diff --git "a/checkpoint-457920/trainer_state.json" "b/checkpoint-457920/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-457920/trainer_state.json"
@@ -0,0 +1,5437 @@
+{
+  "best_metric": 3.8769750595092773,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/rel-cl2/transformer/3/checkpoints/checkpoint-457920",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 457920,
+  "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": 10.9751,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8389,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1964,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9913,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.831,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.729,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6181,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5466,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4708,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4107,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3693,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3226,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2769,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 5.2265,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 5.2016,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 5.1534,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 5.1278,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.1024,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0704,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0395,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.024,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9922,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9741,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9492,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9421,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.9073,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 4.8925,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8648,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8574,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.835,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 4.8275,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740117521192533e-05,
+      "loss": 4.8153,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731731573682013e-05,
+      "loss": 4.7978,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7871,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7747,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7605,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969820416244367e-05,
+      "loss": 4.7562,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.7263,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7161,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6945,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6983,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6842,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6841,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6636,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 4.658,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6533,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6467,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6345,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.6175,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.607,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958085003370758e-05,
+      "loss": 4.626,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957246408619706e-05,
+      "loss": 4.6135,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 4.594,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.58,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5894,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5654,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5763,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 4.5455,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5663,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5519,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5308,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5254,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5371,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4981,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.5097,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5177,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.5039,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4962,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4801,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9421582546222625e-05,
+      "loss": 4.4776,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4747,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4897,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4655,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.477,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4706,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4679,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.451,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.4509,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4463,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.4422,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.4351,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.4272,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4292,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 4.429,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.4078,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 4.4161,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 4.4141,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 4.4159,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.4003,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925392911122715e-05,
+      "loss": 4.3988,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 4.389,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 4.3841,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922878764749932e-05,
+      "loss": 4.386,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 4.3884,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.3837,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.3889,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.3802,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.3615,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.363,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.3707,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.3667,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.365,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.355,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.3449,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.3576,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.3364,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.3404,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.3355,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.3354,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.327,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907787334991742e-05,
+      "loss": 4.3271,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90694874024069e-05,
+      "loss": 4.3205,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.3286,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3227,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.3148,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.3247,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.3281,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901920447495124e-05,
+      "loss": 4.3174,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901081852744072e-05,
+      "loss": 4.3046,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.3016,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.3007,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.3064,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.3021,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968905168691855e-05,
+      "loss": 4.2912,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.3,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.2918,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.2915,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.2856,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.2863,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918605862432463e-05,
+      "loss": 4.2782,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.2889,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.2873,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2855,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.2637,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.2777,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886830655617308e-05,
+      "loss": 4.2728,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.2745,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.2644,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884316509244525e-05,
+      "loss": 4.2725,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883477914493473e-05,
+      "loss": 4.2626,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.2801,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2486,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.2673,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.2524,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8792882164989586e-05,
+      "loss": 4.2501,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2526,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877611026996855e-05,
+      "loss": 4.2388,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876772432245803e-05,
+      "loss": 4.2436,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875933837494751e-05,
+      "loss": 4.2508,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8750952427436986e-05,
+      "loss": 4.238,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.206758975982666,
+      "eval_runtime": 303.3129,
+      "eval_samples_per_second": 1258.077,
+      "eval_steps_per_second": 39.316,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87425828587302e-05,
+      "loss": 4.2339,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873419691121968e-05,
