diff --git "a/checkpoint-915840/trainer_state.json" "b/checkpoint-915840/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-915840/trainer_state.json"
@@ -0,0 +1,10849 @@
+{
+  "best_metric": 4.0708794593811035,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/existential-there-quantifier/lstm/4/checkpoints/checkpoint-915840",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 915840,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.8207,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5526,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0612,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9978,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9523,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9164,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7353,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6349,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993292879871958e-05,
+      "loss": 6.5372,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992454285120906e-05,
+      "loss": 6.4491,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.389,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.3339,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.2708,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.2072,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.159,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.1045,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0613,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 6.0227,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.986,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9484,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.9142,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8747,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8528,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.8107,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.793,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 5.7615,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 5.7385,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 5.7165,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6853,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6601,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 5.6473,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 5.6355,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 5.6129,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5888,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 5.5844,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 5.5564,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.5426,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.5091,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.5112,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 5.4898,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966467675120532e-05,
+      "loss": 5.4805,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96562908036948e-05,
+      "loss": 5.4686,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964790485618428e-05,
+      "loss": 5.4613,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963951890867376e-05,
+      "loss": 5.4309,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 5.4245,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 5.4194,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.4028,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.4054,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9597605549924894e-05,
+      "loss": 5.3842,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 5.3631,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 5.3727,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.3514,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 5.3465,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 5.3189,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 5.3297,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 5.3066,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 5.3341,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 5.2953,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 5.2958,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 5.2888,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 5.2691,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948863736869932e-05,
+      "loss": 5.2564,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 5.2567,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 5.2384,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 5.2349,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 5.2508,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 5.232,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 5.2178,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 5.195,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 5.1929,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 5.2004,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 5.1974,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 5.189,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 5.1869,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 5.1788,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 5.1709,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 5.1581,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 5.1483,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 5.1428,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 5.1312,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 5.1383,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 5.1284,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 5.1309,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930422841748654e-05,
+      "loss": 5.1123,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 5.0928,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 5.1066,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 5.1026,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 5.1002,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 5.0864,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925391273242342e-05,
+      "loss": 5.0878,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 5.0741,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 5.0695,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 5.0666,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922038532118507e-05,
+      "loss": 5.0613,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921199937367455e-05,
+      "loss": 5.0657,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920361342616403e-05,
+      "loss": 5.059,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919522747865351e-05,
+      "loss": 5.0564,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918684153114299e-05,
+      "loss": 5.0342,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 5.0306,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 5.029,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 5.0287,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 5.0209,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 5.0135,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 5.0041,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 5.0116,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.9868,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.9936,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103014813645245e-05,
+      "loss": 4.9901,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.9818,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.9765,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.9817,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.9692,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.971,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.9626,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.9535,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.9588,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.9556,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901920447495124e-05,
+      "loss": 4.9494,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901081852744072e-05,
+      "loss": 4.9361,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.9426,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.9439,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.9322,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977291116202375e-05,
+      "loss": 4.9193,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968905168691855e-05,
+      "loss": 4.9288,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.9278,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.9157,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.9097,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.8988,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.9044,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.8949,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.9094,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.8986,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.8797,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888509482999785e-05,
+      "loss": 4.8879,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887670888248733e-05,
+      "loss": 4.8857,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886832293497681e-05,
+      "loss": 4.8811,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.8845,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.8712,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.873,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.8796,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.8623,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.8601,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.8617,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.8468,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.8461,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878447983867534e-05,
+      "loss": 4.8518,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.8371,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.8392,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.8462,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8750952427436986e-05,
+      "loss": 4.8393,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.799422264099121,
+      "eval_runtime": 295.7715,
+      "eval_samples_per_second": 1290.155,
+      "eval_steps_per_second": 40.318,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87425828587302e-05,
+      "loss": 4.8219,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873419691121968e-05,
+      "loss": 4.8218,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872584372131662e-05,
+      "loss": 4.8371,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87174577738061e-05,
+      "loss": 4.8213,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870907182629558e-05,
+      "loss": 4.832,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870068587878506e-05,
+      "loss": 4.8041,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869229993127454e-05,
+      "loss": 4.8236,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683913983764027e-05,
+      "loss": 4.8059,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675528036253507e-05,
+      "loss": 4.8006,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667142088742986e-05,
+      "loss": 4.8069,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865875614123246e-05,
+      "loss": 4.8087,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865037019372194e-05,
+      "loss": 4.8094,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642000625015155e-05,
+      "loss": 4.7923,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633614677504635e-05,
+      "loss": 4.791,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862522872999411e-05,
+      "loss": 4.783,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861684278248359e-05,
+      "loss": 4.7842,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.7828,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860008726626628e-05,
+      "loss": 4.7821,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591701318755764e-05,
+      "loss": 4.7787,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583315371245244e-05,
+      "loss": 4.7981,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574929423734724e-05,
+      "loss": 4.7739,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566543476224204e-05,
+      "loss": 4.7823,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558157528713684e-05,
+      "loss": 4.7685,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549771581203164e-05,
+      "loss": 4.7764,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.7643,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532999686182124e-05,
+      "loss": 4.764,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524613738671604e-05,
+      "loss": 4.7594,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8516227791161084e-05,
+      "loss": 4.7562,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8507841843650564e-05,
+      "loss": 4.7383,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.7541,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849110270623698e-05,
+      "loss": 4.7518,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848271675872646e-05,
+      "loss": 4.7549,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847433081121595e-05,
+      "loss": 4.7442,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846594486370543e-05,
+      "loss": 4.7474,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845755891619491e-05,
+      "loss": 4.7392,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844917296868439e-05,
+      "loss": 4.7355,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84408033999776e-05,
+      "loss": 4.7197,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.7267,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842404788376029e-05,
+      "loss": 4.7266,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841566193624977e-05,
+      "loss": 4.7228,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840727598873925e-05,
+      "loss": 4.7273,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839889004122873e-05,
+      "loss": 4.7262,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839050409371821e-05,
+      "loss": 4.7139,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382118146207687e-05,
+      "loss": 4.72,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373732198697167e-05,
+      "loss": 4.7204,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365346251186646e-05,
+      "loss": 4.7064,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8356960303676126e-05,
+      "loss": 4.7187,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8348574356165606e-05,
+      "loss": 4.7039,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8340188408655086e-05,
+      "loss": 4.6918,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331802461144566e-05,
+      "loss": 4.7119,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323432892437775e-05,
+      "loss": 4.6953,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315046944927255e-05,
+      "loss": 4.6944,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306660997416735e-05,
+      "loss": 4.6834,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298275049906215e-05,
+      "loss": 4.6934,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289889102395695e-05,
+      "loss": 4.6804,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281519533688904e-05,
+      "loss": 4.7121,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273133586178384e-05,
+      "loss": 4.6815,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826474763866787e-05,
+      "loss": 4.6921,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825636169115735e-05,
+      "loss": 4.6893,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824799212245056e-05,
+      "loss": 4.6755,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823960617494004e-05,
+      "loss": 4.6665,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823122022742952e-05,
+      "loss": 4.6787,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222834279919e-05,
+      "loss": 4.6655,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821444833240848e-05,
+      "loss": 4.6638,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820606238489796e-05,
+      "loss": 4.6798,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819769281619117e-05,
+      "loss": 4.6746,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818930686868065e-05,
+      "loss": 4.6598,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818092092117013e-05,
+      "loss": 4.6541,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817253497365961e-05,
+      "loss": 4.6529,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816414902614909e-05,
+      "loss": 4.6649,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155779457442305e-05,
+      "loss": 4.6638,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147393509931785e-05,
+      "loss": 4.662,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8139007562421265e-05,
+      "loss": 4.6623,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8130621614910745e-05,
+      "loss": 4.662,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8122252046203954e-05,
+      "loss": 4.6613,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113866098693434e-05,
+      "loss": 4.6441,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105480151182914e-05,
+      "loss": 4.6484,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8097094203672393e-05,
+      "loss": 4.6472,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80887246349656e-05,
+      "loss": 4.6311,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808033868745508e-05,
+      "loss": 4.6473,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807196911874829e-05,
+      "loss": 4.6442,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806358317123777e-05,
+      "loss": 4.6475,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805519722372726e-05,
+      "loss": 4.6287,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804681127621674e-05,
+      "loss": 4.6191,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803842532870622e-05,
+      "loss": 4.6375,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80300393811957e-05,
+      "loss": 4.6317,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802165343368518e-05,
+      "loss": 4.6388,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801326748617466e-05,
+      "loss": 4.6295,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800489791746787e-05,
+      "loss": 4.632,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799651196995735e-05,
+      "loss": 4.626,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798812602244683e-05,
+      "loss": 4.6175,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79797400749363e-05,
+      "loss": 4.6234,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797135412742578e-05,
+      "loss": 4.6214,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7962984558718996e-05,
