diff --git "a/checkpoint-686872/trainer_state.json" "b/checkpoint-686872/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-686872/trainer_state.json"
@@ -0,0 +1,8143 @@
+{
+  "best_metric": 3.8498642444610596,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/rel-cl/transformer/4/checkpoints/checkpoint-686872",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 686872,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 11.029,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8404,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1927,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9723,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8266,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7094,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6055,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5396,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4816,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4052,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.37,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3294,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2682,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 5.2185,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 5.1914,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 5.1621,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 5.1269,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0824,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0643,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0336,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0079,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 4.991,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.971,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9444,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9248,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.9093,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 4.8848,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8758,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8413,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756873037409844e-05,
+      "loss": 4.844,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.8119,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.8013,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.8054,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 4.7875,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 4.7611,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 4.7497,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 4.7519,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 4.7328,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7225,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.7064,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6896,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6788,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6652,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6668,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.658,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6515,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6365,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6329,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.6361,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 4.6124,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.6073,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.5957,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.6055,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 4.5713,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5885,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5631,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5613,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 4.5631,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 4.5521,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5384,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5301,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5188,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 4.5323,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.5162,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.5061,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5068,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 4.497,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4784,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4989,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9421582546222625e-05,
+      "loss": 4.4788,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4718,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4837,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4629,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.475,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4723,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4627,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.4649,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.4411,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4346,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.4397,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.4399,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.4329,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4182,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 4.4086,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.4121,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.4112,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.4136,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 4.4133,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.4103,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925391273242342e-05,
+      "loss": 4.3964,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 4.3856,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 4.3843,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 4.3899,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922038532118507e-05,
+      "loss": 4.3846,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921199937367455e-05,
+      "loss": 4.368,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920361342616403e-05,
+      "loss": 4.3717,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.3727,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.3754,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.3713,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.3657,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.3689,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.3575,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.3558,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136542224883596e-05,
+      "loss": 4.3629,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.3457,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.352,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.3323,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103014813645245e-05,
+      "loss": 4.3288,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.3283,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.3371,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.333,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3099,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3137,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3214,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.3294,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.3155,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.3216,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901920447495124e-05,
+      "loss": 4.3114,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901081852744072e-05,
+      "loss": 4.3141,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.3047,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.3,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.2944,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.3025,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968888789888125e-05,
+      "loss": 4.2873,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.2973,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.2901,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.2974,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2737,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.2843,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2938,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.273,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2906,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.277,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.2788,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.272,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886832293497681e-05,
+      "loss": 4.2642,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.2681,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.2727,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2771,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.2662,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.2701,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2618,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.2498,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.2565,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.2448,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2476,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877611026996855e-05,
+      "loss": 4.2473,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876772432245803e-05,
+      "loss": 4.2398,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875933837494751e-05,
+      "loss": 4.2388,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8750952427436986e-05,
+      "loss": 4.2388,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.204409122467041,
+      "eval_runtime": 570.6558,
+      "eval_samples_per_second": 668.689,
+      "eval_steps_per_second": 20.897,
+      "step": 76319
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.87425828587302e-05,
+      "loss": 4.2439,
+      "step": 76800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.873419691121968e-05,
+      "loss": 4.2382,
+      "step": 77312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.872581096370916e-05,
+      "loss": 4.2318,
+      "step": 77824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8717425016198635e-05,
+      "loss": 4.2303,
+      "step": 78336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8709039068688115e-05,
+      "loss": 4.229,
+      "step": 78848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8700653121177595e-05,
+      "loss": 4.2065,
+      "step": 79360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8692267173667075e-05,
+      "loss": 4.2219,
+      "step": 79872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8683881226156555e-05,
+      "loss": 4.2209,
+      "step": 80384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.867549527864604e-05,
+      "loss": 4.2114,
+      "step": 80896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.866712570993925e-05,
+      "loss": 4.2214,
+      "step": 81408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865873976242873e-05,
+      "loss": 4.2207,
+      "step": 81920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865035381491821e-05,
+      "loss": 4.2069,
+      "step": 82432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.864196786740769e-05,
+      "loss": 4.2041,
+      "step": 82944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.863358191989717e-05,
+      "loss": 4.2063,
+      "step": 83456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.862519597238665e-05,
+      "loss": 4.2052,
+      "step": 83968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.861681002487613e-05,
+      "loss": 4.196,
+      "step": 84480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860842407736561e-05,
+      "loss": 4.1959,
+      "step": 84992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860003812985509e-05,
+      "loss": 4.1975,
+      "step": 85504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.859165218234457e-05,
+      "loss": 4.1948,
+      "step": 86016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.1986,
+      "step": 86528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.857489666612726e-05,
+      "loss": 4.1964,
+      "step": 87040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.856651071861674e-05,
+      "loss": 4.1982,
+      "step": 87552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8558124771106226e-05,
+      "loss": 4.192,
+      "step": 88064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.85497388235957e-05,
+      "loss": 4.1916,
+      "step": 88576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.854135287608518e-05,
+      "loss": 4.1881,
+      "step": 89088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.853296692857466e-05,
+      "loss": 4.1835,
+      "step": 89600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.852458098106414e-05,
+      "loss": 4.19,
+      "step": 90112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.851619503355362e-05,
+      "loss": 4.1696,
+      "step": 90624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.1848,
+      "step": 91136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.849943951733631e-05,
+      "loss": 4.1681,
+      "step": 91648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.849105356982579e-05,
+      "loss": 4.172,
+      "step": 92160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.848266762231527e-05,
+      "loss": 4.1821,
+      "step": 92672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.847428167480475e-05,
+      "loss": 4.1801,
+      "step": 93184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8465912106097964e-05,
+      "loss": 4.1667,
+      "step": 93696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8457526158587444e-05,
+      "loss": 4.1603,
+      "step": 94208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8449140211076924e-05,
+      "loss": 4.1751,
+      "step": 94720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8440754263566404e-05,
+      "loss": 4.1665,
+      "step": 95232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8432368316055884e-05,
+      "loss": 4.1672,
+      "step": 95744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8423982368545363e-05,
+      "loss": 4.161,
+      "step": 96256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8415596421034843e-05,
+      "loss": 4.1476,
+      "step": 96768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.840721047352432e-05,
+      "loss": 4.1472,
+      "step": 97280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839884090481753e-05,
+      "loss": 4.1481,
+      "step": 97792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839045495730701e-05,
+      "loss": 4.1568,
+      "step": 98304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.838208538860022e-05,
+      "loss": 4.1476,
+      "step": 98816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.83736994410897e-05,
+      "loss": 4.1501,
+      "step": 99328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.836531349357918e-05,
+      "loss": 4.1476,
+      "step": 99840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.835692754606866e-05,
+      "loss": 4.1481,
+      "step": 100352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834854159855814e-05,
+      "loss": 4.1576,
+      "step": 100864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834017202985136e-05,
+      "loss": 4.1364,
+      "step": 101376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.833178608234084e-05,
+      "loss": 4.1401,
+      "step": 101888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.832340013483032e-05,
+      "loss": 4.1323,
+      "step": 102400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.1535,
+      "step": 102912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.1264,
+      "step": 103424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8298258671102486e-05,
+      "loss": 4.1429,