+      "loss": 4.2289,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872581096370916e-05,
+      "loss": 4.2439,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717425016198635e-05,
+      "loss": 4.2338,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870905544749185e-05,
+      "loss": 4.2424,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870066949998133e-05,
+      "loss": 4.2215,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869228355247081e-05,
+      "loss": 4.2239,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.868389760496029e-05,
+      "loss": 4.2119,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675528036253507e-05,
+      "loss": 4.2221,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667142088742986e-05,
+      "loss": 4.2176,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865875614123246e-05,
+      "loss": 4.2141,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865037019372194e-05,
+      "loss": 4.2205,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864198424621142e-05,
+      "loss": 4.1997,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633614677504635e-05,
+      "loss": 4.2117,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862522872999411e-05,
+      "loss": 4.197,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861684278248359e-05,
+      "loss": 4.1969,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.2043,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860008726626628e-05,
+      "loss": 4.1963,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591701318755764e-05,
+      "loss": 4.198,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583315371245244e-05,
+      "loss": 4.2132,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574929423734724e-05,
+      "loss": 4.1907,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856655985502793e-05,
+      "loss": 4.1932,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855817390751741e-05,
+      "loss": 4.1915,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854978796000689e-05,
+      "loss": 4.1981,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854140201249637e-05,
+      "loss": 4.1901,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853303244378958e-05,
+      "loss": 4.1893,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852464649627906e-05,
+      "loss": 4.1739,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851626054876854e-05,
+      "loss": 4.1814,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850787460125802e-05,
+      "loss": 4.1703,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84994886537475e-05,
+      "loss": 4.1817,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849111908504072e-05,
+      "loss": 4.1757,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482749516333934e-05,
+      "loss": 4.1788,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847436356882341e-05,
+      "loss": 4.1832,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846597762131289e-05,
+      "loss": 4.1711,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845759167380237e-05,
+      "loss": 4.1762,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844920572629185e-05,
+      "loss": 4.172,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844081977878133e-05,
+      "loss": 4.1624,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843243383127081e-05,
+      "loss": 4.1557,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8424064262564016e-05,
+      "loss": 4.1473,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415678315053496e-05,
+      "loss": 4.1549,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8407292367542976e-05,
+      "loss": 4.1553,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398906420032456e-05,
+      "loss": 4.1624,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839053685132567e-05,
+      "loss": 4.1507,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838215090381515e-05,
+      "loss": 4.1486,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837376495630463e-05,
+      "loss": 4.1542,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836537900879411e-05,
+      "loss": 4.1477,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835700944008732e-05,
+      "loss": 4.1483,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83486234925768e-05,
+      "loss": 4.1356,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834023754506628e-05,
+      "loss": 4.1328,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833185159755576e-05,
+      "loss": 4.1569,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832346565004524e-05,
+      "loss": 4.1533,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831509608133845e-05,
+      "loss": 4.137,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830671013382793e-05,
+      "loss": 4.1392,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829834056512114e-05,
+      "loss": 4.1433,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289954617610625e-05,
+      "loss": 4.1246,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281568670100105e-05,
+      "loss": 4.1434,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273182722589585e-05,
+      "loss": 4.1177,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264796775079065e-05,
+      "loss": 4.1445,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256410827568545e-05,
+      "loss": 4.1392,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248024880058025e-05,
+      "loss": 4.121,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239655311351234e-05,
+      "loss": 4.1194,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231269363840714e-05,
+      "loss": 4.1344,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222883416330194e-05,
+      "loss": 4.1004,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214497468819674e-05,
+      "loss": 4.1178,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820612790011288e-05,
+      "loss": 4.1285,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819774195260236e-05,
+      "loss": 4.1218,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818935600509184e-05,
+      "loss": 4.1126,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818097005758132e-05,
+      "loss": 4.1008,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817260048887454e-05,
+      "loss": 4.1089,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816421454136402e-05,
+      "loss": 4.1099,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81558285938535e-05,
+      "loss": 4.1186,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814744264634298e-05,
+      "loss": 4.1073,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813907307763619e-05,
+      "loss": 4.1162,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813068713012567e-05,
+      "loss": 4.1182,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812230118261515e-05,
+      "loss": 4.1164,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811391523510463e-05,
+      "loss": 4.1007,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810554566639784e-05,
+      "loss": 4.1078,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809715971888732e-05,
+      "loss": 4.1074,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80887737713768e-05,
+      "loss": 4.1014,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808038782386628e-05,
+      "loss": 4.1036,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807201825515949e-05,
+      "loss": 4.0974,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806363230764897e-05,
+      "loss": 4.1048,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805524636013845e-05,
+      "loss": 4.1024,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804686041262793e-05,
+      "loss": 4.0836,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803849084392114e-05,
+      "loss": 4.0912,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803010489641062e-05,
+      "loss": 4.0954,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80217189489001e-05,
+      "loss": 4.1019,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801333300138958e-05,
+      "loss": 4.0921,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800496343268279e-05,
+      "loss": 4.0892,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799657748517227e-05,
+      "loss": 4.0846,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798819153766175e-05,
+      "loss": 4.0763,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797980559015123e-05,
+      "loss": 4.0809,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971436021444446e-05,
+      "loss": 4.0893,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7963050073933926e-05,
+      "loss": 4.0859,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954664126423406e-05,
+      "loss": 4.0948,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946278178912886e-05,
+      "loss": 4.0854,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937908610206095e-05,
+      "loss": 4.0774,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929522662695575e-05,
+      "loss": 4.0753,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7921136715185055e-05,
+      "loss": 4.0832,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912750767674535e-05,
+      "loss": 4.0869,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7904381198967744e-05,
+      "loss": 4.0845,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895995251457224e-05,
+      "loss": 4.0764,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887609303946704e-05,
+      "loss": 4.0669,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879223356436184e-05,
+      "loss": 4.0843,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78708537877294e-05,
+      "loss": 4.064,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786246784021888e-05,
+      "loss": 4.0713,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785408189270836e-05,
+      "loss": 4.0704,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784569594519783e-05,
+      "loss": 4.069,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783734275529478e-05,
+      "loss": 4.0601,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782895680778426e-05,