+      "loss": 4.628,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954598611208476e-05,
+      "loss": 4.6271,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946212663697956e-05,
+      "loss": 4.6239,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937826716187436e-05,
+      "loss": 4.6118,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792947352628438e-05,
+      "loss": 4.6111,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792108757877386e-05,
+      "loss": 4.6092,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791270163126334e-05,
+      "loss": 4.6179,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790433206255655e-05,
+      "loss": 4.6047,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789594611504603e-05,
+      "loss": 4.604,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788756016753551e-05,
+      "loss": 4.5983,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787917422002499e-05,
+      "loss": 4.6117,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787078827251447e-05,
+      "loss": 4.5861,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786240232500395e-05,
+      "loss": 4.5974,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785401637749343e-05,
+      "loss": 4.6002,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784563042998291e-05,
+      "loss": 4.592,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783724448247239e-05,
+      "loss": 4.5896,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782885853496187e-05,
+      "loss": 4.6012,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782047258745135e-05,
+      "loss": 4.5868,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781208663994083e-05,
+      "loss": 4.5885,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780370069243031e-05,
+      "loss": 4.595,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779531474491979e-05,
+      "loss": 4.5808,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778692879740927e-05,
+      "loss": 4.5913,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777854284989875e-05,
+      "loss": 4.5959,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777017328119196e-05,
+      "loss": 4.5871,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776178733368144e-05,
+      "loss": 4.5757,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775340138617092e-05,
+      "loss": 4.5863,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7745031817464134e-05,
+      "loss": 4.5848,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7736645869953614e-05,
+      "loss": 4.575,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7728259922443094e-05,
+      "loss": 4.5728,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719873974932574e-05,
+      "loss": 4.5759,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771150440622578e-05,
+      "loss": 4.5863,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770311845871526e-05,
+      "loss": 4.5743,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769473251120474e-05,
+      "loss": 4.5702,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768634656369422e-05,
+      "loss": 4.5606,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767797699498743e-05,
+      "loss": 4.5676,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766960742628064e-05,
+      "loss": 4.5603,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766122147877012e-05,
+      "loss": 4.5779,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76528355312596e-05,
+      "loss": 4.5716,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764444958374909e-05,
+      "loss": 4.5538,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763606363623857e-05,
+      "loss": 4.5601,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762767768872805e-05,
+      "loss": 4.5664,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761929174121753e-05,
+      "loss": 4.5628,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761090579370701e-05,
+      "loss": 4.5687,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760251984619649e-05,
+      "loss": 4.5555,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75941502774897e-05,
+      "loss": 4.5572,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758576432997918e-05,
+      "loss": 4.5652,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757737838246866e-05,
+      "loss": 4.5521,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7568992434958137e-05,
+      "loss": 4.5536,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756060648744761e-05,
+      "loss": 4.5588,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755222053993709e-05,
+      "loss": 4.5431,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543850971230306e-05,
+      "loss": 4.5473,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535465023719786e-05,
+      "loss": 4.554,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527079076209266e-05,
+      "loss": 4.5381,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518693128698745e-05,
+      "loss": 4.5445,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510307181188225e-05,
+      "loss": 4.5499,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7501921233677705e-05,
+      "loss": 4.545,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.51077938079834,
+      "eval_runtime": 296.1256,
+      "eval_samples_per_second": 1288.612,
+      "eval_steps_per_second": 40.27,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493535286167185e-05,
+      "loss": 4.5317,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485149338656665e-05,
+      "loss": 4.5335,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476763391146145e-05,
+      "loss": 4.5516,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468377443635625e-05,
+      "loss": 4.5319,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7459991496125105e-05,
+      "loss": 4.5478,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451605548614585e-05,
+      "loss": 4.5273,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443219601104065e-05,
+      "loss": 4.5422,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434833653593545e-05,
+      "loss": 4.5257,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426447706083025e-05,
+      "loss": 4.5349,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418061758572505e-05,
+      "loss": 4.5307,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409675811061985e-05,
+      "loss": 4.5336,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401289863551465e-05,
+      "loss": 4.5395,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739292029484468e-05,
+      "loss": 4.5194,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384534347334154e-05,
+      "loss": 4.5241,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376148399823634e-05,
+      "loss": 4.5182,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7367762452313114e-05,
+      "loss": 4.5163,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359376504802594e-05,
+      "loss": 4.5198,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350990557292074e-05,
+      "loss": 4.5236,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342604609781554e-05,
+      "loss": 4.5175,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7334218662271034e-05,
+      "loss": 4.5393,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7325832714760514e-05,
+      "loss": 4.5188,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731746314605372e-05,
+      "loss": 4.5246,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73090771985432e-05,
+      "loss": 4.5168,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730069125103269e-05,
+      "loss": 4.5255,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729230530352217e-05,
+      "loss": 4.516,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728391935601165e-05,
+      "loss": 4.5102,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727553340850113e-05,
+      "loss": 4.5144,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726714746099061e-05,
+      "loss": 4.5079,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725876151348009e-05,
+      "loss": 4.492,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72503919447733e-05,
+      "loss": 4.5085,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724200599726278e-05,
+      "loss": 4.5147,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723362004975226e-05,
+      "loss": 4.5094,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722523410224174e-05,
+      "loss": 4.5059,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721686453353495e-05,
+      "loss": 4.5083,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208494964828156e-05,
+      "loss": 4.502,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720010901731764e-05,
+      "loss": 4.4971,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719172306980712e-05,
+      "loss": 4.4914,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71833371222966e-05,
+      "loss": 4.491,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717495117478608e-05,
+      "loss": 4.4923,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716658160607929e-05,
+      "loss": 4.4927,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715819565856877e-05,
+      "loss": 4.5006,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714980971105825e-05,
+      "loss": 4.4996,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714142376354773e-05,
+      "loss": 4.4878,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713303781603721e-05,
+      "loss": 4.4938,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124651868526685e-05,
+      "loss": 4.4986,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116265921016165e-05,
+      "loss": 4.481,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7107879973505645e-05,
+      "loss": 4.4947,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709951040479886e-05,
+      "loss": 4.4869,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709112445728834e-05,
+      "loss": 4.4691,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708273850977782e-05,
+      "loss": 4.4891,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074368941071037e-05,
+      "loss": 4.4825,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7065982993560517e-05,
+      "loss": 4.4762,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705759704604999e-05,
+      "loss": 4.4726,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704921109853947e-05,
+      "loss": 4.4774,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040841529832686e-05,
+      "loss": 4.4642,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703245558232216e-05,
+      "loss": 4.493,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702406963481164e-05,
+      "loss": 4.4736,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701568368730112e-05,
+      "loss": 4.4812,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70072977397906e-05,
+      "loss": 4.4824,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699891179228008e-05,
+      "loss": 4.4666,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699052584476956e-05,
+      "loss": 4.459,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982139897259045e-05,
+      "loss": 4.4757,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973753949748525e-05,
+      "loss": 4.4626,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965384381041734e-05,
+      "loss": 4.4569,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956998433531214e-05,
+      "loss": 4.4763,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6948612486020694e-05,
+      "loss": 4.4744,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6940226538510174e-05,
+      "loss": 4.4593,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931840590999654e-05,
+      "loss": 4.4525,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6923454643489134e-05,
+      "loss": 4.4538,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915068695978614e-05,
+      "loss": 4.464,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906682748468094e-05,
+      "loss": 4.4673,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68983131797613e-05,
+      "loss": 4.4641,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688992723225078e-05,
+      "loss": 4.4672,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688154128474026e-05,
+      "loss": 4.4666,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687315533722974e-05,
+      "loss": 4.4672,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686480214732669e-05,
+      "loss": 4.4526,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685641619981617e-05,
+      "loss": 4.4574,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684803025230565e-05,
+      "loss": 4.46,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683964430479513e-05,
+      "loss": 4.4398,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683127473608834e-05,
+      "loss": 4.4563,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682288878857782e-05,
+      "loss": 4.4604,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68145028410673e-05,
+      "loss": 4.4588,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680611689355678e-05,
+      "loss": 4.4444,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679773094604626e-05,
+      "loss": 4.4319,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789361377339466e-05,
+      "loss": 4.4518,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678097542982895e-05,
+      "loss": 4.4431,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677258948231843e-05,
+      "loss": 4.459,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676420353480791e-05,
+      "loss": 4.4469,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675583396610112e-05,
+      "loss": 4.4511,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.67474480185906e-05,
+      "loss": 4.4466,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673906207108008e-05,
+      "loss": 4.4341,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673067612356956e-05,
+      "loss": 4.4437,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672229017605904e-05,
+      "loss": 4.4402,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671390422854852e-05,
+      "loss": 4.4522,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705518281037995e-05,
+      "loss": 4.4542,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697132333527474e-05,
+      "loss": 4.4461,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668877914362442e-05,
+      "loss": 4.4318,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680393196113906e-05,
+      "loss": 4.4381,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6672007248603386e-05,
+      "loss": 4.4364,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663621301092866e-05,
+      "loss": 4.4444,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655235353582346e-05,
+      "loss": 4.4318,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664684940607182e-05,
+      "loss": 4.4352,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66384634585613e-05,
+      "loss": 4.4249,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663007751105078e-05,
+      "loss": 4.4439,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621707942343995e-05,
+      "loss": 4.4166,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661332199483347e-05,
+      "loss": 4.4321,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604952426126684e-05,
+      "loss": 4.4229,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596566478616164e-05,
+      "loss": 4.4278,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588180531105644e-05,
+      "loss": 4.4241,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579794583595124e-05,
+      "loss": 4.4338,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571408636084604e-05,
+      "loss": 4.4272,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563022688574084e-05,
+      "loss": 4.4203,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554636741063564e-05,
+      "loss": 4.43,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654626717235677e-05,
+      "loss": 4.4137,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653788122484625e-05,
+      "loss": 4.4331,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652949527733573e-05,
+      "loss": 4.4359,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652110932982521e-05,
+      "loss": 4.4217,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651272338231469e-05,
+      "loss": 4.418,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650433743480417e-05,
+      "loss": 4.4215,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649595148729365e-05,
+      "loss": 4.4238,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648756553978313e-05,
+      "loss": 4.4157,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647919597107634e-05,
+      "loss": 4.4147,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647081002356582e-05,
+      "loss": 4.4135,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646242407605531e-05,
+      "loss": 4.4294,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645403812854479e-05,
+      "loss": 4.4187,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445668559838e-05,
+      "loss": 4.4136,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643728261232748e-05,
+      "loss": 4.4039,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6428913043620687e-05,
+      "loss": 4.4146,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420527096110166e-05,
+      "loss": 4.4041,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412141148599646e-05,
+      "loss": 4.4242,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403755201089126e-05,
+      "loss": 4.4192,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395369253578606e-05,
+      "loss": 4.4064,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6386983306068086e-05,
+      "loss": 4.4034,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378597358557566e-05,
+      "loss": 4.4133,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370227789850775e-05,
+      "loss": 4.4097,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636184184234026e-05,
+      "loss": 4.4212,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635345589482974e-05,
+      "loss": 4.4026,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634506994731922e-05,
+      "loss": 4.4128,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63366839998087e-05,
+      "loss": 4.4107,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632829805229818e-05,
+      "loss": 4.4083,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6319912104787655e-05,
+      "loss": 4.4065,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311526157277135e-05,
+      "loss": 4.4099,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630315658857035e-05,
+      "loss": 4.396,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629477064105983e-05,
+      "loss": 4.4043,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286384693549304e-05,