+      "step": 103936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8289872723591966e-05,
+      "loss": 4.124,
+      "step": 104448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8281486776081446e-05,
+      "loss": 4.1348,
+      "step": 104960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8273100828570926e-05,
+      "loss": 4.1357,
+      "step": 105472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8264714881060406e-05,
+      "loss": 4.1331,
+      "step": 105984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8256328933549886e-05,
+      "loss": 4.1225,
+      "step": 106496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8247959364843095e-05,
+      "loss": 4.1246,
+      "step": 107008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.823957341733258e-05,
+      "loss": 4.1114,
+      "step": 107520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.823118746982206e-05,
+      "loss": 4.1297,
+      "step": 108032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8222801522311535e-05,
+      "loss": 4.1224,
+      "step": 108544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.821443195360475e-05,
+      "loss": 4.1149,
+      "step": 109056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.820604600609423e-05,
+      "loss": 4.1166,
+      "step": 109568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.819766005858371e-05,
+      "loss": 4.1172,
+      "step": 110080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8189274111073184e-05,
+      "loss": 4.0991,
+      "step": 110592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.81809045423664e-05,
+      "loss": 4.1235,
+      "step": 111104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.817251859485588e-05,
+      "loss": 4.1051,
+      "step": 111616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.816413264734535e-05,
+      "loss": 4.108,
+      "step": 112128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.815574669983483e-05,
+      "loss": 4.1184,
+      "step": 112640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.814736075232432e-05,
+      "loss": 4.1035,
+      "step": 113152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8138991183617535e-05,
+      "loss": 4.1157,
+      "step": 113664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813060523610701e-05,
+      "loss": 4.1213,
+      "step": 114176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.812221928859649e-05,
+      "loss": 4.1123,
+      "step": 114688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.811383334108597e-05,
+      "loss": 4.1176,
+      "step": 115200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8105463772379184e-05,
+      "loss": 4.0999,
+      "step": 115712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.809707782486866e-05,
+      "loss": 4.0918,
+      "step": 116224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808869187735814e-05,
+      "loss": 4.1019,
+      "step": 116736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808030592984762e-05,
+      "loss": 4.1095,
+      "step": 117248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8071936361140827e-05,
+      "loss": 4.1003,
+      "step": 117760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8063550413630307e-05,
+      "loss": 4.0928,
+      "step": 118272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8055164466119786e-05,
+      "loss": 4.0827,
+      "step": 118784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.804677851860927e-05,
+      "loss": 4.0874,
+      "step": 119296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.803840894990248e-05,
+      "loss": 4.0887,
+      "step": 119808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.803002300239196e-05,
+      "loss": 4.0922,
+      "step": 120320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.802163705488144e-05,
+      "loss": 4.1003,
+      "step": 120832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.801325110737092e-05,
+      "loss": 4.099,
+      "step": 121344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.800488153866413e-05,
+      "loss": 4.087,
+      "step": 121856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.799649559115361e-05,
+      "loss": 4.0837,
+      "step": 122368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.798810964364309e-05,
+      "loss": 4.0785,
+      "step": 122880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797972369613257e-05,
+      "loss": 4.0835,
+      "step": 123392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797135412742578e-05,
+      "loss": 4.0859,
+      "step": 123904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.796296817991526e-05,
+      "loss": 4.078,
+      "step": 124416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.795458223240474e-05,
+      "loss": 4.0754,
+      "step": 124928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.794619628489423e-05,
+      "loss": 4.0838,
+      "step": 125440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7937826716187436e-05,
+      "loss": 4.0864,
+      "step": 125952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7929440768676916e-05,
+      "loss": 4.08,
+      "step": 126464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7921054821166396e-05,
+      "loss": 4.0812,
+      "step": 126976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7912668873655876e-05,
+      "loss": 4.0875,
+      "step": 127488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7904299304949085e-05,
+      "loss": 4.0769,
+      "step": 128000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7895913357438565e-05,
+      "loss": 4.0741,
+      "step": 128512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7887527409928045e-05,
+      "loss": 4.0884,
+      "step": 129024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7879157841221254e-05,
+      "loss": 4.072,
+      "step": 129536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7870771893710734e-05,
+      "loss": 4.0823,
+      "step": 130048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7862385946200214e-05,
+      "loss": 4.0681,
+      "step": 130560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7853999998689694e-05,
+      "loss": 4.0584,
+      "step": 131072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.784561405117918e-05,
+      "loss": 4.066,
+      "step": 131584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.783722810366866e-05,
+      "loss": 4.0724,
+      "step": 132096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782885853496187e-05,
+      "loss": 4.068,
+      "step": 132608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782047258745135e-05,
+      "loss": 4.0535,
+      "step": 133120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.781208663994083e-05,
+      "loss": 4.0514,
+      "step": 133632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.780370069243031e-05,
+      "loss": 4.0653,
+      "step": 134144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.779531474491979e-05,
+      "loss": 4.0726,
+      "step": 134656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.778692879740927e-05,
+      "loss": 4.0607,
+      "step": 135168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777855922870248e-05,
+      "loss": 4.0738,
+      "step": 135680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777017328119196e-05,
+      "loss": 4.0617,
+      "step": 136192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.776178733368144e-05,
+      "loss": 4.0631,
+      "step": 136704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.775340138617092e-05,
+      "loss": 4.0612,
+      "step": 137216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77450154386604e-05,
+      "loss": 4.0544,
+      "step": 137728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.773662949114988e-05,
+      "loss": 4.0496,
+      "step": 138240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.772824354363936e-05,
+      "loss": 4.0605,
+      "step": 138752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7719857596128845e-05,
+      "loss": 4.0495,
+      "step": 139264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7711488027422054e-05,
+      "loss": 4.0516,
+      "step": 139776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7703102079911534e-05,
+      "loss": 4.0555,
+      "step": 140288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7694716132401014e-05,
+      "loss": 4.0632,
+      "step": 140800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7686330184890494e-05,
+      "loss": 4.0425,
+      "step": 141312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.76779606161837e-05,
+      "loss": 4.0454,
+      "step": 141824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.766957466867318e-05,
+      "loss": 4.0645,
+      "step": 142336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.766118872116266e-05,
+      "loss": 4.0426,
+      "step": 142848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7652802773652136e-05,
+      "loss": 4.0625,
+      "step": 143360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.764443320494535e-05,
+      "loss": 4.0456,
+      "step": 143872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.763604725743483e-05,
+      "loss": 4.0541,
+      "step": 144384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.762766130992431e-05,
+      "loss": 4.0463,
+      "step": 144896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761927536241379e-05,
+      "loss": 4.0399,
+      "step": 145408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761090579370701e-05,
+      "loss": 4.0453,
+      "step": 145920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.760251984619649e-05,
+      "loss": 4.0497,
+      "step": 146432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.759413389868596e-05,
+      "loss": 4.06,
+      "step": 146944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.758574795117544e-05,
+      "loss": 4.0497,
+      "step": 147456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.757737838246866e-05,
+      "loss": 4.0504,
+      "step": 147968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7568992434958137e-05,
+      "loss": 4.0442,
+      "step": 148480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756060648744761e-05,
+      "loss": 4.0322,
+      "step": 148992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.755222053993709e-05,
+      "loss": 4.0462,
+      "step": 149504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7543850971230306e-05,
+      "loss": 4.0332,
+      "step": 150016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7535465023719786e-05,
+      "loss": 4.0356,
+      "step": 150528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7527079076209266e-05,
+      "loss": 4.041,
+      "step": 151040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7518693128698745e-05,
+      "loss": 4.0264,
+      "step": 151552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751032355999196e-05,
+      "loss": 4.0344,
+      "step": 152064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7501937612481435e-05,
+      "loss": 4.0326,
+      "step": 152576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0317535400390625,
+      "eval_runtime": 576.2911,
+      "eval_samples_per_second": 662.15,
+      "eval_steps_per_second": 20.693,
+      "step": 152638
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749356804377465e-05,
+      "loss": 4.0391,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748518209626413e-05,
+      "loss": 4.0369,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747679614875361e-05,
+      "loss": 4.031,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468410201243084e-05,
+      "loss": 4.0281,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460024253732563e-05,
+      "loss": 4.0324,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451638306222043e-05,
+      "loss": 4.0124,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744325235871152e-05,
+      "loss": 4.0282,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434866411201e-05,
+      "loss": 4.0237,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742648046369049e-05,
+      "loss": 4.0141,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74181108949837e-05,
+      "loss": 4.0292,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740972494747318e-05,
+      "loss": 4.0306,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740133899996266e-05,
+      "loss": 4.0161,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739295305245214e-05,
+      "loss": 4.0123,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738456710494162e-05,
+      "loss": 4.0177,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73761811574311e-05,
+      "loss": 4.0181,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736779520992058e-05,
+      "loss": 4.0094,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735940926241006e-05,
+      "loss": 4.0083,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735103969370327e-05,
+      "loss": 4.0108,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734265374619275e-05,
+      "loss": 4.0134,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733426779868223e-05,
+      "loss": 4.0159,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732588185117171e-05,
+      "loss": 4.0139,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7317512282464924e-05,
+      "loss": 4.0184,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7309126334954404e-05,
+      "loss": 4.0157,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300740387443884e-05,
+      "loss": 4.0101,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292354439933364e-05,
+      "loss": 4.0126,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728398487122657e-05,
+      "loss": 4.0103,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727559892371605e-05,
+      "loss": 4.0139,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726721297620553e-05,
+      "loss": 3.9983,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725882702869501e-05,
+      "loss": 4.0085,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725045745998822e-05,
+      "loss": 3.9964,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72420715124777e-05,
+      "loss": 4.0009,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723368556496718e-05,
+      "loss": 4.0126,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722529961745666e-05,
+      "loss": 4.0076,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721693004874988e-05,
+      "loss": 3.9989,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720854410123936e-05,
+      "loss": 3.994,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720015815372884e-05,
+      "loss": 4.0111,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719177220621832e-05,
+      "loss": 3.9951,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183402637511526e-05,
+      "loss": 4.0015,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175016690001006e-05,
+      "loss": 3.9963,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7166630742490486e-05,
+      "loss": 3.9836,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158244794979966e-05,
+      "loss": 3.9841,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149858847469446e-05,
+      "loss": 3.9829,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714147289995892e-05,
+      "loss": 4.0007,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71330869524484e-05,
+      "loss": 3.9882,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712470100493788e-05,
+      "loss": 3.9886,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116331436231095e-05,
+      "loss": 3.987,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7107945488720575e-05,
+      "loss": 3.9874,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099559541210055e-05,
+      "loss": 4.0021,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709118997250327e-05,