+      "loss": 4.0697,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782057086027374e-05,
+      "loss": 4.0576,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781218491276322e-05,
+      "loss": 4.0676,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78037989652527e-05,
+      "loss": 4.0658,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779541301774218e-05,
+      "loss": 4.0582,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778702707023166e-05,
+      "loss": 4.0777,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777864112272114e-05,
+      "loss": 4.0754,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770271554014354e-05,
+      "loss": 4.0677,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761885606503834e-05,
+      "loss": 4.058,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775349965899331e-05,
+      "loss": 4.0522,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774511371148279e-05,
+      "loss": 4.0587,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773672776397227e-05,
+      "loss": 4.0553,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7728341816461747e-05,
+      "loss": 4.0645,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719955868951227e-05,
+      "loss": 4.046,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7711569921440707e-05,
+      "loss": 4.0568,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703200352733916e-05,
+      "loss": 4.0583,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694814405223396e-05,
+      "loss": 4.06,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686428457712876e-05,
+      "loss": 4.0453,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767805888900609e-05,
+      "loss": 4.0531,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766967294149557e-05,
+      "loss": 4.0514,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766128699398505e-05,
+      "loss": 4.0542,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765290104647453e-05,
+      "loss": 4.0592,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764453147776774e-05,
+      "loss": 4.0587,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763614553025722e-05,
+      "loss": 4.0364,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76277595827467e-05,
+      "loss": 4.0494,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761939001403991e-05,
+      "loss": 4.0513,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761100406652939e-05,
+      "loss": 4.05,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760261811901887e-05,
+      "loss": 4.047,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759423217150835e-05,
+      "loss": 4.0517,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758584622399783e-05,
+      "loss": 4.0453,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757746027648731e-05,
+      "loss": 4.0607,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569090707780525e-05,
+      "loss": 4.0361,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560704760270005e-05,
+      "loss": 4.0509,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552318812759485e-05,
+      "loss": 4.0399,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543932865248965e-05,
+      "loss": 4.0374,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535563296542174e-05,
+      "loss": 4.0418,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527177349031654e-05,
+      "loss": 4.0319,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518791401521134e-05,
+      "loss": 4.0351,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510405454010614e-05,
+      "loss": 4.0369,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7502019506500094e-05,
+      "loss": 4.0326,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.035162925720215,
+      "eval_runtime": 303.1156,
+      "eval_samples_per_second": 1258.896,
+      "eval_steps_per_second": 39.341,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493633558989574e-05,
+      "loss": 4.0293,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485247611479054e-05,
+      "loss": 4.0246,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476861663968534e-05,
+      "loss": 4.041,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468475716458014e-05,
+      "loss": 4.0305,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460089768947494e-05,
+      "loss": 4.048,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451703821436974e-05,
+      "loss": 4.0216,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443317873926453e-05,
+      "loss": 4.0286,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434931926415933e-05,
+      "loss": 4.0127,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426545978905413e-05,
+      "loss": 4.0294,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741816003139489e-05,
+      "loss": 4.0244,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740977408388437e-05,
+      "loss": 4.0201,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740140451517758e-05,
+      "loss": 4.0309,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739303494647079e-05,
+      "loss": 4.008,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738464899896027e-05,
+      "loss": 4.0249,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737626305144975e-05,
+      "loss": 4.0077,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736787710393923e-05,
+      "loss": 4.0097,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735949115642871e-05,
+      "loss": 4.0099,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73511052089182e-05,
+      "loss": 4.0181,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734271926140768e-05,
+      "loss": 4.0111,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733433331389716e-05,
+      "loss": 4.0303,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7325980123994096e-05,
+      "loss": 4.0102,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7317594176483576e-05,
+      "loss": 4.0132,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7309208228973056e-05,
+      "loss": 4.0122,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300822281462536e-05,
+      "loss": 4.0211,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292436333952016e-05,
+      "loss": 4.0076,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7284050386441496e-05,
+      "loss": 4.0136,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275664438930976e-05,
+      "loss": 3.9999,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267278491420456e-05,
+      "loss": 4.0062,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258908922713665e-05,
+      "loss": 3.9943,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725052297520315e-05,
+      "loss": 4.009,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724213702769263e-05,
+      "loss": 4.0067,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723376745898584e-05,
+      "loss": 4.0094,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722538151147532e-05,
+      "loss": 4.0147,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72169955639648e-05,
+      "loss": 4.0008,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720860961645428e-05,
+      "loss": 4.0003,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720022366894376e-05,
+      "loss": 4.0055,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719183772143324e-05,
+      "loss": 3.9971,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718345177392272e-05,
+      "loss": 3.9903,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717508220521593e-05,
+      "loss": 3.9818,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716671263650914e-05,
+      "loss": 3.9915,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715832668899862e-05,
+      "loss": 3.9904,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149940741488105e-05,
+      "loss": 4.0001,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141554793977585e-05,
+      "loss": 3.9924,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133168846467065e-05,
+      "loss": 3.984,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124799277760274e-05,
+      "loss": 3.9977,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116413330249754e-05,
+      "loss": 3.9897,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7108027382739234e-05,
+      "loss": 3.9858,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099641435228714e-05,
+      "loss": 3.9811,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091255487718194e-05,
+      "loss": 3.9748,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082869540207674e-05,
+      "loss": 3.9998,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074483592697154e-05,
+      "loss": 3.9977,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706609764518663e-05,
+      "loss": 3.9833,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705772807647984e-05,
+      "loss": 3.9833,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704934212896932e-05,
+      "loss": 3.9917,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70409561814588e-05,
+      "loss": 3.968,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703257023394828e-05,
+      "loss": 3.9898,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702418428643776e-05,
+      "loss": 3.968,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701579833892724e-05,
+      "loss": 3.9903,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700741239141672e-05,
+      "loss": 3.9943,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69990264439062e-05,
+      "loss": 3.9756,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699067325400315e-05,
+      "loss": 3.9667,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698228730649263e-05,
+      "loss": 3.9871,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69739013589821e-05,
+      "loss": 3.9568,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696551541147158e-05,
+      "loss": 3.9719,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695712946396106e-05,
+      "loss": 3.9813,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694874351645054e-05,
+      "loss": 3.9793,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694035756894003e-05,