+      "loss": 4.4064,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277998746038784e-05,
+      "loss": 4.3988,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269629177332e-05,
+      "loss": 4.3956,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626124322982148e-05,
+      "loss": 4.4035,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625285728231096e-05,
+      "loss": 4.4018,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.373067855834961,
+      "eval_runtime": 292.1016,
+      "eval_samples_per_second": 1306.364,
+      "eval_steps_per_second": 40.825,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624447133480044e-05,
+      "loss": 4.3954,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623608538728992e-05,
+      "loss": 4.3865,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.62276994397794e-05,
+      "loss": 4.4115,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621932987107261e-05,
+      "loss": 4.3908,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621094392356209e-05,
+      "loss": 4.4104,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620255797605157e-05,
+      "loss": 4.3828,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619417202854105e-05,
+      "loss": 4.4047,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185818838637994e-05,
+      "loss": 4.3897,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177432891127474e-05,
+      "loss": 4.3931,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616906332242068e-05,
+      "loss": 4.3933,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616067737491017e-05,
+      "loss": 4.3947,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615229142739965e-05,
+      "loss": 4.4016,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614392185869286e-05,
+      "loss": 4.3859,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613553591118234e-05,
+      "loss": 4.3842,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612714996367182e-05,
+      "loss": 4.3895,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61187640161613e-05,
+      "loss": 4.3743,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611037806865078e-05,
+      "loss": 4.3866,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610199212114025e-05,
+      "loss": 4.3878,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609360617362973e-05,
+      "loss": 4.3805,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608522022611921e-05,
+      "loss": 4.4037,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607683427860869e-05,
+      "loss": 4.39,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606844833109817e-05,
+      "loss": 4.3946,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606006238358765e-05,
+      "loss": 4.3838,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605167643607714e-05,
+      "loss": 4.39,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604329048856662e-05,
+      "loss": 4.3832,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603492091985983e-05,
+      "loss": 4.3807,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602653497234931e-05,
+      "loss": 4.3842,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601814902483879e-05,
+      "loss": 4.3734,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600976307732827e-05,
+      "loss": 4.3701,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600137712981775e-05,
+      "loss": 4.3734,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599299118230723e-05,
+      "loss": 4.3849,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598460523479671e-05,
+      "loss": 4.3831,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597621928728619e-05,
+      "loss": 4.3803,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967849718579396e-05,
+      "loss": 4.3835,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959480149872605e-05,
+      "loss": 4.3671,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595109420236209e-05,
+      "loss": 4.3727,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594270825485157e-05,
+      "loss": 4.3707,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593432230734105e-05,
+      "loss": 4.3607,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592593635983053e-05,
+      "loss": 4.3674,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591755041232001e-05,
+      "loss": 4.3639,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590916446480949e-05,
+      "loss": 4.3763,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59007948961027e-05,
+      "loss": 4.3753,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589240894859218e-05,
+      "loss": 4.3667,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588402300108166e-05,
+      "loss": 4.3648,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587563705357114e-05,
+      "loss": 4.3778,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867251106060614e-05,
+      "loss": 4.3575,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858865158550094e-05,
+      "loss": 4.3723,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5850479211039573e-05,
+      "loss": 4.3639,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5842093263529053e-05,
+      "loss": 4.3481,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583370731601854e-05,
+      "loss": 4.371,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582532136850802e-05,
+      "loss": 4.3605,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581695179980123e-05,
+      "loss": 4.3595,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580856585229071e-05,
+      "loss": 4.3477,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580017990478019e-05,
+      "loss": 4.3573,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579179395726967e-05,
+      "loss": 4.3479,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578340800975915e-05,
+      "loss": 4.3675,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577503844105236e-05,
+      "loss": 4.3558,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576665249354184e-05,
+      "loss": 4.3603,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575826654603132e-05,
+      "loss": 4.3636,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57498805985208e-05,
+      "loss": 4.3508,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574149465101028e-05,
+      "loss": 4.3396,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573310870349976e-05,
+      "loss": 4.3617,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572472275598924e-05,
+      "loss": 4.3456,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716336808478725e-05,
+      "loss": 4.3419,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570798361857566e-05,
+      "loss": 4.3597,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569959767106514e-05,
+      "loss": 4.3565,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569121172355462e-05,
+      "loss": 4.3449,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56828257760441e-05,
+      "loss": 4.3424,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567443982853358e-05,
+      "loss": 4.3347,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566605388102306e-05,
+      "loss": 4.346,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565768431231627e-05,
+      "loss": 4.3558,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564929836480575e-05,
+      "loss": 4.3556,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564091241729523e-05,
+      "loss": 4.3523,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563252646978471e-05,
+      "loss": 4.349,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562414052227419e-05,
+      "loss": 4.3614,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561575457476368e-05,
+      "loss": 4.3392,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560736862725315e-05,
+      "loss": 4.3488,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559898267974263e-05,
+      "loss": 4.3458,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5590629489839576e-05,
+      "loss": 4.3257,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582243542329056e-05,
+      "loss": 4.3517,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5573857594818536e-05,
+      "loss": 4.3484,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5565471647308016e-05,
+      "loss": 4.3457,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557085699797496e-05,
+      "loss": 4.3369,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548716131090705e-05,
+      "loss": 4.3264,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5540330183580185e-05,
+      "loss": 4.3413,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5531944236069665e-05,
+      "loss": 4.3394,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523558288559145e-05,
+      "loss": 4.3455,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551518871985236e-05,
+      "loss": 4.3356,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550680277234184e-05,
+      "loss": 4.3471,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549841682483132e-05,
+      "loss": 4.3318,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54900308773208e-05,
+      "loss": 4.3323,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481644929810274e-05,
+      "loss": 4.3328,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5473258982299754e-05,
+      "loss": 4.3296,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546488941359297e-05,
+      "loss": 4.3484,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545650346608244e-05,
+      "loss": 4.3443,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544811751857192e-05,
+      "loss": 4.3456,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54397315710614e-05,
+      "loss": 4.3248,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543134562355088e-05,
+      "loss": 4.3328,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542295967604037e-05,
+      "loss": 4.3329,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541459010733358e-05,
+      "loss": 4.3374,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540620415982306e-05,
+      "loss": 4.329,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539781821231254e-05,
+      "loss": 4.3271,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538943226480202e-05,
+      "loss": 4.3202,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53810463172915e-05,
+      "loss": 4.341,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537266036978098e-05,
+      "loss": 4.312,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536427442227046e-05,
+      "loss": 4.3272,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535590485356367e-05,
+      "loss": 4.3229,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534751890605315e-05,
+      "loss": 4.3255,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533913295854263e-05,
+      "loss": 4.3215,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533074701103211e-05,
+      "loss": 4.3354,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532236106352159e-05,
+      "loss": 4.3241,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531397511601107e-05,
+      "loss": 4.3164,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530558916850055e-05,
+      "loss": 4.3315,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297203220990034e-05,
+      "loss": 4.3158,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528883365228324e-05,
+      "loss": 4.3296,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528044770477272e-05,
+      "loss": 4.3363,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52720617572622e-05,
+      "loss": 4.3247,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526369218855541e-05,
+      "loss": 4.3186,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525530624104489e-05,
+      "loss": 4.3181,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524692029353437e-05,
+      "loss": 4.3229,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523853434602385e-05,
+      "loss": 4.3191,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523016477731706e-05,
+      "loss": 4.3223,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522177882980654e-05,
+      "loss": 4.3097,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521339288229602e-05,
+      "loss": 4.3295,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52050069347855e-05,
+      "loss": 4.3243,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519662098727499e-05,
+      "loss": 4.3187,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518823503976446e-05,
+      "loss": 4.3035,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517986547105768e-05,
+      "loss": 4.3192,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517147952354716e-05,
+      "loss": 4.3091,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516309357603663e-05,
+      "loss": 4.323,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515470762852611e-05,
+      "loss": 4.3271,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514632168101559e-05,
+      "loss": 4.316,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513793573350507e-05,
+      "loss": 4.2994,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512954978599455e-05,
+      "loss": 4.3174,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512118021728776e-05,
+      "loss": 4.3128,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511279426977724e-05,
+      "loss": 4.3282,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104408322266726e-05,
+      "loss": 4.305,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096022374756206e-05,
+      "loss": 4.3181,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087636427245685e-05,
+      "loss": 4.3173,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079266858538895e-05,
+      "loss": 4.3189,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070880911028375e-05,
+      "loss": 4.3126,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062494963517855e-05,
+      "loss": 4.3143,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5054109016007334e-05,
+      "loss": 4.3022,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5045723068496814e-05,
+      "loss": 4.3137,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5037337120986294e-05,
+      "loss": 4.3108,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5028951173475774e-05,
+      "loss": 4.3092,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020565225965254e-05,
+      "loss": 4.3017,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501221203606219e-05,
+      "loss": 4.3137,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500382608855168e-05,
+      "loss": 4.307,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.287594318389893,
+      "eval_runtime": 291.8841,
+      "eval_samples_per_second": 1307.337,
+      "eval_steps_per_second": 40.855,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499544014104116e-05,
+      "loss": 4.2988,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498705419353064e-05,
+      "loss": 4.2985,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497866824602012e-05,
+      "loss": 4.3173,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49702822985096e-05,
+      "loss": 4.3023,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496189635099908e-05,
+      "loss": 4.3147,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495351040348856e-05,
+      "loss": 4.3015,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494512445597804e-05,
+      "loss": 4.3081,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493673850846752e-05,
+      "loss": 4.3027,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492835256095699e-05,
+      "loss": 4.3009,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491996661344647e-05,
+      "loss": 4.3047,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491158066593595e-05,
+      "loss": 4.3064,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490319471842543e-05,
+      "loss": 4.3111,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489482514971865e-05,
+      "loss": 4.2976,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488643920220813e-05,
+      "loss": 4.2959,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487805325469761e-05,
+      "loss": 4.2984,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486966730718709e-05,
+      "loss": 4.2887,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486128135967657e-05,
+      "loss": 4.2969,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485289541216605e-05,
+      "loss": 4.3003,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484450946465553e-05,
+      "loss": 4.2943,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483612351714501e-05,
+      "loss": 4.3157,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482773756963449e-05,
+      "loss": 4.3019,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481935162212397e-05,
+      "loss": 4.3085,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481096567461345e-05,
+      "loss": 4.2984,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480257972710293e-05,
+      "loss": 4.3035,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4794210158396136e-05,
+      "loss": 4.2944,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4785824210885616e-05,
+      "loss": 4.298,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4777438263375096e-05,
+      "loss": 4.296,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476906869466831e-05,
+      "loss": 4.2899,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476068274715779e-05,
+      "loss": 4.2861,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475229679964727e-05,
+      "loss": 4.2879,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474391085213675e-05,
+      "loss": 4.2958,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473554128342996e-05,
+      "loss": 4.2999,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472715533591944e-05,
+      "loss": 4.2939,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471876938840892e-05,
+      "loss": 4.2968,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47103834408984e-05,
+      "loss": 4.2845,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470199749338788e-05,
+      "loss": 4.2892,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469362792468109e-05,
+      "loss": 4.2882,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468524197717057e-05,
+      "loss": 4.2752,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467685602966005e-05,
+      "loss": 4.2882,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466847008214953e-05,
+      "loss": 4.2789,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466008413463901e-05,
+      "loss": 4.2933,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465169818712849e-05,
+      "loss": 4.2925,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464331223961797e-05,
+      "loss": 4.2857,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463494267091118e-05,
+      "loss": 4.2797,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462655672340066e-05,