+      "loss": 3.98,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082804024992744e-05,
+      "loss": 3.9808,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074418077482224e-05,
+      "loss": 3.9784,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706604850877544e-05,
+      "loss": 3.9992,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705766256126492e-05,
+      "loss": 3.9739,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704927661375439e-05,
+      "loss": 3.9897,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704089066624387e-05,
+      "loss": 3.9676,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703250471873335e-05,
+      "loss": 3.9834,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702411877122283e-05,
+      "loss": 3.9875,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701573282371232e-05,
+      "loss": 3.9758,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70073468762018e-05,
+      "loss": 3.9812,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699897730749501e-05,
+      "loss": 3.976,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699059135998449e-05,
+      "loss": 3.959,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698220541247397e-05,
+      "loss": 3.9791,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697381946496345e-05,
+      "loss": 3.9782,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696544989625666e-05,
+      "loss": 3.9702,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695708032754987e-05,
+      "loss": 3.9666,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694869438003935e-05,
+      "loss": 3.9725,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694030843252883e-05,
+      "loss": 3.9555,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931922485018307e-05,
+      "loss": 3.9807,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6923536537507787e-05,
+      "loss": 3.957,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915150589997267e-05,
+      "loss": 3.9655,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690678102129048e-05,
+      "loss": 3.975,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689839507377996e-05,
+      "loss": 3.9647,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689000912626944e-05,
+      "loss": 3.9717,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688162317875892e-05,
+      "loss": 3.9861,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68732372312484e-05,
+      "loss": 3.9699,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686485128373788e-05,
+      "loss": 3.9807,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685646533622736e-05,
+      "loss": 3.9626,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684807938871684e-05,
+      "loss": 3.9519,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683970982001005e-05,
+      "loss": 3.9639,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683132387249953e-05,
+      "loss": 3.9751,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682293792498901e-05,
+      "loss": 3.964,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681455197747849e-05,
+      "loss": 3.9544,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680618240877171e-05,
+      "loss": 3.9526,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679779646126119e-05,
+      "loss": 3.952,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678941051375067e-05,
+      "loss": 3.9514,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678102456624015e-05,
+      "loss": 3.9573,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772654997533356e-05,
+      "loss": 3.966,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764269050022836e-05,
+      "loss": 3.9652,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755883102512316e-05,
+      "loss": 3.9546,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747513533805525e-05,
+      "loss": 3.9523,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739127586295005e-05,
+      "loss": 3.9475,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730741638784485e-05,
+      "loss": 3.9547,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722355691273965e-05,
+      "loss": 3.9508,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6713969743763445e-05,
+      "loss": 3.9518,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705583796252925e-05,
+      "loss": 3.9457,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697197848742405e-05,
+      "loss": 3.957,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688811901231885e-05,
+      "loss": 3.9555,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668045871132883e-05,
+      "loss": 3.9544,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667207276381831e-05,
+      "loss": 3.9546,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666368681630779e-05,
+      "loss": 3.9595,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665530086879727e-05,
+      "loss": 3.9527,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664691492128675e-05,
+      "loss": 3.946,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663852897377623e-05,
+      "loss": 3.9651,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663015940506944e-05,
+      "loss": 3.9425,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662177345755892e-05,
+      "loss": 3.9643,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66133875100484e-05,
+      "loss": 3.9391,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660500156253788e-05,
+      "loss": 3.9347,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596631993831094e-05,
+      "loss": 3.9415,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588246046320574e-05,
+      "loss": 3.952,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579860098810054e-05,
+      "loss": 3.9462,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657147415129953e-05,
+      "loss": 3.934,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656308820378901e-05,
+      "loss": 3.9313,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655470225627849e-05,
+      "loss": 3.9414,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654631630876797e-05,
+      "loss": 3.9525,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653793036125745e-05,
+      "loss": 3.9393,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529560792550656e-05,
+      "loss": 3.9548,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6521174845040136e-05,
+      "loss": 3.9428,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512788897529616e-05,
+      "loss": 3.9427,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6504402950019096e-05,
+      "loss": 3.9483,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649601700250858e-05,
+      "loss": 3.9332,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648764743380179e-05,
+      "loss": 3.9353,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647926148629127e-05,
+      "loss": 3.9403,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647087553878075e-05,
+      "loss": 3.9319,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646248959127023e-05,
+      "loss": 3.9371,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645410364375971e-05,
+      "loss": 3.9382,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644571769624919e-05,
+      "loss": 3.9448,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643733174873867e-05,
+      "loss": 3.9284,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642894580122815e-05,
+      "loss": 3.9284,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642057623252136e-05,
+      "loss": 3.9516,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641220666381457e-05,
+      "loss": 3.9316,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640382071630405e-05,
+      "loss": 3.9442,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395434768793536e-05,
+      "loss": 3.9354,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387048821283016e-05,
+      "loss": 3.9389,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378662873772496e-05,
+      "loss": 3.934,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370276926261976e-05,
+      "loss": 3.9293,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361907357555185e-05,
+      "loss": 3.9355,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353521410044665e-05,
+      "loss": 3.935,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345135462534145e-05,
+      "loss": 3.9472,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336749515023625e-05,
+      "loss": 3.9415,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328379946316834e-05,
+      "loss": 3.9384,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6319993998806314e-05,
+      "loss": 3.9323,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311608051295794e-05,
+      "loss": 3.9246,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303222103785274e-05,
+      "loss": 3.9352,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629485253507848e-05,
+      "loss": 3.9236,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628646658756797e-05,
+      "loss": 3.9282,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627808064005745e-05,
+      "loss": 3.928,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626969469254693e-05,
+      "loss": 3.9206,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626130874503641e-05,
+      "loss": 3.9266,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625292279752589e-05,
+      "loss": 3.9236,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.957977533340454,
+      "eval_runtime": 576.5675,
+      "eval_samples_per_second": 661.832,
+      "eval_steps_per_second": 20.683,
+      "step": 228957
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.62445532288191e-05,
+      "loss": 3.929,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.623616728130858e-05,
+      "loss": 3.9328,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.622778133379806e-05,
+      "loss": 3.9226,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621939538628753e-05,
+      "loss": 3.9227,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621100943877701e-05,
+      "loss": 3.9275,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.620262349126649e-05,
+      "loss": 3.9079,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.619423754375597e-05,
+      "loss": 3.9209,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.618585159624545e-05,
+      "loss": 3.921,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.617748202753867e-05,
+      "loss": 3.9127,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6169112458831884e-05,
+      "loss": 3.9246,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6160726511321364e-05,
+      "loss": 3.9256,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.615234056381084e-05,
+      "loss": 3.9092,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.614395461630032e-05,
+      "loss": 3.91,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.61355686687898e-05,
+      "loss": 3.9146,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.612718272127928e-05,
+      "loss": 3.9173,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611879677376876e-05,
+      "loss": 3.9046,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6110410826258237e-05,
+      "loss": 3.9058,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6102024878747717e-05,
+      "loss": 3.909,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6093655310040926e-05,
+      "loss": 3.9126,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6085269362530406e-05,
+      "loss": 3.9198,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.607688341501989e-05,
+      "loss": 3.91,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606849746750937e-05,
+      "loss": 3.9201,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606011151999885e-05,
+      "loss": 3.9144,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.605174195129206e-05,
+      "loss": 3.9119,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.604335600378154e-05,
+      "loss": 3.9093,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.603497005627102e-05,
+      "loss": 3.9108,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.60265841087605e-05,
+      "loss": 3.9188,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.601819816124998e-05,
+      "loss": 3.9024,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.600981221373946e-05,
+      "loss": 3.9095,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600144264503267e-05,
+      "loss": 3.8993,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599305669752215e-05,
+      "loss": 3.9018,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598467075001163e-05,
+      "loss": 3.9156,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.597628480250111e-05,
+      "loss": 3.9103,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5967915233794326e-05,
+      "loss": 3.9055,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5959529286283806e-05,
+      "loss": 3.8994,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5951143338773286e-05,
+      "loss": 3.9121,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5942757391262766e-05,
+      "loss": 3.8998,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5934387822555975e-05,
+      "loss": 3.9086,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5926001875045455e-05,
+      "loss": 3.9007,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5917615927534935e-05,
+      "loss": 3.8865,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5909229980024415e-05,
+      "loss": 3.8944,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5900844032513895e-05,
+      "loss": 3.8906,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5892474463807104e-05,
+      "loss": 3.9037,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5884088516296584e-05,
+      "loss": 3.8957,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5875702568786064e-05,
+      "loss": 3.8946,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5867316621275544e-05,
+      "loss": 3.8939,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5858930673765024e-05,
+      "loss": 3.8983,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5850544726254504e-05,
+      "loss": 3.9068,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5842158778743984e-05,
+      "loss": 3.8886,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58337892100372e-05,
+      "loss": 3.8883,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.582540326252667e-05,
+      "loss": 3.8925,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.581701731501615e-05,
+      "loss": 3.902,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580863136750563e-05,
+      "loss": 3.885,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580026179879884e-05,
+      "loss": 3.8966,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.579187585128832e-05,
+      "loss": 3.8794,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.57834899037778e-05,
+      "loss": 3.8884,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.577510395626728e-05,
+      "loss": 3.8989,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.576671800875677e-05,
+      "loss": 3.881,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.575834844004998e-05,
+      "loss": 3.8979,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.574996249253946e-05,
+      "loss": 3.8904,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.574157654502894e-05,
+      "loss": 3.8697,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.573319059751842e-05,
+      "loss": 3.8865,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5724821028811626e-05,
+      "loss": 3.8896,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5716435081301106e-05,
+      "loss": 3.8876,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5708049133790586e-05,
+      "loss": 3.8803,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5699663186280066e-05,