+      "loss": 3.9692,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693197162142951e-05,
+      "loss": 3.9583,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692361843152645e-05,
+      "loss": 3.9626,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915232484015926e-05,
+      "loss": 3.9651,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906846536505406e-05,
+      "loss": 3.9758,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898460588994886e-05,
+      "loss": 3.9646,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6890074641484366e-05,
+      "loss": 3.9762,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881688693973846e-05,
+      "loss": 3.9784,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873319125267055e-05,
+      "loss": 3.978,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864933177756535e-05,
+      "loss": 3.9581,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856547230246015e-05,
+      "loss": 3.9692,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848161282735495e-05,
+      "loss": 3.9716,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683977533522498e-05,
+      "loss": 3.9626,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683140576651819e-05,
+      "loss": 3.9665,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682301981900767e-05,
+      "loss": 3.9638,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681463387149715e-05,
+      "loss": 3.9646,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680624792398663e-05,
+      "loss": 3.9694,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679786197647611e-05,
+      "loss": 3.9479,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678947602896559e-05,
+      "loss": 3.9603,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67811064602588e-05,
+      "loss": 3.9578,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677272051274828e-05,
+      "loss": 3.9692,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676433456523776e-05,
+      "loss": 3.958,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675594861772724e-05,
+      "loss": 3.9579,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674756267021672e-05,
+      "loss": 3.9531,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739193101509935e-05,
+      "loss": 3.9471,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730807153999415e-05,
+      "loss": 3.9469,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722421206488895e-05,
+      "loss": 3.9571,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6714035258978375e-05,
+      "loss": 3.9653,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705665690271584e-05,
+      "loss": 3.9628,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697279742761064e-05,
+      "loss": 3.9553,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688893795250544e-05,
+      "loss": 3.9491,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680507847740024e-05,
+      "loss": 3.9497,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667213827903323e-05,
+      "loss": 3.9537,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666375233152271e-05,
+      "loss": 3.9605,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665536638401219e-05,
+      "loss": 3.9576,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664698043650167e-05,
+      "loss": 3.9503,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663859448899115e-05,
+      "loss": 3.9415,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663022492028437e-05,
+      "loss": 3.9625,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662183897277385e-05,
+      "loss": 3.9389,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661345302526333e-05,
+      "loss": 3.9486,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660506707775281e-05,
+      "loss": 3.9421,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659668113024229e-05,
+      "loss": 3.9482,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65883115615355e-05,
+      "loss": 3.9386,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657992561402498e-05,
+      "loss": 3.948,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657153966651446e-05,
+      "loss": 3.9395,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656315371900393e-05,
+      "loss": 3.9403,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655476777149341e-05,
+      "loss": 3.9496,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546398202786626e-05,
+      "loss": 3.9345,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6538012255276106e-05,
+      "loss": 3.9567,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529626307765586e-05,
+      "loss": 3.9549,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6521240360255066e-05,
+      "loss": 3.9486,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512854412744546e-05,
+      "loss": 3.9397,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650448484403776e-05,
+      "loss": 3.9356,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6496098896527235e-05,
+      "loss": 3.9368,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487712949016715e-05,
+      "loss": 3.9377,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479327001506195e-05,
+      "loss": 3.9462,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470941053995675e-05,
+      "loss": 3.9279,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462555106485155e-05,
+      "loss": 3.9419,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454169158974635e-05,
+      "loss": 3.9439,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445783211464115e-05,
+      "loss": 3.9414,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437413642757324e-05,
+      "loss": 3.9315,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429027695246804e-05,
+      "loss": 3.9368,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642064174773629e-05,
+      "loss": 3.9391,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641225580022577e-05,
+      "loss": 3.9395,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640386985271525e-05,
+      "loss": 3.9414,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639548390520473e-05,
+      "loss": 3.9466,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638711433649794e-05,
+      "loss": 3.9217,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637872838898742e-05,
+      "loss": 3.9372,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63703424414769e-05,
+      "loss": 3.9392,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636195649396638e-05,
+      "loss": 3.9344,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635358692525959e-05,
+      "loss": 3.9331,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634520097774907e-05,
+      "loss": 3.9433,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633681503023855e-05,
+      "loss": 3.9355,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632842908272803e-05,
+      "loss": 3.9488,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632004313521751e-05,
+      "loss": 3.9256,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631165718770699e-05,
+      "loss": 3.9364,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303271240196475e-05,
+      "loss": 3.9325,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294901671489684e-05,
+      "loss": 3.9283,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286515723979164e-05,
+      "loss": 3.9312,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6278129776468644e-05,
+      "loss": 3.9253,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626974382895812e-05,
+      "loss": 3.9216,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626137426025133e-05,
+      "loss": 3.9274,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625298831274081e-05,
+      "loss": 3.9254,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9593703746795654,
+      "eval_runtime": 320.1607,
+      "eval_samples_per_second": 1191.873,
+      "eval_steps_per_second": 37.247,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624460236523029e-05,
+      "loss": 3.9267,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6236216417719766e-05,
+      "loss": 3.9158,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227830470209246e-05,
+      "loss": 3.9321,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621946090150246e-05,
+      "loss": 3.9244,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621107495399194e-05,
+      "loss": 3.9444,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620268900648142e-05,
+      "loss": 3.9171,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61943030589709e-05,
+      "loss": 3.9238,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618591711146038e-05,
+      "loss": 3.9084,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617753116394986e-05,
+      "loss": 3.926,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616914521643934e-05,
+      "loss": 3.9221,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616075926892882e-05,
+      "loss": 3.9155,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615238970022203e-05,
+      "loss": 3.9228,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614400375271151e-05,
+      "loss": 3.9106,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613561780520099e-05,
+      "loss": 3.9223,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612723185769047e-05,
+      "loss": 3.9074,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611886228898368e-05,
+      "loss": 3.9052,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611047634147316e-05,
+      "loss": 3.9094,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610209039396265e-05,
+      "loss": 3.9141,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609370444645213e-05,
+      "loss": 3.9097,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085334877745336e-05,
+      "loss": 3.9284,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076948930234816e-05,
+      "loss": 3.9101,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068562982724296e-05,
+      "loss": 3.9195,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6060177035213776e-05,
+      "loss": 3.9102,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051791087703256e-05,