+      "loss": 4.2974,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461817077589014e-05,
+      "loss": 4.2761,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460978482837962e-05,
+      "loss": 4.2911,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46013988808691e-05,
+      "loss": 4.2803,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459301293335858e-05,
+      "loss": 4.2662,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458464336465179e-05,
+      "loss": 4.2916,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457625741714127e-05,
+      "loss": 4.2793,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456787146963075e-05,
+      "loss": 4.2769,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559485522120234e-05,
+      "loss": 4.2718,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551099574609714e-05,
+      "loss": 4.272,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542713627099194e-05,
+      "loss": 4.273,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534327679588674e-05,
+      "loss": 4.2815,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4525941732078154e-05,
+      "loss": 4.2788,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451757216337136e-05,
+      "loss": 4.2799,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450918621586084e-05,
+      "loss": 4.2853,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450080026835032e-05,
+      "loss": 4.2726,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44924143208398e-05,
+      "loss": 4.2608,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448404475213301e-05,
+      "loss": 4.2826,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447565880462249e-05,
+      "loss": 4.2667,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446727285711197e-05,
+      "loss": 4.2599,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445888690960145e-05,
+      "loss": 4.2835,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445050096209093e-05,
+      "loss": 4.2772,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444211501458041e-05,
+      "loss": 4.2683,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44337290670699e-05,
+      "loss": 4.2661,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442535949836311e-05,
+      "loss": 4.2544,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441697355085259e-05,
+      "loss": 4.269,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440858760334207e-05,
+      "loss": 4.279,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440020165583154e-05,
+      "loss": 4.2739,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439181570832102e-05,
+      "loss": 4.2745,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43834297608105e-05,
+      "loss": 4.2717,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437504381329998e-05,
+      "loss": 4.2873,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436665786578946e-05,
+      "loss": 4.2618,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435828829708267e-05,
+      "loss": 4.2741,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434990234957216e-05,
+      "loss": 4.2715,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434151640206164e-05,
+      "loss": 4.2452,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333130454551117e-05,
+      "loss": 4.2806,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4324744507040597e-05,
+      "loss": 4.2706,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316358559530076e-05,
+      "loss": 4.2722,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4307972612019556e-05,
+      "loss": 4.2645,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299586664509036e-05,
+      "loss": 4.2486,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291217095802245e-05,
+      "loss": 4.2669,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4282831148291725e-05,
+      "loss": 4.2634,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274461579584935e-05,
+      "loss": 4.2726,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266075632074415e-05,
+      "loss": 4.2594,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257689684563894e-05,
+      "loss": 4.27,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249303737053374e-05,
+      "loss": 4.2575,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424093416834659e-05,
+      "loss": 4.2612,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423254822083607e-05,
+      "loss": 4.2586,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422417865212928e-05,
+      "loss": 4.2568,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421579270461876e-05,
+      "loss": 4.2747,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420740675710824e-05,
+      "loss": 4.2687,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419902080959772e-05,
+      "loss": 4.2762,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41906348620872e-05,
+      "loss": 4.2491,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418224891457668e-05,
+      "loss": 4.2597,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417386296706616e-05,
+      "loss": 4.2591,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416547701955564e-05,
+      "loss": 4.263,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415710745084885e-05,
+      "loss": 4.259,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414872150333833e-05,
+      "loss": 4.2521,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414033555582781e-05,
+      "loss": 4.2525,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413194960831729e-05,
+      "loss": 4.2691,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123563660806775e-05,
+      "loss": 4.2415,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115177713296255e-05,
+      "loss": 4.254,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4106791765785735e-05,
+      "loss": 4.2493,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4098422197078944e-05,
+      "loss": 4.2585,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4090036249568424e-05,
+      "loss": 4.2506,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4081650302057904e-05,
+      "loss": 4.2609,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407326435454738e-05,
+      "loss": 4.2553,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406487840703686e-05,
+      "loss": 4.2453,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405649245952634e-05,
+      "loss": 4.26,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404810651201582e-05,
+      "loss": 4.2476,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40397205645053e-05,
+      "loss": 4.2616,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403136737460224e-05,
+      "loss": 4.265,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402298142709173e-05,
+      "loss": 4.2519,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40145954795812e-05,
+      "loss": 4.2483,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400620953207068e-05,
+      "loss": 4.2513,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399782358456016e-05,
+      "loss": 4.2518,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398943763704964e-05,
+      "loss": 4.251,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398105168953912e-05,
+      "loss": 4.252,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397268212083233e-05,
+      "loss": 4.2419,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396429617332181e-05,
+      "loss": 4.2608,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395591022581129e-05,
+      "loss": 4.254,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394752427830077e-05,
+      "loss": 4.252,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393913833079025e-05,
+      "loss": 4.2365,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3930768762083466e-05,
+      "loss": 4.25,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922382814572946e-05,
+      "loss": 4.2408,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3913996867062426e-05,
+      "loss": 4.2537,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905610919551906e-05,
+      "loss": 4.2563,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897224972041386e-05,
+      "loss": 4.2513,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3888839024530866e-05,
+      "loss": 4.2307,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880469455824075e-05,
+      "loss": 4.2537,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872083508313555e-05,
+      "loss": 4.246,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863697560803035e-05,
+      "loss": 4.2581,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855311613292515e-05,
+      "loss": 4.2368,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846942044585724e-05,
+      "loss": 4.2542,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838556097075204e-05,
+      "loss": 4.2484,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830170149564684e-05,
+      "loss": 4.2525,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821784202054164e-05,
+      "loss": 4.243,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813398254543644e-05,
+      "loss": 4.2492,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380501230703313e-05,
+      "loss": 4.2347,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379662635952261e-05,
+      "loss": 4.2496,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378824041201209e-05,
+      "loss": 4.245,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37798708433053e-05,
+      "loss": 4.2428,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377148489579478e-05,
+      "loss": 4.2367,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376311532708799e-05,
+      "loss": 4.2445,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375472937957747e-05,
+      "loss": 4.2466,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.229648113250732,
+      "eval_runtime": 297.36,
+      "eval_samples_per_second": 1283.263,
+      "eval_steps_per_second": 40.103,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374634343206695e-05,
+      "loss": 4.2409,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373795748455643e-05,
+      "loss": 4.232,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372957153704591e-05,
+      "loss": 4.251,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372120196833912e-05,
+      "loss": 4.2372,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37128160208286e-05,
+      "loss": 4.2469,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3704430073318084e-05,
+      "loss": 4.2401,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3696044125807564e-05,
+      "loss": 4.2412,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368767455710077e-05,
+      "loss": 4.2407,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367928860959025e-05,
+      "loss": 4.2369,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367090266207973e-05,
+      "loss": 4.2391,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366251671456921e-05,
+      "loss": 4.2449,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3654130767058686e-05,
+      "loss": 4.244,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36457611983519e-05,
+      "loss": 4.2364,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363737525084138e-05,
+      "loss": 4.2318,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362898930333086e-05,
+      "loss": 4.2383,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3620603355820335e-05,
+      "loss": 4.2203,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361221740830982e-05,
+      "loss": 4.235,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360384783960304e-05,
+      "loss": 4.2344,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359546189209251e-05,
+      "loss": 4.233,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358707594458199e-05,
+      "loss": 4.2523,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357868999707147e-05,
+      "loss": 4.2413,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357030404956095e-05,
+      "loss": 4.2465,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356191810205043e-05,
+      "loss": 4.2355,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355353215453991e-05,
+      "loss": 4.2404,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354516258583312e-05,
+      "loss": 4.2332,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35367766383226e-05,
+      "loss": 4.2357,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352839069081208e-05,
+      "loss": 4.2312,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352000474330156e-05,
+      "loss": 4.2307,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351161879579104e-05,
+      "loss": 4.2264,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350323284828052e-05,
+      "loss": 4.229,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3494846900770006e-05,
+      "loss": 4.233,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486460953259486e-05,
+      "loss": 4.2372,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478091384552696e-05,
+      "loss": 4.2341,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3469705437042175e-05,
+      "loss": 4.2388,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461335868335385e-05,
+      "loss": 4.2239,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3452949920824865e-05,
+      "loss": 4.2307,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3444563973314344e-05,
+      "loss": 4.2255,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3436178025803824e-05,
+      "loss": 4.216,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3427792078293304e-05,
+      "loss": 4.2302,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419406130782784e-05,
+      "loss": 4.2131,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3411036562075993e-05,
+      "loss": 4.2316,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402650614565473e-05,
+      "loss": 4.2321,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339426466705496e-05,
+      "loss": 4.2292,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338587871954444e-05,
+      "loss": 4.2228,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337749277203392e-05,
+      "loss": 4.2364,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33691068245234e-05,
+      "loss": 4.2147,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336073725581661e-05,
+      "loss": 4.2352,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335235130830609e-05,
+      "loss": 4.2211,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334396536079557e-05,
+      "loss": 4.2019,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333557941328505e-05,
+      "loss": 4.2366,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332719346577452e-05,
+      "loss": 4.2191,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3318807518264e-05,
+      "loss": 4.2233,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331042157075348e-05,
+      "loss": 4.2118,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33020520020467e-05,
+      "loss": 4.2117,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329366605453618e-05,
+      "loss": 4.2183,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3285296485829394e-05,
+      "loss": 4.2189,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3276910538318874e-05,
+      "loss": 4.2236,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326852459080835e-05,
+      "loss": 4.2193,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326013864329783e-05,
+      "loss": 4.2326,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325175269578731e-05,
+      "loss": 4.2148,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324336674827679e-05,
+      "loss": 4.2045,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323498080076627e-05,
+      "loss": 4.2244,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322659485325575e-05,
+      "loss": 4.2048,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321820890574523e-05,
+      "loss": 4.2062,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3209822958234707e-05,
+      "loss": 4.2236,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201437010724187e-05,
+      "loss": 4.2227,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193051063213666e-05,
+      "loss": 4.2088,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3184681494506876e-05,
+      "loss": 4.211,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317631192580009e-05,
+      "loss": 4.1991,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316792597828957e-05,
+      "loss": 4.2108,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315954003077905e-05,
+      "loss": 4.2214,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315115408326853e-05,
+      "loss": 4.2185,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314278451456174e-05,
+      "loss": 4.2185,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313439856705122e-05,
+      "loss": 4.2167,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31260126195407e-05,
+      "loss": 4.2311,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311762667203018e-05,
+      "loss": 4.2065,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310924072451966e-05,
+      "loss": 4.2196,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310085477700914e-05,
+      "loss": 4.2152,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309246882949862e-05,
+      "loss": 4.1921,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308409926079183e-05,
+      "loss": 4.2211,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3075713313281316e-05,
+      "loss": 4.2163,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3067327365770796e-05,
+      "loss": 4.2154,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3058941418260276e-05,
+      "loss": 4.2117,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050555470749756e-05,
+      "loss": 4.1984,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042169523239236e-05,
+      "loss": 4.2054,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303378357572871e-05,
+      "loss": 4.2089,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302539762821819e-05,
+      "loss": 4.2163,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3017044438315134e-05,
+      "loss": 4.2038,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008658490804614e-05,
+      "loss": 4.2151,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000272543294094e-05,
+      "loss": 4.2034,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2991886595783574e-05,
+      "loss": 4.2083,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983500648273054e-05,
+      "loss": 4.2026,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297513107956627e-05,
+      "loss": 4.2041,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296674513205575e-05,
+      "loss": 4.2202,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295835918454523e-05,