+      "loss": 3.8831,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5691293617573275e-05,
+      "loss": 3.8679,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5682907670062755e-05,
+      "loss": 3.8934,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5674521722552235e-05,
+      "loss": 3.871,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5666135775041715e-05,
+      "loss": 3.8799,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.56577498275312e-05,
+      "loss": 3.8894,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564938025882441e-05,
+      "loss": 3.876,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564099431131389e-05,
+      "loss": 3.8842,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.563260836380337e-05,
+      "loss": 3.8949,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.562422241629285e-05,
+      "loss": 3.8865,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561585284758606e-05,
+      "loss": 3.8951,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560746690007554e-05,
+      "loss": 3.8749,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559908095256502e-05,
+      "loss": 3.8699,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.55906950050545e-05,
+      "loss": 3.8747,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.558230905754398e-05,
+      "loss": 3.8942,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.557393948883719e-05,
+      "loss": 3.8781,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.556555354132667e-05,
+      "loss": 3.8712,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5557167593816156e-05,
+      "loss": 3.8653,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5548781646305635e-05,
+      "loss": 3.871,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5540412077598845e-05,
+      "loss": 3.8674,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5532026130088325e-05,
+      "loss": 3.8717,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5523640182577804e-05,
+      "loss": 3.8796,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5515254235067284e-05,
+      "loss": 3.8852,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5506868287556764e-05,
+      "loss": 3.8725,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5498498718849973e-05,
+      "loss": 3.8736,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5490112771339453e-05,
+      "loss": 3.8636,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5481726823828933e-05,
+      "loss": 3.8705,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.547334087631841e-05,
+      "loss": 3.872,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546498768641536e-05,
+      "loss": 3.8696,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.545660173890484e-05,
+      "loss": 3.8643,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544821579139432e-05,
+      "loss": 3.8753,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54398298438838e-05,
+      "loss": 3.8705,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543144389637328e-05,
+      "loss": 3.8829,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.542305794886276e-05,
+      "loss": 3.872,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.541467200135224e-05,
+      "loss": 3.8794,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.540628605384172e-05,
+      "loss": 3.872,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.539791648513493e-05,
+      "loss": 3.8665,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538953053762441e-05,
+      "loss": 3.8852,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538114459011389e-05,
+      "loss": 3.8648,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.537275864260337e-05,
+      "loss": 3.8857,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5364389073896576e-05,
+      "loss": 3.859,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.535600312638606e-05,
+      "loss": 3.8594,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.534761717887554e-05,
+      "loss": 3.8617,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533923123136502e-05,
+      "loss": 3.8759,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533087804146196e-05,
+      "loss": 3.8677,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.532249209395144e-05,
+      "loss": 3.8563,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.531410614644092e-05,
+      "loss": 3.8502,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.53057201989304e-05,
+      "loss": 3.861,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.529733425141988e-05,
+      "loss": 3.8782,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528894830390936e-05,
+      "loss": 3.8612,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528056235639884e-05,
+      "loss": 3.8784,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.527217640888832e-05,
+      "loss": 3.8657,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.52637904613778e-05,
+      "loss": 3.8684,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5255420892671017e-05,
+      "loss": 3.8687,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5247034945160496e-05,
+      "loss": 3.8573,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5238648997649976e-05,
+      "loss": 3.8572,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523026305013945e-05,
+      "loss": 3.8597,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5221893481432665e-05,
+      "loss": 3.8576,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5213507533922145e-05,
+      "loss": 3.8595,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5205121586411625e-05,
+      "loss": 3.8648,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51967356389011e-05,
+      "loss": 3.8646,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5188366070194314e-05,
+      "loss": 3.8544,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5179980122683794e-05,
+      "loss": 3.8523,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5171594175173274e-05,
+      "loss": 3.8751,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5163208227662754e-05,
+      "loss": 3.8545,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515483865895597e-05,
+      "loss": 3.8688,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.514645271144545e-05,
+      "loss": 3.8628,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.513806676393492e-05,
+      "loss": 3.8645,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51296808164244e-05,
+      "loss": 3.8577,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512131124771762e-05,
+      "loss": 3.859,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51129253002071e-05,
+      "loss": 3.8588,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.510453935269657e-05,
+      "loss": 3.8605,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.509615340518605e-05,
+      "loss": 3.8752,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.508778383647927e-05,
+      "loss": 3.8621,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507939788896874e-05,
+      "loss": 3.8664,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507101194145822e-05,
+      "loss": 3.8631,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.506262599394771e-05,
+      "loss": 3.8515,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5054256425240924e-05,
+      "loss": 3.8602,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.50458704777304e-05,
+      "loss": 3.8511,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503748453021988e-05,
+      "loss": 3.8551,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502909858270936e-05,
+      "loss": 3.8536,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502072901400257e-05,
+      "loss": 3.8515,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5012343066492046e-05,
+      "loss": 3.8484,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5003957118981526e-05,
+      "loss": 3.8544,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9165520668029785,
+      "eval_runtime": 561.5936,
+      "eval_samples_per_second": 679.479,
+      "eval_steps_per_second": 21.234,
+      "step": 305276
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995571171471006e-05,
+      "loss": 3.8605,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4987185223960486e-05,
+      "loss": 3.8589,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978799276449966e-05,
+      "loss": 3.8537,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970413328939446e-05,
+      "loss": 3.8502,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4962027381428926e-05,
+      "loss": 3.8588,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953641433918406e-05,
+      "loss": 3.8356,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945255486407886e-05,
+      "loss": 3.8538,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493686953889737e-05,
+      "loss": 3.8467,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492849997019058e-05,
+      "loss": 3.8442,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492013040148379e-05,
+      "loss": 3.8515,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491174445397327e-05,
+      "loss": 3.856,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490335850646275e-05,
+      "loss": 3.8372,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489497255895223e-05,
+      "loss": 3.8443,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488658661144171e-05,
+      "loss": 3.844,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487820066393119e-05,
+      "loss": 3.848,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486981471642067e-05,
+      "loss": 3.8325,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486142876891015e-05,
+      "loss": 3.8362,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485304282139963e-05,
+      "loss": 3.84,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484467325269284e-05,
+      "loss": 3.8426,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4836287305182326e-05,
+      "loss": 3.8523,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4827901357671806e-05,
+      "loss": 3.8399,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4819515410161286e-05,
+      "loss": 3.8532,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4811145841454495e-05,
+      "loss": 3.8484,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4802759893943975e-05,
+      "loss": 3.8458,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4794373946433455e-05,
+      "loss": 3.8389,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4786004377726664e-05,
+      "loss": 3.8416,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4777618430216144e-05,
+      "loss": 3.8503,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4769232482705624e-05,
+      "loss": 3.8374,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4760846535195104e-05,
+      "loss": 3.838,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475246058768458e-05,
+      "loss": 3.8389,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474409101897779e-05,
+      "loss": 3.8287,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473570507146728e-05,
+      "loss": 3.8455,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472731912395676e-05,
+      "loss": 3.848,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471893317644623e-05,
+      "loss": 3.8393,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471056360773945e-05,
+      "loss": 3.8344,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470217766022893e-05,
+      "loss": 3.8413,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469379171271841e-05,
+      "loss": 3.8398,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468542214401162e-05,
+      "loss": 3.8377,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46770361965011e-05,
+      "loss": 3.836,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466865024899058e-05,
+      "loss": 3.8237,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466026430148005e-05,
+      "loss": 3.8285,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465189473277327e-05,
+      "loss": 3.8222,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464350878526275e-05,
+      "loss": 3.8392,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463512283775223e-05,
+      "loss": 3.8282,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626736890241707e-05,
+      "loss": 3.8286,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461836732153492e-05,
+      "loss": 3.8288,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46099813740244e-05,
+      "loss": 3.8309,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460159542651388e-05,
+      "loss": 3.8446,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4593209479003356e-05,
+      "loss": 3.8246,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458483991029657e-05,
+      "loss": 3.8224,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457645396278605e-05,
+      "loss": 3.8307,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4568068015275525e-05,
+      "loss": 3.8324,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559682067765004e-05,
+      "loss": 3.8239,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455131249905822e-05,
+      "loss": 3.8291,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45429265515477e-05,
+      "loss": 3.8203,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453454060403718e-05,
+      "loss": 3.8252,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452615465652666e-05,
+      "loss": 3.8345,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517785087819876e-05,
+      "loss": 3.8168,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509399140309356e-05,
+      "loss": 3.8378,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450101319279883e-05,
+      "loss": 3.8245,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449262724528831e-05,
+      "loss": 3.8095,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4484257676581525e-05,
+      "loss": 3.8213,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475871729071e-05,
+      "loss": 3.8255,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446748578156048e-05,
+      "loss": 3.8308,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445909983404996e-05,
+      "loss": 3.8162,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4450730265343174e-05,
+      "loss": 3.8231,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442344317832654e-05,
+      "loss": 3.8018,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4433958370322134e-05,
+      "loss": 3.8318,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425572422811614e-05,
+      "loss": 3.811,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441720285410482e-05,
+      "loss": 3.8181,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44088169065943e-05,
+      "loss": 3.8258,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440043095908378e-05,
+      "loss": 3.8167,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439204501157326e-05,
+      "loss": 3.8203,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438367544286647e-05,
+      "loss": 3.833,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437528949535595e-05,
+      "loss": 3.8236,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436690354784543e-05,
+      "loss": 3.8298,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435851760033491e-05,
+      "loss": 3.8169,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435014803162813e-05,
+      "loss": 3.8085,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434176208411761e-05,
+      "loss": 3.8168,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433337613660709e-05,
+      "loss": 3.8316,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432499018909657e-05,
+      "loss": 3.8171,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431662062038978e-05,
+      "loss": 3.8121,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308234672879257e-05,