+      "loss": 3.9171,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043421518996465e-05,
+      "loss": 3.9085,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6035035571485945e-05,
+      "loss": 3.9153,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026649623975425e-05,
+      "loss": 3.9073,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018263676464904e-05,
+      "loss": 3.9032,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009894107758114e-05,
+      "loss": 3.8995,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.60015081602476e-05,
+      "loss": 3.9082,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599312221273708e-05,
+      "loss": 3.9122,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598473626522656e-05,
+      "loss": 3.9118,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597635031771604e-05,
+      "loss": 3.9205,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596796437020552e-05,
+      "loss": 3.9068,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595959480149873e-05,
+      "loss": 3.9036,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595120885398821e-05,
+      "loss": 3.9084,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594282290647769e-05,
+      "loss": 3.9088,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593443695896717e-05,
+      "loss": 3.8918,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592605101145665e-05,
+      "loss": 3.8886,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591768144274986e-05,
+      "loss": 3.8944,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590929549523934e-05,
+      "loss": 3.9021,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590090954772882e-05,
+      "loss": 3.9043,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58925236002183e-05,
+      "loss": 3.9019,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5884137652707785e-05,
+      "loss": 3.8894,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587575170519726e-05,
+      "loss": 3.9038,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586736575768674e-05,
+      "loss": 3.8976,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585897981017622e-05,
+      "loss": 3.894,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585061024146943e-05,
+      "loss": 3.8865,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584224067276264e-05,
+      "loss": 3.8761,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583385472525212e-05,
+      "loss": 3.9142,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58254687777416e-05,
+      "loss": 3.9028,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5817082830231076e-05,
+      "loss": 3.8936,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580871326152429e-05,
+      "loss": 3.8942,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580032731401377e-05,
+      "loss": 3.8961,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579194136650325e-05,
+      "loss": 3.8807,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578355541899273e-05,
+      "loss": 3.8966,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577516947148221e-05,
+      "loss": 3.882,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576679990277543e-05,
+      "loss": 3.8946,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57584139552649e-05,
+      "loss": 3.9082,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575002800775438e-05,
+      "loss": 3.8864,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574164206024386e-05,
+      "loss": 3.8737,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573325611273334e-05,
+      "loss": 3.9029,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572487016522282e-05,
+      "loss": 3.8694,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571650059651603e-05,
+      "loss": 3.88,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570811464900551e-05,
+      "loss": 3.8978,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569972870149499e-05,
+      "loss": 3.89,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691342753984476e-05,
+      "loss": 3.8799,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5682973185277685e-05,
+      "loss": 3.8763,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674587237767165e-05,
+      "loss": 3.8714,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5666201290256645e-05,
+      "loss": 3.8772,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657815342746125e-05,
+      "loss": 3.8912,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649445774039334e-05,
+      "loss": 3.8758,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5641059826528814e-05,
+      "loss": 3.8917,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632673879018294e-05,
+      "loss": 3.887,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624287931507774e-05,
+      "loss": 3.8945,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561591836280098e-05,
+      "loss": 3.8767,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560753241529046e-05,
+      "loss": 3.8804,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559914646777994e-05,
+      "loss": 3.8874,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559076052026943e-05,
+      "loss": 3.8749,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558239095156264e-05,
+      "loss": 3.8868,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557400500405212e-05,
+      "loss": 3.8786,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55656190565416e-05,
+      "loss": 3.8784,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555723310903108e-05,
+      "loss": 3.8822,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554884716152056e-05,
+      "loss": 3.8667,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554047759281377e-05,
+      "loss": 3.8749,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553209164530325e-05,
+      "loss": 3.8786,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552370569779273e-05,
+      "loss": 3.8814,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551531975028221e-05,
+      "loss": 3.8715,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550693380277169e-05,
+      "loss": 3.8766,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54985642340649e-05,
+      "loss": 3.8692,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5490178286554384e-05,
+      "loss": 3.8645,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481792339043864e-05,
+      "loss": 3.8658,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5473406391533343e-05,
+      "loss": 3.8729,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546503682282655e-05,
+      "loss": 3.884,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545665087531603e-05,
+      "loss": 3.8854,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544826492780551e-05,
+      "loss": 3.8753,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543987898029499e-05,
+      "loss": 3.8639,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543149303278447e-05,
+      "loss": 3.8693,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542312346407768e-05,
+      "loss": 3.8745,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541473751656716e-05,
+      "loss": 3.8786,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540635156905664e-05,
+      "loss": 3.8775,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539796562154612e-05,
+      "loss": 3.8706,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538959605283933e-05,
+      "loss": 3.8615,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538121010532882e-05,
+      "loss": 3.8846,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53728241578183e-05,
+      "loss": 3.8542,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536443821030778e-05,
+      "loss": 3.8714,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5356068641600986e-05,
+      "loss": 3.8601,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347682694090466e-05,
+      "loss": 3.8705,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339296746579946e-05,
+      "loss": 3.8619,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5330910799069426e-05,
+      "loss": 3.8684,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5322524851558906e-05,
+      "loss": 3.8615,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314155282852115e-05,
+      "loss": 3.8624,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305769335341595e-05,
+      "loss": 3.8665,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297383387831075e-05,
+      "loss": 3.8607,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5288997440320555e-05,
+      "loss": 3.877,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528062787161377e-05,
+      "loss": 3.8776,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527224192410325e-05,
+      "loss": 3.8724,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526385597659273e-05,
+      "loss": 3.8615,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525547002908221e-05,
+      "loss": 3.861,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524710046037542e-05,
+      "loss": 3.8594,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52387145128649e-05,
+      "loss": 3.8605,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523032856535438e-05,
+      "loss": 3.8719,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522194261784386e-05,
+      "loss": 3.848,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521355667033333e-05,
+      "loss": 3.8655,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520518710162655e-05,
+      "loss": 3.8628,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519680115411603e-05,
+      "loss": 3.8697,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518841520660551e-05,
+      "loss": 3.8542,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5180045637898725e-05,
+      "loss": 3.8624,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5171659690388204e-05,
+      "loss": 3.8579,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5163273742877684e-05,
+      "loss": 3.8702,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515488779536716e-05,