+      "loss": 4.2167,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294997323703471e-05,
+      "loss": 4.2207,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294158728952418e-05,
+      "loss": 4.1942,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293320134201366e-05,
+      "loss": 4.2078,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292481539450314e-05,
+      "loss": 4.2031,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291642944699262e-05,
+      "loss": 4.2136,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290805987828583e-05,
+      "loss": 4.2041,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289967393077531e-05,
+      "loss": 4.205,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289128798326479e-05,
+      "loss": 4.1963,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288290203575427e-05,
+      "loss": 4.2117,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287453246704749e-05,
+      "loss": 4.1883,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286614651953697e-05,
+      "loss": 4.2068,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285777695083018e-05,
+      "loss": 4.1902,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2849391003319656e-05,
+      "loss": 4.2082,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2841005055809136e-05,
+      "loss": 4.2011,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2832619108298616e-05,
+      "loss": 4.2029,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2824233160788096e-05,
+      "loss": 4.2026,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2815847213277576e-05,
+      "loss": 4.2006,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807461265767056e-05,
+      "loss": 4.2018,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2799075318256536e-05,
+      "loss": 4.1972,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2790722128353474e-05,
+      "loss": 4.2049,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278233618084296e-05,
+      "loss": 4.2183,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277396661213618e-05,
+      "loss": 4.2011,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276558066462566e-05,
+      "loss": 4.1959,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275719471711513e-05,
+      "loss": 4.2015,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274880876960461e-05,
+      "loss": 4.2012,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274042282209409e-05,
+      "loss": 4.1938,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273203687458357e-05,
+      "loss": 4.203,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272365092707305e-05,
+      "loss": 4.1896,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271528135836626e-05,
+      "loss": 4.2076,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270689541085574e-05,
+      "loss": 4.206,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269850946334522e-05,
+      "loss": 4.199,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26901235158347e-05,
+      "loss": 4.1895,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268173756832418e-05,
+      "loss": 4.1958,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267335162081366e-05,
+      "loss": 4.1941,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2664965673303146e-05,
+      "loss": 4.2034,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2656579725792626e-05,
+      "loss": 4.2015,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2648193778282105e-05,
+      "loss": 4.202,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2639807830771585e-05,
+      "loss": 4.1793,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2631438262064795e-05,
+      "loss": 4.2026,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2623052314554274e-05,
+      "loss": 4.1961,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2614666367043754e-05,
+      "loss": 4.2079,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2606280419533234e-05,
+      "loss": 4.1873,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2597910850826443e-05,
+      "loss": 4.2073,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2589524903315923e-05,
+      "loss": 4.1958,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2581138955805403e-05,
+      "loss": 4.2036,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257275300829488e-05,
+      "loss": 4.1898,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256436706078436e-05,
+      "loss": 4.2055,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255599749207758e-05,
+      "loss": 4.18,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254761154456706e-05,
+      "loss": 4.1979,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253922559705654e-05,
+      "loss": 4.1965,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253083964954602e-05,
+      "loss": 4.1881,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252245370203549e-05,
+      "loss": 4.1935,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251406775452497e-05,
+      "loss": 4.1908,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250569818581819e-05,
+      "loss": 4.1988,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.188170909881592,
+      "eval_runtime": 302.2958,
+      "eval_samples_per_second": 1262.31,
+      "eval_steps_per_second": 39.448,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249731223830766e-05,
+      "loss": 4.1928,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248892629079714e-05,
+      "loss": 4.1818,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248054034328662e-05,
+      "loss": 4.1991,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24721543957761e-05,
+      "loss": 4.1877,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246376844826558e-05,
+      "loss": 4.1985,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245538250075506e-05,
+      "loss": 4.1879,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244699655324455e-05,
+      "loss": 4.1929,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243861060573403e-05,
+      "loss": 4.1932,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243022465822351e-05,
+      "loss": 4.1887,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242183871071299e-05,
+      "loss": 4.1912,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241345276320247e-05,
+      "loss": 4.1943,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240506681569195e-05,
+      "loss": 4.1949,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239669724698516e-05,
+      "loss": 4.1888,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2388311299474637e-05,
+      "loss": 4.1838,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2379925351964117e-05,
+      "loss": 4.19,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2371539404453596e-05,
+      "loss": 4.1708,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363153456943076e-05,
+      "loss": 4.1891,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235476750943255e-05,
+      "loss": 4.186,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234638156192203e-05,
+      "loss": 4.1834,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2337995614411516e-05,
+      "loss": 4.2058,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2329609666900996e-05,
+      "loss": 4.1961,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321223719390476e-05,
+      "loss": 4.1984,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2312837771879956e-05,
+      "loss": 4.1851,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304451824369436e-05,
+      "loss": 4.1922,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2296082255662645e-05,
+      "loss": 4.1877,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2287696308152125e-05,
+      "loss": 4.1861,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2279310360641605e-05,
+      "loss": 4.1854,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2270924413131085e-05,
+      "loss": 4.1861,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2262554844424294e-05,
+      "loss": 4.1767,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254168896913774e-05,
+      "loss": 4.1795,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245782949403254e-05,
+      "loss": 4.1822,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223741338069647e-05,
+      "loss": 4.1918,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222902743318595e-05,
+      "loss": 4.1876,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222064148567543e-05,
+      "loss": 4.1924,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221225553816491e-05,
+      "loss": 4.1769,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220386959065439e-05,
+      "loss": 4.1805,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21955000219476e-05,
+      "loss": 4.1829,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218711407443708e-05,
+      "loss": 4.1713,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217872812692656e-05,
+      "loss": 4.1788,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217034217941604e-05,
+      "loss": 4.1723,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216195623190552e-05,
+      "loss": 4.1825,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2153570284395e-05,
+      "loss": 4.1851,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214518433688448e-05,
+      "loss": 4.1818,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213681476817769e-05,
+      "loss": 4.1766,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212842882066717e-05,
+      "loss": 4.1869,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212004287315665e-05,
+      "loss": 4.1732,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2111656925646134e-05,
+      "loss": 4.1862,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103270978135614e-05,
+      "loss": 4.1776,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2094885030625094e-05,
+      "loss": 4.1546,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208649908311457e-05,
+      "loss": 4.1976,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207811313560405e-05,
+      "loss": 4.1687,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206974356689726e-05,
+      "loss": 4.1801,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061357619386736e-05,
+      "loss": 4.1612,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2052971671876216e-05,
+      "loss": 4.1707,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2044585724365696e-05,
+      "loss": 4.1653,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203621615565891e-05,
+      "loss": 4.1761,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2027830208148385e-05,
+      "loss": 4.175,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201944426063787e-05,
+      "loss": 4.177,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201105831312735e-05,
+      "loss": 4.1866,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200267236561683e-05,
+      "loss": 4.1705,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199430279691004e-05,
+      "loss": 4.1586,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198591684939952e-05,
+      "loss": 4.1842,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1977530901889e-05,
+      "loss": 4.1606,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196914495437848e-05,
+      "loss": 4.156,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196075900686796e-05,
+      "loss": 4.1799,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195238943816117e-05,
+      "loss": 4.1786,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194400349065065e-05,
+      "loss": 4.1627,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193563392194386e-05,
+      "loss": 4.1715,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192724797443334e-05,
+      "loss": 4.155,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1918862026922826e-05,
+      "loss": 4.167,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1910476079412306e-05,
+      "loss": 4.1763,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902090131901786e-05,
+      "loss": 4.1742,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1893720563194995e-05,
+      "loss": 4.1749,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1885334615684475e-05,
+      "loss": 4.1738,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1876948668173955e-05,
+      "loss": 4.1821,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1868562720663435e-05,
+      "loss": 4.161,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860176773152915e-05,
+      "loss": 4.176,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1851807204446124e-05,
+      "loss": 4.1659,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843421256935604e-05,
+      "loss": 4.1567,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1835035309425084e-05,
+      "loss": 4.1732,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1826649361914564e-05,
+      "loss": 4.1761,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818263414404044e-05,
+      "loss": 4.1717,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1809877466893524e-05,
+      "loss": 4.1696,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180149151938301e-05,
+      "loss": 4.1541,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179310557187249e-05,
+      "loss": 4.1591,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17847360031657e-05,
+      "loss": 4.1706,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177636643445891e-05,
+      "loss": 4.1721,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176798048694839e-05,
+      "loss": 4.1624,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.175959453943787e-05,
+      "loss": 4.1715,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175120859192735e-05,
+      "loss": 4.1544,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174282264441683e-05,
+      "loss": 4.1682,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173443669690631e-05,
+      "loss": 4.1597,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172605074939579e-05,
+      "loss": 4.1628,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171766480188527e-05,
+      "loss": 4.1746,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170927885437475e-05,
+      "loss": 4.1747,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1700909285667964e-05,
+      "loss": 4.1774,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169253971696117e-05,
+      "loss": 4.1523,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168415376945065e-05,
+      "loss": 4.1671,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167576782194013e-05,
+      "loss": 4.1577,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166738187442961e-05,
+      "loss": 4.1716,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165899592691909e-05,
+      "loss": 4.1617,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165060997940857e-05,
+      "loss": 4.1657,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642224031898046e-05,
+      "loss": 4.1563,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1633838084387526e-05,
+      "loss": 4.1662,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1625452136877006e-05,
+      "loss": 4.15,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161708256817022e-05,
+      "loss": 4.1632,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160871299946343e-05,
+      "loss": 4.1473,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160032705195292e-05,
+      "loss": 4.1662,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159195748324613e-05,
+      "loss": 4.1571,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158357153573561e-05,
+      "loss": 4.161,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1575185588225087e-05,
+      "loss": 4.1595,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1566799640714567e-05,
+      "loss": 4.1617,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558413693204047e-05,
+      "loss": 4.1605,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155002774569352e-05,
+      "loss": 4.1545,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1541641798183e-05,
+      "loss": 4.1658,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153325585067248e-05,
+      "loss": 4.1734,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1524886281965695e-05,
+      "loss": 4.1645,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151650033445517e-05,
+      "loss": 4.1576,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1508114386944655e-05,
+      "loss": 4.1549,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1499728439434135e-05,
+      "loss": 4.1634,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1491342491923615e-05,
+      "loss": 4.1536,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1482956544413095e-05,
+      "loss": 4.1644,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1474570596902575e-05,
+      "loss": 4.1474,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1466184649392055e-05,
+      "loss": 4.1676,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1457815080685264e-05,
+      "loss": 4.163,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1449429133174744e-05,
+      "loss": 4.1591,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1441043185664224e-05,
+      "loss": 4.1496,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1432657238153704e-05,
+      "loss": 4.1544,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424271290643184e-05,
+      "loss": 4.1516,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1415885343132664e-05,
+      "loss": 4.1657,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1407499395622144e-05,
+      "loss": 4.1607,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1399113448111624e-05,
+      "loss": 4.1596,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139074387940483e-05,
+      "loss": 4.1407,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138237431069805e-05,
+      "loss": 4.1642,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137398836318753e-05,
+      "loss": 4.1549,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136560241567701e-05,
+      "loss": 4.1672,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135721646816649e-05,
+      "loss": 4.1476,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134883052065597e-05,
+      "loss": 4.1642,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134046095194918e-05,
+      "loss": 4.1536,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133207500443866e-05,
+      "loss": 4.1696,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132370543573187e-05,
+      "loss": 4.1505,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131531948822135e-05,
+      "loss": 4.1611,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130693354071083e-05,
+      "loss": 4.1455,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129854759320031e-05,
+      "loss": 4.1583,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129016164568979e-05,
+      "loss": 4.1569,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1281775698179273e-05,