+      "loss": 3.8029,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299848725368737e-05,
+      "loss": 3.8121,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291462777858217e-05,
+      "loss": 3.8048,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4283093209151426e-05,
+      "loss": 3.8169,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274707261640906e-05,
+      "loss": 3.8156,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266321314130386e-05,
+      "loss": 3.8268,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257935366619865e-05,
+      "loss": 3.8136,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424956579791308e-05,
+      "loss": 3.8154,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424117985040256e-05,
+      "loss": 3.8048,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423279390289204e-05,
+      "loss": 3.8111,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422440795538152e-05,
+      "loss": 3.8155,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421603838667473e-05,
+      "loss": 3.808,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420765243916421e-05,
+      "loss": 3.8054,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419926649165369e-05,
+      "loss": 3.8195,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419088054414317e-05,
+      "loss": 3.8074,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418251097543638e-05,
+      "loss": 3.8296,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417412502792586e-05,
+      "loss": 3.809,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416573908041534e-05,
+      "loss": 3.821,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415735313290482e-05,
+      "loss": 3.8094,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4148983564198035e-05,
+      "loss": 3.8088,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4140597616687515e-05,
+      "loss": 3.8273,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4132211669176995e-05,
+      "loss": 3.8075,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123825721666475e-05,
+      "loss": 3.824,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115456152959684e-05,
+      "loss": 3.8024,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4107070205449164e-05,
+      "loss": 3.8054,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4098684257938644e-05,
+      "loss": 3.8014,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4090298310428124e-05,
+      "loss": 3.8153,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408192874172133e-05,
+      "loss": 3.8132,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407354279421081e-05,
+      "loss": 3.8013,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406515684670029e-05,
+      "loss": 3.7906,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405677089918977e-05,
+      "loss": 3.8043,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404838495167925e-05,
+      "loss": 3.8202,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403999900416873e-05,
+      "loss": 3.8017,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403161305665822e-05,
+      "loss": 3.8216,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40232271091477e-05,
+      "loss": 3.8089,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401485754044091e-05,
+      "loss": 3.8106,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400647159293039e-05,
+      "loss": 3.8131,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399808564541987e-05,
+      "loss": 3.7998,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398969969790935e-05,
+      "loss": 3.8001,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398133012920256e-05,
+      "loss": 3.8036,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397294418169204e-05,
+      "loss": 3.8039,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396455823418152e-05,
+      "loss": 3.8045,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3956172286671e-05,
+      "loss": 3.8075,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947802717964206e-05,
+      "loss": 3.8082,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3939416770453686e-05,
+      "loss": 3.7975,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393103082294317e-05,
+      "loss": 3.7989,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392266125423638e-05,
+      "loss": 3.8149,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391427530672586e-05,
+      "loss": 3.7996,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390588935921534e-05,
+      "loss": 3.8141,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389750341170482e-05,
+      "loss": 3.8091,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388913384299803e-05,
+      "loss": 3.806,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388074789548751e-05,
+      "loss": 3.8033,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387236194797699e-05,
+      "loss": 3.8035,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386397600046647e-05,
+      "loss": 3.8039,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385560643175968e-05,
+      "loss": 3.8044,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384722048424916e-05,
+      "loss": 3.8191,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383883453673864e-05,
+      "loss": 3.811,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383044858922813e-05,
+      "loss": 3.8108,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3822079020521336e-05,
+      "loss": 3.8073,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813693073010816e-05,
+      "loss": 3.7986,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805307125500296e-05,
+      "loss": 3.8044,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796921177989776e-05,
+      "loss": 3.7968,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788551609282985e-05,
+      "loss": 3.8004,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3780165661772465e-05,
+      "loss": 3.802,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771779714261945e-05,
+      "loss": 3.7979,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763393766751425e-05,
+      "loss": 3.7918,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3755024198044634e-05,
+      "loss": 3.8024,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.89079213142395,
+      "eval_runtime": 570.9046,
+      "eval_samples_per_second": 668.397,
+      "eval_steps_per_second": 20.888,
+      "step": 381595
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3746638250534114e-05,
+      "loss": 3.8119,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3738252303023594e-05,
+      "loss": 3.8034,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372986635551308e-05,
+      "loss": 3.8012,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3721480408002554e-05,
+      "loss": 3.8007,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3713094460492034e-05,
+      "loss": 3.7992,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3704708512981514e-05,
+      "loss": 3.7869,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3696322565470993e-05,
+      "loss": 3.8004,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3687936617960473e-05,
+      "loss": 3.7881,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367956704925368e-05,
+      "loss": 3.7954,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36711974805469e-05,
+      "loss": 3.8001,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366281153303637e-05,
+      "loss": 3.7999,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365442558552585e-05,
+      "loss": 3.7849,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364603963801533e-05,
+      "loss": 3.7928,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363765369050482e-05,
+      "loss": 3.7893,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36292677429943e-05,
+      "loss": 3.7947,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362088179548378e-05,
+      "loss": 3.782,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361249584797326e-05,
+      "loss": 3.7811,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360412627926647e-05,
+      "loss": 3.7884,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359574033175595e-05,
+      "loss": 3.7914,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358735438424543e-05,
+      "loss": 3.7991,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357896843673491e-05,
+      "loss": 3.7862,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3570598868028116e-05,
+      "loss": 3.8044,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3562212920517596e-05,
+      "loss": 3.7914,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3553826973007076e-05,
+      "loss": 3.7962,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3545441025496556e-05,
+      "loss": 3.786,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353707145678977e-05,
+      "loss": 3.7883,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352868550927925e-05,
+      "loss": 3.8,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352029956176873e-05,
+      "loss": 3.7862,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351191361425821e-05,
+      "loss": 3.7903,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350354404555142e-05,
+      "loss": 3.7883,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34951580980409e-05,
+      "loss": 3.7754,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348677215053038e-05,
+      "loss": 3.7967,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347838620301986e-05,
+      "loss": 3.7925,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347001663431307e-05,
+      "loss": 3.7916,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346163068680255e-05,
+      "loss": 3.7817,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345324473929203e-05,
+      "loss": 3.7932,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344485879178151e-05,
+      "loss": 3.7842,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3436489223074726e-05,
+      "loss": 3.7888,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3428103275564206e-05,
+      "loss": 3.7858,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3419717328053685e-05,
+      "loss": 3.7712,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3411331380543165e-05,
+      "loss": 3.7794,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402961811836375e-05,
+      "loss": 3.7704,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3394575864325854e-05,
+      "loss": 3.793,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3386189916815334e-05,
+      "loss": 3.7807,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377803969304814e-05,
+      "loss": 3.7761,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3369434400598024e-05,
+      "loss": 3.7832,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3361048453087503e-05,
+      "loss": 3.7815,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3352662505576983e-05,
+      "loss": 3.7937,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334427655806646e-05,
+      "loss": 3.7728,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333590698935968e-05,
+      "loss": 3.7743,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332752104184916e-05,
+      "loss": 3.7833,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331913509433864e-05,
+      "loss": 3.7823,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331074914682812e-05,
+      "loss": 3.7744,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330237957812133e-05,
+      "loss": 3.7763,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329399363061081e-05,
+      "loss": 3.7722,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328560768310029e-05,
+      "loss": 3.7783,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327722173558977e-05,
+      "loss": 3.7814,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326885216688298e-05,
+      "loss": 3.7749,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326046621937246e-05,
+      "loss": 3.7872,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325208027186194e-05,
+      "loss": 3.7744,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3243710703155146e-05,
+      "loss": 3.7645,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3235324755644626e-05,
+      "loss": 3.7716,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322693880813411e-05,
+      "loss": 3.7711,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321855286062359e-05,
+      "loss": 3.7909,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32101832919168e-05,
+      "loss": 3.7679,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320179734440628e-05,
+      "loss": 3.7759,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319341139689576e-05,
+      "loss": 3.7587,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318502544938524e-05,
+      "loss": 3.7785,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317665588067845e-05,
+      "loss": 3.7645,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316826993316793e-05,
+      "loss": 3.7679,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315988398565741e-05,
+      "loss": 3.7777,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315149803814689e-05,
+      "loss": 3.7715,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31431284694401e-05,
+      "loss": 3.7722,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313474252192958e-05,
+      "loss": 3.7841,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3126356574419067e-05,
+      "loss": 3.7751,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3117970626908546e-05,
+      "loss": 3.7857,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109601058201756e-05,
+      "loss": 3.7709,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3101215110691236e-05,
+      "loss": 3.7611,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3092829163180715e-05,
+      "loss": 3.7667,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3084443215670195e-05,
+      "loss": 3.7853,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3076073646963405e-05,
+      "loss": 3.7689,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067687699452885e-05,
+      "loss": 3.7672,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3059301751942364e-05,
+      "loss": 3.7549,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3050915804431844e-05,
+      "loss": 3.7684,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042546235725054e-05,
+      "loss": 3.7553,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3034160288214533e-05,
+      "loss": 3.7695,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.302577434070402e-05,
+      "loss": 3.7697,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30173883931935e-05,
+      "loss": 3.7767,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300901882448671e-05,
+      "loss": 3.7684,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300063287697619e-05,
+      "loss": 3.7698,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299224692946567e-05,
+      "loss": 3.7585,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298386098195515e-05,
+      "loss": 3.7643,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297549141324836e-05,
+      "loss": 3.7667,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296710546573784e-05,
+      "loss": 3.7647,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295871951822732e-05,
+      "loss": 3.7578,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29503335707168e-05,
+      "loss": 3.7749,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294196400201001e-05,
+      "loss": 3.7598,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293357805449949e-05,
+      "loss": 3.7856,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29252084857927e-05,
+      "loss": 3.7596,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291682253828218e-05,
+      "loss": 3.7752,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290843659077166e-05,