+      "loss": 3.8645,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146518226660373e-05,
+      "loss": 3.8712,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5138132279149853e-05,
+      "loss": 3.8475,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129746331639333e-05,
+      "loss": 3.8646,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121360384128807e-05,
+      "loss": 3.8655,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511299081542202e-05,
+      "loss": 3.8547,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51046048679115e-05,
+      "loss": 3.8613,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096218920400976e-05,
+      "loss": 3.8702,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508783297289046e-05,
+      "loss": 3.8621,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507944702537994e-05,
+      "loss": 3.8769,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507107745667316e-05,
+      "loss": 3.8509,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506269150916263e-05,
+      "loss": 3.8642,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505430556165211e-05,
+      "loss": 3.8531,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504591961414159e-05,
+      "loss": 3.8555,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50375500454348e-05,
+      "loss": 3.8588,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502916409792428e-05,
+      "loss": 3.8521,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502077815041376e-05,
+      "loss": 3.8509,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012408581706976e-05,
+      "loss": 3.8512,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500402263419645e-05,
+      "loss": 3.8541,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.917726993560791,
+      "eval_runtime": 316.9205,
+      "eval_samples_per_second": 1204.059,
+      "eval_steps_per_second": 37.628,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499563668668593e-05,
+      "loss": 3.8512,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4987250739175416e-05,
+      "loss": 3.8443,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978864791664896e-05,
+      "loss": 3.8611,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970478844154376e-05,
+      "loss": 3.8527,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4962092896643856e-05,
+      "loss": 3.8695,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953706949133336e-05,
+      "loss": 3.8503,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945321001622816e-05,
+      "loss": 3.8525,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936935054112296e-05,
+      "loss": 3.8392,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928549106601776e-05,
+      "loss": 3.853,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4920179537894985e-05,
+      "loss": 3.8507,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911793590384465e-05,
+      "loss": 3.8482,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903407642873945e-05,
+      "loss": 3.8493,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4895038074167154e-05,
+      "loss": 3.8386,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886652126656634e-05,
+      "loss": 3.8529,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878266179146114e-05,
+      "loss": 3.8366,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48698802316356e-05,
+      "loss": 3.8353,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486149428412508e-05,
+      "loss": 3.838,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485310833661456e-05,
+      "loss": 3.8466,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484472238910404e-05,
+      "loss": 3.8391,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483633644159352e-05,
+      "loss": 3.8573,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482796687288673e-05,
+      "loss": 3.8426,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481958092537621e-05,
+      "loss": 3.8525,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481119497786569e-05,
+      "loss": 3.8416,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480280903035516e-05,
+      "loss": 3.8455,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479442308284464e-05,
+      "loss": 3.8402,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478605351413786e-05,
+      "loss": 3.8482,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477766756662734e-05,
+      "loss": 3.8392,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476928161911682e-05,
+      "loss": 3.8387,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47608956716063e-05,
+      "loss": 3.8305,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752526102899514e-05,
+      "loss": 3.8399,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4744140155388994e-05,
+      "loss": 3.8454,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473575420787847e-05,
+      "loss": 3.8458,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472736826036795e-05,
+      "loss": 3.8487,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471899869166116e-05,
+      "loss": 3.8361,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710612744150636e-05,
+      "loss": 3.8379,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4702226796640116e-05,
+      "loss": 3.846,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693840849129596e-05,
+      "loss": 3.8433,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468547128042281e-05,
+      "loss": 3.8223,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4677085332912285e-05,
+      "loss": 3.8286,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466869938540177e-05,
+      "loss": 3.8277,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466031343789125e-05,
+      "loss": 3.8316,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465194386918447e-05,
+      "loss": 3.8423,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464355792167394e-05,
+      "loss": 3.8332,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463517197416342e-05,
+      "loss": 3.8279,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46267860266529e-05,
+      "loss": 3.8358,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461841645794611e-05,
+      "loss": 3.8293,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461003051043559e-05,
+      "loss": 3.8322,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460164456292507e-05,
+      "loss": 3.8246,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459325861541455e-05,
+      "loss": 3.8067,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458487266790403e-05,
+      "loss": 3.8514,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457650309919724e-05,
+      "loss": 3.837,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4568117151686725e-05,
+      "loss": 3.8337,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559731204176205e-05,
+      "loss": 3.8281,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551361635469415e-05,
+      "loss": 3.8294,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542975687958895e-05,
+      "loss": 3.8167,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534589740448374e-05,
+      "loss": 3.8302,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4526203792937854e-05,
+      "loss": 3.8228,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517817845427334e-05,
+      "loss": 3.8281,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509431897916814e-05,
+      "loss": 3.8432,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4501045950406294e-05,
+      "loss": 3.8224,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492660002895774e-05,
+      "loss": 3.8139,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448429043418898e-05,
+      "loss": 3.8387,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447590448667846e-05,
+      "loss": 3.8075,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446751853916794e-05,
+      "loss": 3.82,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445913259165742e-05,
+      "loss": 3.8307,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445076302295064e-05,
+      "loss": 3.8277,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444237707544012e-05,
+      "loss": 3.8187,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44339911279296e-05,
+      "loss": 3.8162,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442562155922281e-05,
+      "loss": 3.806,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441723561171229e-05,
+      "loss": 3.8156,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440884966420177e-05,
+      "loss": 3.8274,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440046371669125e-05,
+      "loss": 3.8183,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439209414798446e-05,
+      "loss": 3.8258,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438370820047394e-05,
+      "loss": 3.8259,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437532225296342e-05,
+      "loss": 3.8354,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43669363054529e-05,
+      "loss": 3.8147,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435856673674611e-05,
+      "loss": 3.8166,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435018078923559e-05,
+      "loss": 3.826,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434179484172507e-05,
+      "loss": 3.8113,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433340889421455e-05,
+      "loss": 3.83,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432503932550776e-05,
+      "loss": 3.8182,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431665337799724e-05,
+      "loss": 3.8171,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430826743048672e-05,
+      "loss": 3.8245,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42998814829762e-05,
+      "loss": 3.8049,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429149553546568e-05,
+      "loss": 3.8117,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428312596675889e-05,
+      "loss": 3.8155,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427474001924837e-05,
+      "loss": 3.8252,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426635407173785e-05,