+      "loss": 4.1461,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127340612947248e-05,
+      "loss": 4.1579,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126502018196196e-05,
+      "loss": 4.1471,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125663423445144e-05,
+      "loss": 4.1617,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.156265735626221,
+      "eval_runtime": 300.6979,
+      "eval_samples_per_second": 1269.018,
+      "eval_steps_per_second": 39.658,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124824828694092e-05,
+      "loss": 4.1541,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12398623394304e-05,
+      "loss": 4.142,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123147639191988e-05,
+      "loss": 4.1594,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1223090444409356e-05,
+      "loss": 4.1475,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1214704496898835e-05,
+      "loss": 4.1596,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1206318549388315e-05,
+      "loss": 4.1519,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1197932601877795e-05,
+      "loss": 4.1541,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1189546654367275e-05,
+      "loss": 4.1555,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1181160706856755e-05,
+      "loss": 4.149,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117279113814997e-05,
+      "loss": 4.1544,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116440519063945e-05,
+      "loss": 4.1563,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115601924312893e-05,
+      "loss": 4.157,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114764967442214e-05,
+      "loss": 4.1556,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113926372691162e-05,
+      "loss": 4.1431,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11308777794011e-05,
+      "loss": 4.1458,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112249183189058e-05,
+      "loss": 4.1371,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111412226318379e-05,
+      "loss": 4.1481,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110573631567327e-05,
+      "loss": 4.1514,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109735036816275e-05,
+      "loss": 4.1424,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108896442065223e-05,
+      "loss": 4.1663,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108057847314171e-05,
+      "loss": 4.1585,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1072192525631196e-05,
+      "loss": 4.1607,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1063806578120676e-05,
+      "loss": 4.1487,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1055420630610156e-05,
+      "loss": 4.151,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1047051061903365e-05,
+      "loss": 4.1567,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1038665114392845e-05,
+      "loss": 4.1461,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1030279166882325e-05,
+      "loss": 4.1472,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1021893219371805e-05,
+      "loss": 4.1537,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1013523650665014e-05,
+      "loss": 4.1356,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005137703154494e-05,
+      "loss": 4.1419,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0996751755643974e-05,
+      "loss": 4.1464,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988365808133454e-05,
+      "loss": 4.1496,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0979979860622933e-05,
+      "loss": 4.152,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0971593913112413e-05,
+      "loss": 4.153,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096320796560189e-05,
+      "loss": 4.1396,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095483839689511e-05,
+      "loss": 4.1419,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094645244938459e-05,
+      "loss": 4.1481,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09380828806778e-05,
+      "loss": 4.1328,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092969693316728e-05,
+      "loss": 4.1441,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092131098565676e-05,
+      "loss": 4.1329,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091292503814624e-05,
+      "loss": 4.1478,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090453909063572e-05,
+      "loss": 4.1442,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089616952192893e-05,
+      "loss": 4.1447,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088778357441841e-05,
+      "loss": 4.1445,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087939762690789e-05,
+      "loss": 4.1462,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087101167939737e-05,
+      "loss": 4.1387,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086262573188685e-05,
+      "loss": 4.1483,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085423978437633e-05,
+      "loss": 4.14,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084585383686581e-05,
+      "loss": 4.1211,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083746788935529e-05,
+      "loss": 4.1587,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082908194184477e-05,
+      "loss": 4.1326,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082069599433425e-05,
+      "loss": 4.1449,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081231004682373e-05,
+      "loss": 4.1306,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080392409931321e-05,
+      "loss": 4.1323,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079553815180269e-05,
+      "loss": 4.1289,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0787184961899625e-05,
+      "loss": 4.1367,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0778799014389105e-05,
+      "loss": 4.1439,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0770413066878585e-05,
+      "loss": 4.1359,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0762027119368065e-05,
+      "loss": 4.153,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075365755066128e-05,
+      "loss": 4.1401,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074527160315076e-05,
+      "loss": 4.1196,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073688565564024e-05,
+      "loss": 4.1447,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072849970812972e-05,
+      "loss": 4.1272,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07201137606192e-05,
+      "loss": 4.1182,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071172781310868e-05,
+      "loss": 4.1488,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070334186559816e-05,
+      "loss": 4.1412,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069497229689137e-05,
+      "loss": 4.1281,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068658634938085e-05,
+      "loss": 4.137,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067820040187033e-05,
+      "loss": 4.121,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066981445435981e-05,
+      "loss": 4.1321,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066142850684929e-05,
+      "loss": 4.1366,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0653058938142505e-05,
+      "loss": 4.1392,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0644672990631985e-05,
+      "loss": 4.1412,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636287043121465e-05,
+      "loss": 4.1351,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0627901095610945e-05,
+      "loss": 4.1502,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0619515148100425e-05,
+      "loss": 4.1219,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611129200589905e-05,
+      "loss": 4.1436,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0602759631883114e-05,
+      "loss": 4.128,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0594373684372594e-05,
+      "loss": 4.1252,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0585987736862074e-05,
+      "loss": 4.1372,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0577601789351554e-05,
+      "loss": 4.1413,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056921584184103e-05,
+      "loss": 4.1368,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056082989433051e-05,
+      "loss": 4.14,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055244394681999e-05,
+      "loss": 4.1162,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05440743781132e-05,
+      "loss": 4.1219,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053568843060268e-05,
+      "loss": 4.1342,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052730248309216e-05,
+      "loss": 4.1391,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051891653558164e-05,
+      "loss": 4.131,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051053058807112e-05,
+      "loss": 4.1362,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.05021446405606e-05,
+      "loss": 4.1214,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049375869305008e-05,
+      "loss": 4.1334,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048538912434329e-05,
+      "loss": 4.1266,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04770195556365e-05,
+      "loss": 4.1269,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046863360812598e-05,
+      "loss": 4.14,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046024766061546e-05,
+      "loss": 4.1388,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045186171310494e-05,
+      "loss": 4.1436,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044347576559443e-05,
+      "loss": 4.1219,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043508981808391e-05,
+      "loss": 4.1293,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042670387057339e-05,
+      "loss": 4.1277,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041831792306287e-05,
+      "loss": 4.1335,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040993197555235e-05,
+      "loss": 4.1281,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040154602804183e-05,
+      "loss": 4.1333,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039316008053131e-05,
+      "loss": 4.1226,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038477413302079e-05,
+      "loss": 4.1274,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037638818551027e-05,
+      "loss": 4.1236,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036800223799974e-05,
+      "loss": 4.1262,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0359649048096685e-05,
+      "loss": 4.1126,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0351263100586165e-05,
+      "loss": 4.1342,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0342877153075645e-05,
+      "loss": 4.1266,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0334491205565125e-05,
+      "loss": 4.1267,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0326105258054605e-05,
+      "loss": 4.1249,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031771931054409e-05,
+      "loss": 4.1297,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03093497418373e-05,
+      "loss": 4.1293,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030096379432678e-05,
+      "loss": 4.1211,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029257784681626e-05,
+      "loss": 4.1334,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028419189930574e-05,
+      "loss": 4.1388,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027582233059895e-05,
+      "loss": 4.1258,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026743638308843e-05,
+      "loss": 4.1307,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025905043557791e-05,
+      "loss": 4.1196,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025066448806738e-05,
+      "loss": 4.126,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024227854055686e-05,
+      "loss": 4.1273,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023389259304634e-05,
+      "loss": 4.1278,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022550664553583e-05,
+      "loss": 4.1113,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021713707682904e-05,
+      "loss": 4.133,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020875112931852e-05,
+      "loss": 4.1295,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0200365181808e-05,
+      "loss": 4.1279,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019197923429748e-05,
+      "loss": 4.1165,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018359328678696e-05,
+      "loss": 4.1222,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017520733927644e-05,
+      "loss": 4.1198,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016682139176592e-05,
+      "loss": 4.1314,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01584354442554e-05,
+      "loss": 4.1289,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015006587554861e-05,
+      "loss": 4.1292,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014167992803809e-05,
+      "loss": 4.1078,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013329398052757e-05,
+      "loss": 4.1305,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012492441182078e-05,
+      "loss": 4.1235,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011653846431026e-05,
+      "loss": 4.135,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010815251679974e-05,
+      "loss": 4.1184,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009976656928922e-05,
+      "loss": 4.1281,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00913806217787e-05,
+      "loss": 4.1202,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008299467426818e-05,
+      "loss": 4.1395,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007460872675766e-05,
+      "loss": 4.1195,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006622277924714e-05,
+      "loss": 4.1295,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005783683173662e-05,
+      "loss": 4.119,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004946726302983e-05,
+      "loss": 4.1255,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004108131551931e-05,
+      "loss": 4.1199,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003269536800879e-05,
+      "loss": 4.1181,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002430942049827e-05,
+      "loss": 4.1264,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001592347298775e-05,
+      "loss": 4.111,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000755390428097e-05,
+      "loss": 4.1348,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.132048606872559,
+      "eval_runtime": 299.9171,
+      "eval_samples_per_second": 1272.322,
+      "eval_steps_per_second": 39.761,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999916795677045e-05,
+      "loss": 4.1197,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999078200925993e-05,
+      "loss": 4.1127,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99823960617494e-05,
+      "loss": 4.1229,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997401011423888e-05,
+      "loss": 4.1185,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996562416672836e-05,
+      "loss": 4.1296,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995723821921784e-05,
+      "loss": 4.1216,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994885227170732e-05,
+      "loss": 4.1239,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99404663241968e-05,
+      "loss": 4.1227,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993208037668628e-05,
+      "loss": 4.1162,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992369442917576e-05,
+      "loss": 4.1237,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991530848166524e-05,
+      "loss": 4.1222,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990692253415472e-05,
+      "loss": 4.1248,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989855296544793e-05,
+      "loss": 4.1263,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890167017937416e-05,
+      "loss": 4.1092,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9881781070426896e-05,
+      "loss": 4.1194,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9873395122916376e-05,
+      "loss": 4.1035,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865009175405856e-05,
+      "loss": 4.1208,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9856623227895336e-05,
+      "loss": 4.1181,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848237280384816e-05,
+      "loss": 4.1104,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983985133287429e-05,
+      "loss": 4.134,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9831481764167505e-05,
+      "loss": 4.1276,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823095816656985e-05,
+      "loss": 4.1293,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981470986914646e-05,
+      "loss": 4.1203,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980632392163594e-05,
+      "loss": 4.1196,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979793797412542e-05,
+      "loss": 4.1263,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97895520266149e-05,
+      "loss": 4.1118,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781166079104385e-05,
+      "loss": 4.1151,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9772780131593865e-05,
+      "loss": 4.1211,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9764410562887074e-05,
+      "loss": 4.1081,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9756024615376554e-05,
+      "loss": 4.109,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9747638667866034e-05,
+      "loss": 4.1175,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973926909915924e-05,
+      "loss": 4.1203,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973088315164872e-05,
+      "loss": 4.1151,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.97224972041382e-05,
+      "loss": 4.1271,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971411125662768e-05,
+      "loss": 4.1087,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970574168792089e-05,
+      "loss": 4.1113,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969735574041037e-05,
+      "loss": 4.1162,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968896979289985e-05,
+      "loss": 4.1,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968058384538934e-05,
+      "loss": 4.1161,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967219789787882e-05,
+      "loss": 4.1065,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96638119503683e-05,
+      "loss": 4.1124,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965542600285778e-05,
+      "loss": 4.1144,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964704005534726e-05,
+      "loss": 4.1149,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963867048664047e-05,
+      "loss": 4.1172,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963028453912995e-05,