+      "loss": 3.7634,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290005064326114e-05,
+      "loss": 3.764,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289166469575062e-05,
+      "loss": 3.7839,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28832787482401e-05,
+      "loss": 3.7581,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2874892800729576e-05,
+      "loss": 3.7798,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286652323202279e-05,
+      "loss": 3.7572,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285813728451227e-05,
+      "loss": 3.7585,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284975133700175e-05,
+      "loss": 3.7563,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2841365389491225e-05,
+      "loss": 3.7683,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283299582078444e-05,
+      "loss": 3.7681,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282460987327393e-05,
+      "loss": 3.7599,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28162239257634e-05,
+      "loss": 3.7425,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280783797825288e-05,
+      "loss": 3.7606,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279945203074236e-05,
+      "loss": 3.7735,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279106608323184e-05,
+      "loss": 3.764,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278268013572132e-05,
+      "loss": 3.7729,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27742941882108e-05,
+      "loss": 3.7652,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276592461950401e-05,
+      "loss": 3.7648,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275753867199349e-05,
+      "loss": 3.7688,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274915272448297e-05,
+      "loss": 3.7598,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274076677697245e-05,
+      "loss": 3.7537,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2732397208265665e-05,
+      "loss": 3.7589,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2724011260755145e-05,
+      "loss": 3.7598,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715625313244625e-05,
+      "loss": 3.7612,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2707239365734105e-05,
+      "loss": 3.7612,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269888617583105e-05,
+      "loss": 3.7643,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2690500228320523e-05,
+      "loss": 3.7577,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268211428081e-05,
+      "loss": 3.752,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267372833329948e-05,
+      "loss": 3.7702,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266534238578896e-05,
+      "loss": 3.7553,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265695643827844e-05,
+      "loss": 3.767,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264857049076792e-05,
+      "loss": 3.7693,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26401845432574e-05,
+      "loss": 3.7603,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263181497455062e-05,
+      "loss": 3.7618,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26234290270401e-05,
+      "loss": 3.7603,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261504307952958e-05,
+      "loss": 3.7587,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260667351082279e-05,
+      "loss": 3.7577,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.259828756331227e-05,
+      "loss": 3.7748,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258990161580175e-05,
+      "loss": 3.7715,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258153204709496e-05,
+      "loss": 3.7629,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257314609958444e-05,
+      "loss": 3.7644,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256476015207392e-05,
+      "loss": 3.7567,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25563742045634e-05,
+      "loss": 3.7634,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254798825705288e-05,
+      "loss": 3.7572,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253960230954236e-05,
+      "loss": 3.7513,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253123274083557e-05,
+      "loss": 3.7615,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252284679332505e-05,
+      "loss": 3.7548,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251446084581453e-05,
+      "loss": 3.748,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250607489830401e-05,
+      "loss": 3.7553,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8751182556152344,
+      "eval_runtime": 570.4198,
+      "eval_samples_per_second": 668.965,
+      "eval_steps_per_second": 20.906,
+      "step": 457914
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249768895079349e-05,
+      "loss": 3.7668,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248930300328297e-05,
+      "loss": 3.759,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248091705577245e-05,
+      "loss": 3.7574,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247253110826193e-05,
+      "loss": 3.7576,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246414516075141e-05,
+      "loss": 3.7584,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455759213240885e-05,
+      "loss": 3.7461,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2447373265730365e-05,
+      "loss": 3.7567,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2438987318219845e-05,
+      "loss": 3.7442,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2430617749513054e-05,
+      "loss": 3.7544,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242224818080627e-05,
+      "loss": 3.7571,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241386223329575e-05,
+      "loss": 3.7617,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240547628578524e-05,
+      "loss": 3.7373,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239709033827471e-05,
+      "loss": 3.7542,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238870439076419e-05,
+      "loss": 3.7454,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238031844325367e-05,
+      "loss": 3.7558,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237193249574315e-05,
+      "loss": 3.738,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236354654823263e-05,
+      "loss": 3.7397,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235517697952584e-05,
+      "loss": 3.7498,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346807410819055e-05,
+      "loss": 3.7469,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233842146330853e-05,
+      "loss": 3.7573,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233003551579801e-05,
+      "loss": 3.7476,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232164956828749e-05,
+      "loss": 3.7617,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2313279999580704e-05,
+      "loss": 3.7535,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304894052070184e-05,
+      "loss": 3.7521,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2296508104559664e-05,
+      "loss": 3.744,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2288122157049144e-05,
+      "loss": 3.7494,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2279736209538624e-05,
+      "loss": 3.7579,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2271350262028104e-05,
+      "loss": 3.7482,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2262964314517584e-05,
+      "loss": 3.7451,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254578367007064e-05,
+      "loss": 3.7492,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224620879830027e-05,
+      "loss": 3.7373,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223782285078975e-05,
+      "loss": 3.7548,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222943690327923e-05,
+      "loss": 3.7549,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222105095576871e-05,
+      "loss": 3.7517,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221268138706193e-05,
+      "loss": 3.7405,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220429543955141e-05,
+      "loss": 3.754,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219590949204089e-05,
+      "loss": 3.7422,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218752354453037e-05,
+      "loss": 3.7499,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217915397582358e-05,
+      "loss": 3.746,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217076802831306e-05,
+      "loss": 3.7342,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216238208080254e-05,
+      "loss": 3.7336,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215399613329202e-05,
+      "loss": 3.7272,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2145626564585226e-05,
+      "loss": 3.7586,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2137256995878436e-05,
+      "loss": 3.7387,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128871048367915e-05,
+      "loss": 3.7356,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2120485100857395e-05,
+      "loss": 3.7447,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211209915334688e-05,
+      "loss": 3.7416,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210371320583636e-05,
+      "loss": 3.7499,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209532725832584e-05,
+      "loss": 3.7353,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208694131081532e-05,
+      "loss": 3.7385,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20785553633048e-05,
+      "loss": 3.7368,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207018579459801e-05,
+      "loss": 3.7427,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206181622589122e-05,
+      "loss": 3.7362,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20534302783807e-05,
+      "loss": 3.7326,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204504433087018e-05,
+      "loss": 3.7362,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203665838335966e-05,
+      "loss": 3.7398,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.202827243584914e-05,
+      "loss": 3.7395,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201988648833862e-05,
+      "loss": 3.734,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20115005408281e-05,
+      "loss": 3.7474,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2003130972121316e-05,
+      "loss": 3.7364,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1994761403414525e-05,
+      "loss": 3.7261,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1986375455904005e-05,
+      "loss": 3.7326,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1977989508393485e-05,
+      "loss": 3.7321,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1969603560882965e-05,
+      "loss": 3.7511,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1961217613372445e-05,
+      "loss": 3.7304,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1952848044665654e-05,
+      "loss": 3.7336,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1944462097155134e-05,
+      "loss": 3.72,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1936076149644614e-05,
+      "loss": 3.7379,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1927690202134094e-05,
+      "loss": 3.725,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1919304254623574e-05,
+      "loss": 3.7326,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1910918307113054e-05,
+      "loss": 3.7366,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190254873840627e-05,
+      "loss": 3.7322,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189416279089575e-05,
+      "loss": 3.7375,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188577684338523e-05,
+      "loss": 3.7426,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187739089587471e-05,
+      "loss": 3.7389,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186902132716792e-05,
+      "loss": 3.7431,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18606353796574e-05,
+      "loss": 3.7383,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185224943214688e-05,
+      "loss": 3.7224,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184386348463636e-05,
+      "loss": 3.7287,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183547753712584e-05,
+      "loss": 3.7415,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182710796841905e-05,
+      "loss": 3.7358,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181872202090853e-05,
+      "loss": 3.7315,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181033607339801e-05,
+      "loss": 3.7128,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180195012588749e-05,
+      "loss": 3.7315,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17935805571807e-05,
+      "loss": 3.7175,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178519460967018e-05,
+      "loss": 3.7338,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177680866215966e-05,
+      "loss": 3.7296,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1768422714649136e-05,
+      "loss": 3.7386,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1760036767138616e-05,
+      "loss": 3.7302,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1751650819628096e-05,
+      "loss": 3.7314,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174328125092131e-05,
+      "loss": 3.721,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1734895303410785e-05,
+      "loss": 3.7318,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1726509355900265e-05,
+      "loss": 3.7245,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1718123408389745e-05,
+      "loss": 3.7277,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1709737460879225e-05,
+      "loss": 3.7195,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1701351513368705e-05,
+      "loss": 3.7326,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169298194466192e-05,
+      "loss": 3.723,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16845959971514e-05,
+      "loss": 3.7499,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167621004964088e-05,
+      "loss": 3.7239,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166782410213036e-05,
+      "loss": 3.7379,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165943815461984e-05,
+      "loss": 3.7284,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165105220710932e-05,
+      "loss": 3.7242,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164268263840253e-05,
+      "loss": 3.7431,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163429669089201e-05,
+      "loss": 3.725,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162591074338149e-05,
+      "loss": 3.7406,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161752479587097e-05,
+      "loss": 3.7247,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160913884836045e-05,
+      "loss": 3.7182,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160075290084993e-05,
+      "loss": 3.7195,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1592383332143145e-05,
+      "loss": 3.7324,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1583997384632625e-05,
+      "loss": 3.7305,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1575611437122105e-05,
+      "loss": 3.7274,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1567225489611585e-05,
+      "loss": 3.7024,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558839542101065e-05,
+      "loss": 3.7243,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1550469973394274e-05,
+      "loss": 3.7358,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1542084025883754e-05,
+      "loss": 3.7253,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1533698078373234e-05,
+      "loss": 3.7355,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1525312130862714e-05,
+      "loss": 3.7294,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1516926183352194e-05,