+      "loss": 3.8093,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425796812422733e-05,
+      "loss": 3.8168,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249598555520546e-05,
+      "loss": 3.8111,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4241212608010026e-05,
+      "loss": 3.8098,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232826660499506e-05,
+      "loss": 3.8032,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224440712988986e-05,
+      "loss": 3.8152,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216071144282195e-05,
+      "loss": 3.8244,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207685196771675e-05,
+      "loss": 3.8225,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4199299249261155e-05,
+      "loss": 3.8212,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190913301750635e-05,
+      "loss": 3.804,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182543733043844e-05,
+      "loss": 3.8103,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4174157785533324e-05,
+      "loss": 3.8115,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165771838022804e-05,
+      "loss": 3.82,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157385890512284e-05,
+      "loss": 3.8185,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41490163218055e-05,
+      "loss": 3.8125,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414063037429498e-05,
+      "loss": 3.8052,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413224442678446e-05,
+      "loss": 3.8253,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412385847927394e-05,
+      "loss": 3.7991,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411548891056715e-05,
+      "loss": 3.8115,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410710296305663e-05,
+      "loss": 3.8024,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409871701554611e-05,
+      "loss": 3.8135,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409033106803559e-05,
+      "loss": 3.8024,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40819614993288e-05,
+      "loss": 3.8114,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407357555181828e-05,
+      "loss": 3.8035,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406518960430776e-05,
+      "loss": 3.8075,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405680365679724e-05,
+      "loss": 3.8093,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4048434088090454e-05,
+      "loss": 3.8048,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4040048140579934e-05,
+      "loss": 3.8173,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403167857187314e-05,
+      "loss": 3.8205,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402329262436262e-05,
+      "loss": 3.8118,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40149066768521e-05,
+      "loss": 3.8107,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400652072934158e-05,
+      "loss": 3.801,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399813478183106e-05,
+      "loss": 3.8058,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398974883432054e-05,
+      "loss": 3.8023,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981362886810016e-05,
+      "loss": 3.8182,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3972976939299496e-05,
+      "loss": 3.7888,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396460737059271e-05,
+      "loss": 3.8082,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395622142308219e-05,
+      "loss": 3.8069,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394783547557167e-05,
+      "loss": 3.8122,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393944952806115e-05,
+      "loss": 3.8012,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393107995935437e-05,
+      "loss": 3.8073,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392269401184384e-05,
+      "loss": 3.8008,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391430806433332e-05,
+      "loss": 3.8109,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39059221168228e-05,
+      "loss": 3.8116,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897552548116016e-05,
+      "loss": 3.8137,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388916660060549e-05,
+      "loss": 3.7958,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388078065309497e-05,
+      "loss": 3.8085,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387239470558445e-05,
+      "loss": 3.8127,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3864025136877665e-05,
+      "loss": 3.8015,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855639189367145e-05,
+      "loss": 3.8047,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3847253241856625e-05,
+      "loss": 3.8129,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838867294346105e-05,
+      "loss": 3.8052,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830497725639314e-05,
+      "loss": 3.8257,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3822111778128794e-05,
+      "loss": 3.7962,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813725830618274e-05,
+      "loss": 3.8079,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805339883107754e-05,
+      "loss": 3.7983,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379697031440096e-05,
+      "loss": 3.8015,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378858436689044e-05,
+      "loss": 3.8064,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378019841937992e-05,
+      "loss": 3.7971,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37718124718694e-05,
+      "loss": 3.8009,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376344290316262e-05,
+      "loss": 3.7875,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37550569556521e-05,
+      "loss": 3.8051,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8928475379943848,
+      "eval_runtime": 302.8902,
+      "eval_samples_per_second": 1259.833,
+      "eval_steps_per_second": 39.371,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374667100814158e-05,
+      "loss": 3.7998,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373828506063106e-05,
+      "loss": 3.7915,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372989911312054e-05,
+      "loss": 3.8059,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372151316561002e-05,
+      "loss": 3.8032,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.37131272180995e-05,
+      "loss": 3.8136,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370474127058898e-05,
+      "loss": 3.7949,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369635532307846e-05,
+      "loss": 3.7975,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368796937556794e-05,
+      "loss": 3.7864,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367959980686115e-05,
+      "loss": 3.8005,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367121385935063e-05,
+      "loss": 3.7981,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366282791184011e-05,
+      "loss": 3.7974,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365444196432959e-05,
+      "loss": 3.7909,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36460723956228e-05,
+      "loss": 3.7909,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363768644811228e-05,
+      "loss": 3.7933,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362930050060176e-05,
+      "loss": 3.787,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362091455309124e-05,
+      "loss": 3.7797,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361252860558072e-05,
+      "loss": 3.7859,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36041426580702e-05,
+      "loss": 3.794,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3595756710559676e-05,
+      "loss": 3.787,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3587370763049156e-05,
+      "loss": 3.8049,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357900119434237e-05,
+      "loss": 3.7932,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3570615246831845e-05,
+      "loss": 3.8024,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3562229299321325e-05,
+      "loss": 3.7893,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3553843351810805e-05,
+      "loss": 3.7921,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354547378310402e-05,
+      "loss": 3.787,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35370878355935e-05,
+      "loss": 3.7993,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352870188808298e-05,
+      "loss": 3.7884,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352031594057246e-05,
+      "loss": 3.7868,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351194637186568e-05,
+      "loss": 3.78,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350356042435515e-05,
+      "loss": 3.789,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349517447684463e-05,
+      "loss": 3.7927,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348678852933411e-05,
+      "loss": 3.7962,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347841896062732e-05,
+      "loss": 3.796,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34700330131168e-05,
+      "loss": 3.7928,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346164706560628e-05,
+      "loss": 3.7877,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345326111809576e-05,
+      "loss": 3.7959,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3444891549388975e-05,
+      "loss": 3.7919,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3436505601878455e-05,
+      "loss": 3.7758,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3428119654367935e-05,
+      "loss": 3.779,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3419733706857415e-05,
+      "loss": 3.7745,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3411364138150624e-05,
+      "loss": 3.7782,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402978190640104e-05,
+      "loss": 3.7938,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3394592243129584e-05,
+      "loss": 3.7857,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3386206295619064e-05,
+      "loss": 3.7755,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337783672691227e-05,