+      "loss": 4.1157,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962189859161943e-05,
+      "loss": 4.1095,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.961351264410891e-05,
+      "loss": 4.1133,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960512669659839e-05,
+      "loss": 4.1138,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959674074908787e-05,
+      "loss": 4.0896,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958835480157735e-05,
+      "loss": 4.1245,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957996885406682e-05,
+      "loss": 4.1032,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9571599285360036e-05,
+      "loss": 4.1171,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956321333784952e-05,
+      "loss": 4.1006,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9554827390339e-05,
+      "loss": 4.0995,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9546441442828476e-05,
+      "loss": 4.1025,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953808825292542e-05,
+      "loss": 4.1021,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95297023054149e-05,
+      "loss": 4.1157,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952131635790438e-05,
+      "loss": 4.102,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951293041039386e-05,
+      "loss": 4.1238,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950456084168707e-05,
+      "loss": 4.1103,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949617489417655e-05,
+      "loss": 4.0882,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948778894666603e-05,
+      "loss": 4.1127,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947940299915551e-05,
+      "loss": 4.1049,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947101705164499e-05,
+      "loss": 4.0896,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9462647482938206e-05,
+      "loss": 4.1151,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454261535427686e-05,
+      "loss": 4.1102,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9445875587917166e-05,
+      "loss": 4.0989,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9437489640406646e-05,
+      "loss": 4.1077,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429103692896126e-05,
+      "loss": 4.0869,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94207177453856e-05,
+      "loss": 4.104,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941233179787508e-05,
+      "loss": 4.1086,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9403962229168295e-05,
+      "loss": 4.1076,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939557628165777e-05,
+      "loss": 4.1157,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938719033414725e-05,
+      "loss": 4.1033,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937880438663673e-05,
+      "loss": 4.1191,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9370418439126214e-05,
+      "loss": 4.1008,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9362032491615694e-05,
+      "loss": 4.1114,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9353646544105174e-05,
+      "loss": 4.0957,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9345260596594654e-05,
+      "loss": 4.0967,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933689102788786e-05,
+      "loss": 4.109,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932850508037734e-05,
+      "loss": 4.1113,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932011913286682e-05,
+      "loss": 4.1061,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93117331853563e-05,
+      "loss": 4.1116,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930334723784578e-05,
+      "loss": 4.0855,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929496129033526e-05,
+      "loss": 4.0927,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.928657534282474e-05,
+      "loss": 4.1072,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927818939531422e-05,
+      "loss": 4.1069,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92698034478037e-05,
+      "loss": 4.0985,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926143387909691e-05,
+      "loss": 4.1049,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925304793158639e-05,
+      "loss": 4.0941,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924466198407588e-05,
+      "loss": 4.1027,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923627603656536e-05,
+      "loss": 4.0958,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922789008905484e-05,
+      "loss": 4.0994,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921952052034805e-05,
+      "loss": 4.1101,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921113457283753e-05,
+      "loss": 4.1104,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920274862532701e-05,
+      "loss": 4.1149,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919437905662022e-05,
+      "loss": 4.0944,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91859931091097e-05,
+      "loss": 4.1013,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917760716159918e-05,
+      "loss": 4.0986,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916922121408866e-05,
+      "loss": 4.1038,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916083526657813e-05,
+      "loss": 4.1006,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9152449319067617e-05,
+      "loss": 4.1001,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914407975036083e-05,
+      "loss": 4.0999,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913569380285031e-05,
+      "loss": 4.0956,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9127307855339786e-05,
+      "loss": 4.0957,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9118921907829266e-05,
+      "loss": 4.0965,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911055233912248e-05,
+      "loss": 4.0872,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9102166391611955e-05,
+      "loss": 4.1026,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9093780444101435e-05,
+      "loss": 4.0992,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9085394496590914e-05,
+      "loss": 4.0939,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9077008549080394e-05,
+      "loss": 4.101,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9068622601569874e-05,
+      "loss": 4.0965,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9060236654059354e-05,
+      "loss": 4.1006,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9051850706548834e-05,
+      "loss": 4.0917,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9043464759038314e-05,
+      "loss": 4.1085,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90350788115278e-05,
+      "loss": 4.1104,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902669286401728e-05,
+      "loss": 4.1016,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901830691650676e-05,
+      "loss": 4.0994,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90099537266037e-05,
+      "loss": 4.0939,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900156777909318e-05,
+      "loss": 4.0951,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899318183158266e-05,
+      "loss": 4.1019,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898479588407214e-05,
+      "loss": 4.1009,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897640993656162e-05,
+      "loss": 4.0814,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89680239890511e-05,
+      "loss": 4.1079,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895963804154058e-05,
+      "loss": 4.101,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895125209403006e-05,
+      "loss": 4.1,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894286614651954e-05,
+      "loss": 4.0901,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8934496577812755e-05,
+      "loss": 4.0928,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8926110630302235e-05,
+      "loss": 4.0923,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8917724682791715e-05,
+      "loss": 4.1033,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8909338735281195e-05,
+      "loss": 4.1009,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8900969166574404e-05,
+      "loss": 4.1022,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8892583219063884e-05,
+      "loss": 4.0785,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888421365035709e-05,
+      "loss": 4.1028,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887582770284657e-05,
+      "loss": 4.0966,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886744175533605e-05,
+      "loss": 4.1104,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885905580782553e-05,
+      "loss": 4.0909,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885066986031501e-05,
+      "loss": 4.0949,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884230029160822e-05,
+      "loss": 4.0979,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88339143440977e-05,
+      "loss": 4.1079,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882552839658719e-05,
+      "loss": 4.0897,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88171588278804e-05,
+      "loss": 4.1027,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880877288036988e-05,
+      "loss": 4.0922,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880038693285936e-05,
+      "loss": 4.1025,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879200098534884e-05,
+      "loss": 4.0901,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878361503783832e-05,
+      "loss": 4.087,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877522909032779e-05,
+      "loss": 4.0962,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876684314281727e-05,
+      "loss": 4.0887,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875845719530675e-05,
+      "loss": 4.1044,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.112194061279297,
+      "eval_runtime": 306.7862,
+      "eval_samples_per_second": 1243.834,
+      "eval_steps_per_second": 38.871,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.875007124779623e-05,
+      "loss": 4.091,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874168530028571e-05,
+      "loss": 4.0853,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.873329935277519e-05,
+      "loss": 4.0954,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8724929784068406e-05,
+      "loss": 4.0917,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8716543836557886e-05,
+      "loss": 4.1044,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8708157889047366e-05,
+      "loss": 4.0939,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8699771941536846e-05,
+      "loss": 4.0906,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8691385994026326e-05,
+      "loss": 4.1002,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8683016425319535e-05,
+      "loss": 4.0896,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8674630477809015e-05,
+      "loss": 4.0957,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8666244530298495e-05,
+      "loss": 4.0987,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8657858582787975e-05,
+      "loss": 4.0988,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8649489014081184e-05,
+      "loss": 4.099,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8641103066570664e-05,
+      "loss": 4.0843,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8632717119060144e-05,
+      "loss": 4.0891,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8624331171549624e-05,
+      "loss": 4.075,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861594522403911e-05,
+      "loss": 4.0928,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.860755927652859e-05,
+      "loss": 4.0949,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859920608662553e-05,
+      "loss": 4.0808,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859082013911501e-05,
+      "loss": 4.1058,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858243419160449e-05,
+      "loss": 4.1006,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857404824409397e-05,
+      "loss": 4.0988,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.856566229658345e-05,
+      "loss": 4.096,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855727634907293e-05,
+      "loss": 4.0952,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854889040156241e-05,
+      "loss": 4.0978,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854050445405189e-05,
+      "loss": 4.0936,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853211850654137e-05,
+      "loss": 4.0819,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852373255903085e-05,
+      "loss": 4.0964,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851534661152033e-05,
+      "loss": 4.0801,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.850696066400981e-05,
+      "loss": 4.0864,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849857471649929e-05,
+      "loss": 4.0914,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8490205147792504e-05,
+      "loss": 4.094,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848181920028198e-05,
+      "loss": 4.0884,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847343325277146e-05,
+      "loss": 4.0997,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846504730526094e-05,
+      "loss": 4.0863,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.845667773655415e-05,
+      "loss": 4.0827,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8448291789043626e-05,
+      "loss": 4.0945,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8439905841533106e-05,
+      "loss": 4.0728,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8431519894022586e-05,
+      "loss": 4.0869,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8423133946512066e-05,
+      "loss": 4.0874,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8414747999001546e-05,
+      "loss": 4.084,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840637843029476e-05,
+      "loss": 4.0878,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839799248278424e-05,
+      "loss": 4.0892,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838960653527372e-05,
+      "loss": 4.0889,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83812205877632e-05,
+      "loss": 4.0894,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837283464025268e-05,
+      "loss": 4.0832,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836444869274216e-05,
+      "loss": 4.088,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835606274523164e-05,
+      "loss": 4.0898,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834767679772112e-05,
+      "loss": 4.0667,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833930722901433e-05,
+      "loss": 4.0971,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833092128150381e-05,
+      "loss": 4.0805,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832253533399329e-05,
+      "loss": 4.0862,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831414938648277e-05,
+      "loss": 4.0751,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8305779817775986e-05,
+      "loss": 4.0751,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8297393870265466e-05,
+      "loss": 4.0777,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8289024301558675e-05,
+      "loss": 4.0788,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8280638354048155e-05,
+      "loss": 4.0919,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8272252406537635e-05,
+      "loss": 4.0771,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8263866459027115e-05,
+      "loss": 4.0987,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8255480511516595e-05,
+      "loss": 4.0886,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8247094564006075e-05,
+      "loss": 4.0609,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8238724995299284e-05,
+      "loss": 4.0839,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8230339047788764e-05,
+      "loss": 4.0751,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8221953100278244e-05,
+      "loss": 4.0715,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8213567152767724e-05,
+      "loss": 4.087,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8205181205257204e-05,
+      "loss": 4.0843,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8196795257746684e-05,
+      "loss": 4.0748,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81884256890399e-05,
+      "loss": 4.085,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818003974152938e-05,
+      "loss": 4.0648,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817165379401886e-05,
+      "loss": 4.0711,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816326784650834e-05,
+      "loss": 4.0836,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815488189899781e-05,
+      "loss": 4.0829,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814651233029103e-05,
+      "loss": 4.0892,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813812638278051e-05,
+      "loss": 4.0777,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812974043526999e-05,
+      "loss": 4.0955,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812135448775946e-05,
+      "loss": 4.075,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811296854024894e-05,
+      "loss": 4.0913,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810458259273842e-05,
+      "loss": 4.0706,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80961966452279e-05,
+      "loss": 4.0744,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808782707652112e-05,
+      "loss": 4.0809,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80794411290106e-05,
+      "loss": 4.0887,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807105518150008e-05,
+      "loss": 4.0818,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806266923398956e-05,
+      "loss": 4.0877,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805428328647904e-05,
+      "loss": 4.062,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804591371777225e-05,
+      "loss": 4.0649,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8037527770261727e-05,
+      "loss": 4.0843,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8029141822751207e-05,
+      "loss": 4.0809,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8020755875240687e-05,
+      "loss": 4.0757,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8012369927730166e-05,
+      "loss": 4.0817,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8003983980219646e-05,
+      "loss": 4.0766,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7995598032709126e-05,
+      "loss": 4.0744,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7987212085198606e-05,
+      "loss": 4.0736,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797884251649182e-05,
+      "loss": 4.0732,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79704565689813e-05,
+      "loss": 4.0867,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796207062147078e-05,