+      "loss": 3.7294,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1508540235841674e-05,
+      "loss": 3.7312,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150015428833115e-05,
+      "loss": 3.7243,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149178471962436e-05,
+      "loss": 3.717,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148339877211384e-05,
+      "loss": 3.7223,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147501282460333e-05,
+      "loss": 3.7248,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14666268770928e-05,
+      "loss": 3.7244,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145824092958228e-05,
+      "loss": 3.7248,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14498713608755e-05,
+      "loss": 3.7286,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144148541336497e-05,
+      "loss": 3.72,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143309946585445e-05,
+      "loss": 3.7166,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142471351834393e-05,
+      "loss": 3.7337,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141632757083341e-05,
+      "loss": 3.7175,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140795800212662e-05,
+      "loss": 3.7336,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13995720546161e-05,
+      "loss": 3.7313,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139118610710558e-05,
+      "loss": 3.7254,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138280015959507e-05,
+      "loss": 3.7277,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137441421208455e-05,
+      "loss": 3.7271,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136602826457403e-05,
+      "loss": 3.7207,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135764231706351e-05,
+      "loss": 3.7208,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1349272748356717e-05,
+      "loss": 3.7396,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340886800846197e-05,
+      "loss": 3.7337,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332517232139406e-05,
+      "loss": 3.7303,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1324131284628886e-05,
+      "loss": 3.7294,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315745337118366e-05,
+      "loss": 3.7198,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1307359389607845e-05,
+      "loss": 3.7251,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1298973442097325e-05,
+      "loss": 3.7196,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1290603873390535e-05,
+      "loss": 3.7136,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128221792588002e-05,
+      "loss": 3.7298,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12738319783695e-05,
+      "loss": 3.7194,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126544603085898e-05,
+      "loss": 3.7127,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125706008334846e-05,
+      "loss": 3.7234,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8647923469543457,
+      "eval_runtime": 311.9235,
+      "eval_samples_per_second": 1223.348,
+      "eval_steps_per_second": 38.231,
+      "step": 534233
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124869051464167e-05,
+      "loss": 3.7242,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124030456713115e-05,
+      "loss": 3.7211,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123191861962063e-05,
+      "loss": 3.7279,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122353267211011e-05,
+      "loss": 3.7213,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121516310340332e-05,
+      "loss": 3.7213,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12067771558928e-05,
+      "loss": 3.7143,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119839120838228e-05,
+      "loss": 3.7211,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119000526087176e-05,
+      "loss": 3.7126,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118163569216497e-05,
+      "loss": 3.7175,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1173249744654455e-05,
+      "loss": 3.7159,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1164863797143935e-05,
+      "loss": 3.7272,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1156477849633415e-05,
+      "loss": 3.7084,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1148091902122895e-05,
+      "loss": 3.7188,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1139705954612375e-05,
+      "loss": 3.7079,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1131336385905584e-05,
+      "loss": 3.7235,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1122950438395064e-05,
+      "loss": 3.7042,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1114564490884544e-05,
+      "loss": 3.7061,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1106178543374024e-05,
+      "loss": 3.7133,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109780897466723e-05,
+      "loss": 3.711,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108942302715671e-05,
+      "loss": 3.7232,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108103707964619e-05,
+      "loss": 3.7116,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107265113213567e-05,
+      "loss": 3.7275,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106428156342889e-05,
+      "loss": 3.7143,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105589561591837e-05,
+      "loss": 3.7189,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104750966840785e-05,
+      "loss": 3.7126,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103912372089733e-05,
+      "loss": 3.7151,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103073777338681e-05,
+      "loss": 3.7198,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102236820468002e-05,
+      "loss": 3.7135,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10139822571695e-05,
+      "loss": 3.7129,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.100559630965898e-05,
+      "loss": 3.7142,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.099721036214846e-05,
+      "loss": 3.702,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098882441463793e-05,
+      "loss": 3.7186,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0980454845931146e-05,
+      "loss": 3.7223,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0972068898420626e-05,
+      "loss": 3.7156,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0963682950910106e-05,
+      "loss": 3.7088,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0955297003399586e-05,
+      "loss": 3.7159,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0946911055889066e-05,
+      "loss": 3.7103,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093854148718228e-05,
+      "loss": 3.7164,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0930155539671755e-05,
+      "loss": 3.7116,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0921769592161235e-05,
+      "loss": 3.7038,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0913383644650715e-05,
+      "loss": 3.7006,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0904997697140195e-05,
+      "loss": 3.6944,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0896628128433404e-05,
+      "loss": 3.7198,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0888242180922884e-05,
+      "loss": 3.7069,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0879856233412364e-05,
+      "loss": 3.7016,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0871470285901844e-05,
+      "loss": 3.7115,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086308433839133e-05,
+      "loss": 3.7076,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085471476968454e-05,
+      "loss": 3.7159,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084632882217402e-05,
+      "loss": 3.7002,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08379428746635e-05,
+      "loss": 3.7063,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082955692715298e-05,
+      "loss": 3.7028,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082117097964246e-05,
+      "loss": 3.7089,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081280141093567e-05,
+      "loss": 3.7077,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080441546342515e-05,
+      "loss": 3.6978,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079602951591463e-05,
+      "loss": 3.7058,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078764356840411e-05,
+      "loss": 3.705,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077927399969732e-05,
+      "loss": 3.7075,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07708880521868e-05,
+      "loss": 3.7025,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0762502104676284e-05,
+      "loss": 3.716,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0754116157165764e-05,
+      "loss": 3.6986,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0745746588458974e-05,
+      "loss": 3.6967,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0737360640948453e-05,
+      "loss": 3.6988,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0728974693437933e-05,
+      "loss": 3.7021,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072058874592741e-05,
+      "loss": 3.7139,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071220279841689e-05,
+      "loss": 3.6967,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07038332297101e-05,
+      "loss": 3.7014,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069544728219958e-05,
+      "loss": 3.6913,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068706133468906e-05,
+      "loss": 3.7057,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067867538717854e-05,
+      "loss": 3.6937,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067030581847175e-05,
+      "loss": 3.6999,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066191987096124e-05,
+      "loss": 3.7027,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065353392345072e-05,
+      "loss": 3.7002,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06451479759402e-05,
+      "loss": 3.7025,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.063676202842968e-05,
+      "loss": 3.7134,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062839245972289e-05,
+      "loss": 3.7034,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062000651221237e-05,
+      "loss": 3.7089,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061162056470185e-05,
+      "loss": 3.7055,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060323461719133e-05,
+      "loss": 3.6943,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0594865048484536e-05,
+      "loss": 3.6966,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0586479100974016e-05,
+      "loss": 3.7098,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0578093153463496e-05,
+      "loss": 3.7049,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0569707205952976e-05,
+      "loss": 3.6967,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056133763724619e-05,
+      "loss": 3.6846,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055295168973567e-05,
+      "loss": 3.6971,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054456574222515e-05,
+      "loss": 3.6863,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053617979471463e-05,
+      "loss": 3.7003,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052781022600784e-05,
+      "loss": 3.6969,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051942427849732e-05,
+      "loss": 3.7062,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05110383309868e-05,
+      "loss": 3.7038,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050265238347628e-05,
+      "loss": 3.6946,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049426643596576e-05,
+      "loss": 3.6936,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048589686725897e-05,
+      "loss": 3.6979,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047751091974845e-05,
+      "loss": 3.6909,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046912497223793e-05,
+      "loss": 3.6973,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046073902472741e-05,
+      "loss": 3.6897,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045235307721689e-05,
+      "loss": 3.7014,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0443983508510105e-05,
+      "loss": 3.694,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0435597560999585e-05,
+      "loss": 3.7193,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0427211613489065e-05,
+      "loss": 3.6878,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041882566597854e-05,
+      "loss": 3.7058,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041043971846802e-05,
+      "loss": 3.6994,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0402070149761234e-05,
+      "loss": 3.6923,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039368420225071e-05,
+      "loss": 3.7068,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038529825474019e-05,
+      "loss": 3.6977,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037691230722967e-05,
+      "loss": 3.7058,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036854273852288e-05,
+      "loss": 3.6941,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036015679101236e-05,
+      "loss": 3.687,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035177084350184e-05,
+      "loss": 3.6894,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034338489599132e-05,
+      "loss": 3.7005,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033501532728454e-05,
+      "loss": 3.7002,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032662937977401e-05,
+      "loss": 3.6933,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031824343226349e-05,
+      "loss": 3.6709,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030985748475297e-05,
+      "loss": 3.694,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030148791604618e-05,
+      "loss": 3.7048,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029310196853566e-05,
+      "loss": 3.6944,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028471602102514e-05,
+      "loss": 3.7058,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027633007351462e-05,
+      "loss": 3.6993,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02679441260041e-05,
+      "loss": 3.6942,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025957455729732e-05,
+      "loss": 3.7036,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02511886097868e-05,
+      "loss": 3.6951,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024280266227628e-05,
+      "loss": 3.6878,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023441671476576e-05,
+      "loss": 3.6902,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022603076725524e-05,
+      "loss": 3.6931,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0217661198548446e-05,
+      "loss": 3.6918,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0209275251037926e-05,
+      "loss": 3.6893,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0200889303527406e-05,
+      "loss": 3.702,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0192503356016886e-05,
+      "loss": 3.6933,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0184133787310095e-05,
+      "loss": 3.6828,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0175747839799575e-05,
+      "loss": 3.7054,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0167361892289055e-05,
+      "loss": 3.6866,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0158975944778535e-05,
+      "loss": 3.7008,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015060637607175e-05,
+      "loss": 3.705,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014222042856123e-05,
+      "loss": 3.6942,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013383448105071e-05,
+      "loss": 3.6933,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012544853354019e-05,