+      "loss": 3.7858,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336945077940175e-05,
+      "loss": 3.7779,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336106483189123e-05,
+      "loss": 3.7839,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335267888438071e-05,
+      "loss": 3.7734,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334430931567393e-05,
+      "loss": 3.7568,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333592336816341e-05,
+      "loss": 3.8026,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332753742065289e-05,
+      "loss": 3.7843,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331915147314237e-05,
+      "loss": 3.7876,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331078190443558e-05,
+      "loss": 3.775,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330241233572879e-05,
+      "loss": 3.7844,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3294026388218266e-05,
+      "loss": 3.7685,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3285640440707746e-05,
+      "loss": 3.7784,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3277254493197226e-05,
+      "loss": 3.7779,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3268868545686706e-05,
+      "loss": 3.7782,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3260482598176186e-05,
+      "loss": 3.797,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3252096650665666e-05,
+      "loss": 3.7735,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3243710703155146e-05,
+      "loss": 3.764,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323534113444836e-05,
+      "loss": 3.7907,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322695518693784e-05,
+      "loss": 3.7582,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321856923942732e-05,
+      "loss": 3.7678,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32101832919168e-05,
+      "loss": 3.7852,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320181372321001e-05,
+      "loss": 3.7817,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319342777569949e-05,
+      "loss": 3.766,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318504182818897e-05,
+      "loss": 3.773,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317665588067845e-05,
+      "loss": 3.7558,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316828631197166e-05,
+      "loss": 3.7705,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315990036446114e-05,
+      "loss": 3.7786,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315151441695062e-05,
+      "loss": 3.7666,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3143144848243836e-05,
+      "loss": 3.7792,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3134758900733316e-05,
+      "loss": 3.7771,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3126372953222796e-05,
+      "loss": 3.7887,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3117987005712276e-05,
+      "loss": 3.7657,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109617437005485e-05,
+      "loss": 3.7736,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3101231489494965e-05,
+      "loss": 3.7784,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3092845541984445e-05,
+      "loss": 3.7641,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3084459594473925e-05,
+      "loss": 3.7786,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3076090025767134e-05,
+      "loss": 3.7733,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067704078256614e-05,
+      "loss": 3.7698,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3059318130746094e-05,
+      "loss": 3.7798,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3050932183235574e-05,
+      "loss": 3.7597,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304256261452879e-05,
+      "loss": 3.7627,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.303417666701827e-05,
+      "loss": 3.771,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.302579071950775e-05,
+      "loss": 3.7747,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.301740477199723e-05,
+      "loss": 3.7649,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300903520329044e-05,
+      "loss": 3.7683,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300064925577992e-05,
+      "loss": 3.7598,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29922633082694e-05,
+      "loss": 3.7661,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298387736075888e-05,
+      "loss": 3.7587,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297550779205209e-05,
+      "loss": 3.7669,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296712184454157e-05,
+      "loss": 3.7812,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295873589703105e-05,
+      "loss": 3.7724,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295034994952053e-05,
+      "loss": 3.774,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294198038081374e-05,
+      "loss": 3.76,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293359443330322e-05,
+      "loss": 3.7637,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29252084857927e-05,
+      "loss": 3.7651,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291682253828218e-05,
+      "loss": 3.7767,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290845296957539e-05,
+      "loss": 3.7683,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290006702206487e-05,
+      "loss": 3.7726,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289168107455435e-05,
+      "loss": 3.7597,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288329512704383e-05,
+      "loss": 3.7765,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287492555833704e-05,
+      "loss": 3.7561,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286653961082652e-05,
+      "loss": 3.7673,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2858153663316e-05,
+      "loss": 3.7539,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284976771580548e-05,
+      "loss": 3.7721,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28413981470987e-05,
+      "loss": 3.7607,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283301219958818e-05,
+      "loss": 3.7631,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282462625207766e-05,
+      "loss": 3.7592,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281624030456714e-05,
+      "loss": 3.7628,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2807870735860346e-05,
+      "loss": 3.7633,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2799484788349826e-05,
+      "loss": 3.7623,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2791098840839306e-05,
+      "loss": 3.7706,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2782729272132515e-05,
+      "loss": 3.774,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2774343324621995e-05,
+      "loss": 3.7669,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2765957377111475e-05,
+      "loss": 3.7701,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2757571429600955e-05,
+      "loss": 3.755,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2749185482090435e-05,
+      "loss": 3.7593,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2740799534579915e-05,
+      "loss": 3.7618,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2732413587069394e-05,
+      "loss": 3.7696,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272404401836261e-05,
+      "loss": 3.7455,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271565807085209e-05,
+      "loss": 3.7638,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270727212334157e-05,
+      "loss": 3.7656,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269888617583105e-05,
+      "loss": 3.7628,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269051660712426e-05,
+      "loss": 3.7591,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268213065961374e-05,
+      "loss": 3.7663,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267374471210322e-05,
+      "loss": 3.7567,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26653587645927e-05,
+      "loss": 3.7689,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265698919588591e-05,
+      "loss": 3.766,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264860324837539e-05,
+      "loss": 3.7703,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264021730086487e-05,
+      "loss": 3.7483,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263183135335435e-05,
+      "loss": 3.7645,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2623461784647564e-05,
+      "loss": 3.7675,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2615075837137044e-05,
+      "loss": 3.7568,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2606689889626524e-05,
+      "loss": 3.7614,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2598303942116e-05,
+      "loss": 3.7678,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258993437340921e-05,
+      "loss": 3.7648,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258154842589869e-05,
+      "loss": 3.7768,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257316247838817e-05,
+      "loss": 3.7557,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2564776530877646e-05,
+      "loss": 3.7604,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255640696217086e-05,
+      "loss": 3.7568,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254802101466034e-05,
+      "loss": 3.7599,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253963506714982e-05,
+      "loss": 3.7569,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25312491196393e-05,
+      "loss": 3.7548,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252287955093252e-05,
+      "loss": 3.7608,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2514493603422e-05,
+      "loss": 3.7444,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250610765591147e-05,
+      "loss": 3.7636,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8769750595092773,
+      "eval_runtime": 303.6628,
+      "eval_samples_per_second": 1256.627,
+      "eval_steps_per_second": 39.271,
+      "step": 457920
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 3.127030926167716e+17,
+  "trial_name": null,
+  "trial_params": null
+}