499298004472e-05,
+      "loss": 3.3837,
+      "step": 2119168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5296623411337934e-05,
+      "loss": 3.3863,
+      "step": 2119680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5288237463827414e-05,
+      "loss": 3.3833,
+      "step": 2120192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5279851516316894e-05,
+      "loss": 3.3856,
+      "step": 2120704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5271465568806374e-05,
+      "loss": 3.3832,
+      "step": 2121216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5263096000099583e-05,
+      "loss": 3.3843,
+      "step": 2121728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5254710052589064e-05,
+      "loss": 3.3775,
+      "step": 2122240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5246324105078544e-05,
+      "loss": 3.3742,
+      "step": 2122752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5237938157568024e-05,
+      "loss": 3.3835,
+      "step": 2123264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5229568588861235e-05,
+      "loss": 3.381,
+      "step": 2123776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5221182641350715e-05,
+      "loss": 3.3851,
+      "step": 2124288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5212796693840195e-05,
+      "loss": 3.3829,
+      "step": 2124800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5204410746329675e-05,
+      "loss": 3.3771,
+      "step": 2125312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5196024798819156e-05,
+      "loss": 3.3764,
+      "step": 2125824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5187655230112366e-05,
+      "loss": 3.3688,
+      "step": 2126336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5179269282601846e-05,
+      "loss": 3.3877,
+      "step": 2126848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5170883335091327e-05,
+      "loss": 3.3801,
+      "step": 2127360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5162497387580807e-05,
+      "loss": 3.3933,
+      "step": 2127872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5154127818874018e-05,
+      "loss": 3.3742,
+      "step": 2128384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5145741871363498e-05,
+      "loss": 3.3907,
+      "step": 2128896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5137355923852978e-05,
+      "loss": 3.3857,
+      "step": 2129408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5128969976342456e-05,
+      "loss": 3.3749,
+      "step": 2129920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5120600407635669e-05,
+      "loss": 3.3799,
+      "step": 2130432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5112214460125149e-05,
+      "loss": 3.3774,
+      "step": 2130944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5103828512614629e-05,
+      "loss": 3.3916,
+      "step": 2131456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5095442565104107e-05,
+      "loss": 3.3906,
+      "step": 2131968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5087056617593587e-05,
+      "loss": 3.3801,
+      "step": 2132480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.50786870488868e-05,
+      "loss": 3.3776,
+      "step": 2132992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5070301101376281e-05,
+      "loss": 3.3857,
+      "step": 2133504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5061915153865757e-05,
+      "loss": 3.3885,
+      "step": 2134016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5053529206355237e-05,
+      "loss": 3.3767,
+      "step": 2134528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5045143258844719e-05,
+      "loss": 3.3897,
+      "step": 2135040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5036757311334199e-05,
+      "loss": 3.3792,
+      "step": 2135552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5028387742627408e-05,
+      "loss": 3.3872,
+      "step": 2136064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.502000179511689e-05,
+      "loss": 3.3795,
+      "step": 2136576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.925368070602417,
+      "eval_runtime": 306.9713,
+      "eval_samples_per_second": 1243.084,
+      "eval_steps_per_second": 38.847,
+      "step": 2136960
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.386509678298501e+18,
+  "trial_name": null,
+  "trial_params": null
+}