+      "loss": 3.6946,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01170789648334e-05,
+      "loss": 3.691,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010869301732288e-05,
+      "loss": 3.6926,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010030706981236e-05,
+      "loss": 3.7081,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009192112230184e-05,
+      "loss": 3.7008,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008355155359505e-05,
+      "loss": 3.7015,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007516560608453e-05,
+      "loss": 3.697,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006677965857401e-05,
+      "loss": 3.689,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005839371106349e-05,
+      "loss": 3.698,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005000776355297e-05,
+      "loss": 3.6861,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0041638194846184e-05,
+      "loss": 3.6852,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0033252247335664e-05,
+      "loss": 3.6953,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0024866299825144e-05,
+      "loss": 3.693,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0016480352314624e-05,
+      "loss": 3.6793,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000811078360783e-05,
+      "loss": 3.6968,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.858268976211548,
+      "eval_runtime": 314.2337,
+      "eval_samples_per_second": 1214.354,
+      "eval_steps_per_second": 37.949,
+      "step": 610552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999972483609731e-05,
+      "loss": 3.7022,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999133888858679e-05,
+      "loss": 3.6902,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998295294107627e-05,
+      "loss": 3.6963,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997456699356575e-05,
+      "loss": 3.6988,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996618104605523e-05,
+      "loss": 3.7044,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995781147734844e-05,
+      "loss": 3.6969,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994942552983792e-05,
+      "loss": 3.7002,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994103958232741e-05,
+      "loss": 3.6882,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993265363481689e-05,
+      "loss": 3.6848,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99242840661101e-05,
+      "loss": 3.7,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991589811859958e-05,
+      "loss": 3.6948,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990751217108906e-05,
+      "loss": 3.6867,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989912622357854e-05,
+      "loss": 3.695,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989075665487175e-05,
+      "loss": 3.6813,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9882370707361227e-05,
+      "loss": 3.6865,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9873984759850707e-05,
+      "loss": 3.677,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865598812340187e-05,
+      "loss": 3.6825,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9857212864829666e-05,
+      "loss": 3.6912,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848843296122876e-05,
+      "loss": 3.6844,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9840457348612356e-05,
+      "loss": 3.7017,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983207140110184e-05,
+      "loss": 3.6989,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823685453591315e-05,
+      "loss": 3.6972,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981531588488453e-05,
+      "loss": 3.6875,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980692993737401e-05,
+      "loss": 3.691,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979854398986349e-05,
+      "loss": 3.6924,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9790158042352964e-05,
+      "loss": 3.699,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978178847364618e-05,
+      "loss": 3.6832,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977340252613566e-05,
+      "loss": 3.6935,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976501657862514e-05,
+      "loss": 3.6775,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975663063111461e-05,
+      "loss": 3.6899,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974826106240783e-05,
+      "loss": 3.6868,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973987511489731e-05,
+      "loss": 3.6933,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973148916738679e-05,
+      "loss": 3.6965,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972310321987627e-05,
+      "loss": 3.7004,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971471727236575e-05,
+      "loss": 3.6826,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9706347703658965e-05,
+      "loss": 3.6894,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969796175614844e-05,
+      "loss": 3.6945,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968957580863792e-05,
+      "loss": 3.6788,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96811898611274e-05,
+      "loss": 3.6802,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9672820292420614e-05,
+      "loss": 3.6724,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966443434491009e-05,
+      "loss": 3.6773,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965604839739957e-05,
+      "loss": 3.6824,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964766244988905e-05,
+      "loss": 3.691,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963929288118226e-05,
+      "loss": 3.6858,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963090693367174e-05,
+      "loss": 3.6795,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962252098616122e-05,
+      "loss": 3.6829,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96141350386507e-05,
+      "loss": 3.6894,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960576546994391e-05,
+      "loss": 3.6737,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959737952243339e-05,
+      "loss": 3.6606,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958899357492287e-05,
+      "loss": 3.6931,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958060762741235e-05,
+      "loss": 3.6862,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957223805870556e-05,
+      "loss": 3.6931,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956385211119504e-05,
+      "loss": 3.6827,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955546616368452e-05,
+      "loss": 3.6838,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9547080216174e-05,
+      "loss": 3.6754,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9538710647467217e-05,
+      "loss": 3.6756,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9530324699956696e-05,
+      "loss": 3.6879,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9521938752446176e-05,
+      "loss": 3.6711,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9513552804935656e-05,
+      "loss": 3.7021,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9505183236228865e-05,
+      "loss": 3.681,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9496797288718345e-05,
+      "loss": 3.6644,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9488411341207825e-05,
+      "loss": 3.6933,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9480025393697305e-05,
+      "loss": 3.6724,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9471655824990514e-05,
+      "loss": 3.6635,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9463269877479994e-05,
+      "loss": 3.6883,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454883929969474e-05,
+      "loss": 3.6849,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9446497982458954e-05,
+      "loss": 3.6754,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943812841375217e-05,
+      "loss": 3.6748,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942974246624165e-05,
+      "loss": 3.6616,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942135651873113e-05,
+      "loss": 3.6733,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941297057122061e-05,
+      "loss": 3.6854,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940460100251382e-05,
+      "loss": 3.6724,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93962150550033e-05,
+      "loss": 3.681,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938782910749278e-05,
+      "loss": 3.677,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937944315998226e-05,
+      "loss": 3.6892,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937107359127547e-05,
+      "loss": 3.6792,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936268764376495e-05,
+      "loss": 3.6803,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935430169625443e-05,
+      "loss": 3.6757,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934591574874391e-05,
+      "loss": 3.6705,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9337529801233395e-05,
+      "loss": 3.6849,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9329160232526604e-05,
+      "loss": 3.6789,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9320774285016084e-05,
+      "loss": 3.6706,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9312388337505564e-05,
+      "loss": 3.6867,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9304002389995044e-05,
+      "loss": 3.6594,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929563282128825e-05,
+      "loss": 3.6702,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.928724687377773e-05,
+      "loss": 3.6776,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927886092626721e-05,
+      "loss": 3.6779,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927047497875669e-05,
+      "loss": 3.6663,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92621054100499e-05,
+      "loss": 3.6803,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925371946253938e-05,
+      "loss": 3.6685,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924533351502886e-05,
+      "loss": 3.6727,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923694756751835e-05,
+      "loss": 3.6644,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922857799881156e-05,
+      "loss": 3.6721,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922019205130104e-05,
+      "loss": 3.6871,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921180610379052e-05,
+      "loss": 3.6787,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920342015628e-05,
+      "loss": 3.679,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9195050587573206e-05,
+      "loss": 3.6699,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9186664640062686e-05,
+      "loss": 3.6683,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9178278692552166e-05,
+      "loss": 3.6719,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9169892745041646e-05,
+      "loss": 3.6799,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9161523176334855e-05,
+      "loss": 3.6763,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9153137228824335e-05,
+      "loss": 3.677,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9144751281313815e-05,
+      "loss": 3.6718,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91363653338033e-05,
+      "loss": 3.6767,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912797938629278e-05,
+      "loss": 3.6738,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911960981758599e-05,
+      "loss": 3.6713,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911122387007547e-05,
+      "loss": 3.6595,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910283792256495e-05,
+      "loss": 3.6829,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909445197505443e-05,
+      "loss": 3.6654,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908608240634764e-05,
+      "loss": 3.6696,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907769645883712e-05,
+      "loss": 3.6665,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90693105113266e-05,
+      "loss": 3.6671,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906092456381608e-05,
+      "loss": 3.6748,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905255499510929e-05,
+      "loss": 3.6687,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904416904759877e-05,
+      "loss": 3.6788,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9035783100088256e-05,
+      "loss": 3.6833,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9027397152577736e-05,
+      "loss": 3.6804,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9019027583870945e-05,
+      "loss": 3.6748,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9010641636360425e-05,
+      "loss": 3.6695,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9002255688849905e-05,
+      "loss": 3.6664,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8993869741339385e-05,
+      "loss": 3.6715,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8985500172632594e-05,
+      "loss": 3.6748,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8977114225122074e-05,
+      "loss": 3.6628,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8968728277611554e-05,
+      "loss": 3.671,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8960342330101034e-05,
+      "loss": 3.6744,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895197276139424e-05,
+      "loss": 3.6716,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894358681388372e-05,
+      "loss": 3.6707,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89352008663732e-05,
+      "loss": 3.6731,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892681491886269e-05,
+      "loss": 3.6654,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89184453501559e-05,
+      "loss": 3.6757,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891005940264538e-05,
+      "loss": 3.6726,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890167345513486e-05,
+      "loss": 3.6863,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889328750762434e-05,
+      "loss": 3.6592,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888491793891755e-05,
+      "loss": 3.6707,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887653199140703e-05,
+      "loss": 3.6735,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886814604389651e-05,
+      "loss": 3.6713,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885976009638599e-05,
+      "loss": 3.6728,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885137414887546e-05,
+      "loss": 3.6726,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8843004580168676e-05,
+      "loss": 3.6747,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8834618632658156e-05,
+      "loss": 3.6877,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882623268514764e-05,
+      "loss": 3.6675,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8817846737637116e-05,
+      "loss": 3.6747,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880947716893033e-05,
+      "loss": 3.6758,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880109122141981e-05,
+      "loss": 3.6647,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8792705273909285e-05,
+      "loss": 3.6658,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8784319326398765e-05,
+      "loss": 3.6656,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877594975769198e-05,
+      "loss": 3.667,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876756381018146e-05,
+      "loss": 3.6605,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8759177862670934e-05,
+      "loss": 3.6689,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8498642444610596,
+      "eval_runtime": 303.0057,
+      "eval_samples_per_second": 1259.353,
+      "eval_steps_per_second": 39.356,
+      "step": 686872
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 4.686192460937595e+17,
+  "trial_name": null,
+  "trial_params": null
+}