+      "loss": 4.0881,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.795370105276399e-05,
+      "loss": 4.0888,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.794531510525347e-05,
+      "loss": 4.0681,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793692915774295e-05,
+      "loss": 4.0776,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792854321023243e-05,
+      "loss": 4.0756,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792015726272191e-05,
+      "loss": 4.0807,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791177131521139e-05,
+      "loss": 4.0796,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.790338536770087e-05,
+      "loss": 4.0738,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789499942019035e-05,
+      "loss": 4.0764,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788662985148356e-05,
+      "loss": 4.0691,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787824390397304e-05,
+      "loss": 4.0734,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786985795646253e-05,
+      "loss": 4.0727,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7861472008952e-05,
+      "loss": 4.065,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785308606144148e-05,
+      "loss": 4.0774,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.784470011393096e-05,
+      "loss": 4.0757,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7836330545224176e-05,
+      "loss": 4.0719,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782794459771365e-05,
+      "loss": 4.0742,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781955865020313e-05,
+      "loss": 4.0743,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781117270269261e-05,
+      "loss": 4.0785,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7802803133985825e-05,
+      "loss": 4.0676,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77944171864753e-05,
+      "loss": 4.0836,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778603123896478e-05,
+      "loss": 4.0838,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7777645291454264e-05,
+      "loss": 4.0796,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7769259343943744e-05,
+      "loss": 4.0761,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7760873396433224e-05,
+      "loss": 4.0729,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7752503827726433e-05,
+      "loss": 4.0698,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7744117880215913e-05,
+      "loss": 4.0797,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7735731932705393e-05,
+      "loss": 4.0731,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772734598519487e-05,
+      "loss": 4.0612,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771897641648808e-05,
+      "loss": 4.0851,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771059046897756e-05,
+      "loss": 4.0788,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770222090027077e-05,
+      "loss": 4.0756,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769383495276025e-05,
+      "loss": 4.0674,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768544900524973e-05,
+      "loss": 4.0672,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767706305773921e-05,
+      "loss": 4.0692,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76686771102287e-05,
+      "loss": 4.0811,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766029116271818e-05,
+      "loss": 4.0781,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765190521520766e-05,
+      "loss": 4.0814,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764351926769714e-05,
+      "loss": 4.0537,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763513332018662e-05,
+      "loss": 4.085,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76267473726761e-05,
+      "loss": 4.0709,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761836142516558e-05,
+      "loss": 4.0845,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760997547765506e-05,
+      "loss": 4.0709,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760158953014454e-05,
+      "loss": 4.0677,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759321996143775e-05,
+      "loss": 4.0752,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758483401392723e-05,
+      "loss": 4.087,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.757644806641671e-05,
+      "loss": 4.0674,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756806211890618e-05,
+      "loss": 4.0792,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755967617139567e-05,
+      "loss": 4.0688,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755130660268888e-05,
+      "loss": 4.0772,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754292065517836e-05,
+      "loss": 4.0669,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7534534707667836e-05,
+      "loss": 4.067,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7526148760157316e-05,
+      "loss": 4.0757,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7517762812646796e-05,
+      "loss": 4.0663,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7509376865136276e-05,
+      "loss": 4.0846,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.096063613891602,
+      "eval_runtime": 319.5353,
+      "eval_samples_per_second": 1194.206,
+      "eval_steps_per_second": 37.32,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7501007296429485e-05,
+      "loss": 4.0681,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7492621348918965e-05,
+      "loss": 4.0622,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7484235401408445e-05,
+      "loss": 4.0738,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7475849453897924e-05,
+      "loss": 4.0704,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7467463506387404e-05,
+      "loss": 4.0807,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7459077558876884e-05,
+      "loss": 4.0716,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7450691611366364e-05,
+      "loss": 4.0685,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744230566385585e-05,
+      "loss": 4.0789,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.743391971634533e-05,
+      "loss": 4.0664,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.742553376883481e-05,
+      "loss": 4.0664,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.741714782132429e-05,
+      "loss": 4.078,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740876187381377e-05,
+      "loss": 4.0745,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740039230510698e-05,
+      "loss": 4.0782,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.739200635759646e-05,
+      "loss": 4.06,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738362041008594e-05,
+      "loss": 4.066,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737523446257542e-05,
+      "loss": 4.0517,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73668485150649e-05,
+      "loss": 4.0724,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735846256755437e-05,
+      "loss": 4.0762,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735007662004385e-05,
+      "loss": 4.0554,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734169067253333e-05,
+      "loss": 4.0833,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733330472502282e-05,
+      "loss": 4.0802,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73249187775123e-05,
+      "loss": 4.077,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731653283000178e-05,
+      "loss": 4.0738,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730816326129499e-05,
+      "loss": 4.0724,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729977731378447e-05,
+      "loss": 4.0735,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729139136627395e-05,
+      "loss": 4.0696,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728300541876343e-05,
+      "loss": 4.0612,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727461947125291e-05,
+      "loss": 4.0752,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726623352374239e-05,
+      "loss": 4.0593,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725784757623187e-05,
+      "loss": 4.0597,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724946162872135e-05,
+      "loss": 4.0709,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724109206001456e-05,
+      "loss": 4.0677,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723270611250404e-05,
+      "loss": 4.07,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722432016499352e-05,
+      "loss": 4.0747,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7215934217483e-05,
+      "loss": 4.0599,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720756464877621e-05,
+      "loss": 4.0654,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719917870126569e-05,
+      "loss": 4.0684,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719079275375517e-05,
+      "loss": 4.051,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.718240680624465e-05,
+      "loss": 4.0649,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717402085873413e-05,
+      "loss": 4.0656,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.716563491122361e-05,
+      "loss": 4.0584,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7157248963713086e-05,
+      "loss": 4.0681,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7148863016202566e-05,
+      "loss": 4.0721,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714049344749578e-05,
+      "loss": 4.0635,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.713210749998526e-05,
+      "loss": 4.0673,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.712373793127847e-05,
+      "loss": 4.0604,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711535198376795e-05,
+      "loss": 4.0674,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.710696603625744e-05,
+      "loss": 4.0644,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709858008874691e-05,
+      "loss": 4.0474,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709019414123639e-05,
+      "loss": 4.0719,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.708180819372587e-05,
+      "loss": 4.0583,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.707342224621535e-05,
+      "loss": 4.0669,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.706503629870483e-05,
+      "loss": 4.0548,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.705665035119431e-05,
+      "loss": 4.0509,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704828078248752e-05,
+      "loss": 4.0531,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7039894834977e-05,
+      "loss": 4.0546,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703150888746648e-05,
+      "loss": 4.0743,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702313931875969e-05,
+      "loss": 4.053,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7014753371249176e-05,
+      "loss": 4.0774,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7006367423738655e-05,
+      "loss": 4.0688,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6997981476228135e-05,
+      "loss": 4.0421,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6989595528717615e-05,
+      "loss": 4.0649,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6981209581207095e-05,
+      "loss": 4.0548,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6972823633696575e-05,
+      "loss": 4.0475,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6964454064989784e-05,
+      "loss": 4.0626,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6956068117479264e-05,
+      "loss": 4.0643,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6947682169968744e-05,
+      "loss": 4.0523,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6939296222458224e-05,
+      "loss": 4.0644,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6930910274947704e-05,
+      "loss": 4.044,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6922524327437184e-05,
+      "loss": 4.0546,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6914138379926664e-05,
+      "loss": 4.06,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6905752432416144e-05,
+      "loss": 4.0663,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689738286370936e-05,
+      "loss": 4.0632,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688899691619884e-05,
+      "loss": 4.0589,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688061096868832e-05,
+      "loss": 4.0738,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68722250211778e-05,
+      "loss": 4.0554,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686385545247101e-05,
+      "loss": 4.0689,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685546950496049e-05,
+      "loss": 4.048,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68470999362537e-05,
+      "loss": 4.0543,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683871398874318e-05,
+      "loss": 4.0591,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683032804123266e-05,
+      "loss": 4.0619,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682194209372214e-05,
+      "loss": 4.0627,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681355614621162e-05,
+      "loss": 4.0668,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680518657750483e-05,
+      "loss": 4.048,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679680062999431e-05,
+      "loss": 4.0409,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6788414682483794e-05,
+      "loss": 4.0639,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6780028734973274e-05,
+      "loss": 4.0606,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677164278746275e-05,
+      "loss": 4.0539,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676325683995223e-05,
+      "loss": 4.0626,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675487089244171e-05,
+      "loss": 4.0567,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6746484944931187e-05,
+      "loss": 4.0524,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6738115376224396e-05,
+      "loss": 4.0559,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6729729428713876e-05,
+      "loss": 4.0519,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6721343481203356e-05,
+      "loss": 4.0663,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6712957533692836e-05,
+      "loss": 4.0662,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6704571586182315e-05,
+      "loss": 4.0684,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6696185638671795e-05,
+      "loss": 4.0465,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668783244876875e-05,
+      "loss": 4.0554,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667944650125822e-05,
+      "loss": 4.0566,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66710605537477e-05,
+      "loss": 4.0561,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.666267460623718e-05,
+      "loss": 4.0603,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.665428865872666e-05,
+      "loss": 4.0517,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664590271121614e-05,
+      "loss": 4.0584,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663751676370562e-05,
+      "loss": 4.0462,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66291308161951e-05,
+      "loss": 4.0534,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662076124748831e-05,
+      "loss": 4.0507,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.661239167878152e-05,
+      "loss": 4.0487,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6604005731271e-05,
+      "loss": 4.0578,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6595619783760485e-05,
+      "loss": 4.0526,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6587250215053694e-05,
+      "loss": 4.0501,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6578864267543174e-05,
+      "loss": 4.0554,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6570478320032654e-05,
+      "loss": 4.0489,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6562092372522134e-05,
+      "loss": 4.0577,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6553706425011614e-05,
+      "loss": 4.0523,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6545320477501094e-05,
+      "loss": 4.0589,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6536934529990574e-05,
+      "loss": 4.0673,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6528548582480054e-05,
+      "loss": 4.0576,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6520162634969534e-05,
+      "loss": 4.0535,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651179306626274e-05,
+      "loss": 4.0556,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650340711875222e-05,
+      "loss": 4.0491,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64950211712417e-05,
+      "loss": 4.0622,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648663522373118e-05,
+      "loss": 4.052,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647824927622067e-05,
+      "loss": 4.0438,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646987970751388e-05,
+      "loss": 4.0626,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646149376000336e-05,
+      "loss": 4.0598,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.645310781249284e-05,
+      "loss": 4.0565,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644472186498232e-05,
+      "loss": 4.0452,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643635229627553e-05,
+      "loss": 4.0482,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642796634876501e-05,
+      "loss": 4.0439,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641958040125449e-05,
+      "loss": 4.0629,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641119445374397e-05,
+      "loss": 4.0565,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.640280850623345e-05,
+      "loss": 4.0605,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.639442255872293e-05,
+      "loss": 4.0322,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6386052990016136e-05,
+      "loss": 4.0671,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637766704250562e-05,
+      "loss": 4.0487,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63692810949951e-05,
+      "loss": 4.0694,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636089514748458e-05,
+      "loss": 4.0469,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6352509199974056e-05,
+      "loss": 4.0481,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634413963126727e-05,
+      "loss": 4.0519,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633575368375675e-05,
+      "loss": 4.0709,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6327367736246225e-05,
+      "loss": 4.0455,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6318981788735705e-05,
+      "loss": 4.0622,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6310595841225185e-05,
+      "loss": 4.0464,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6302209893714665e-05,
+      "loss": 4.0582,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6293840325007874e-05,
+      "loss": 4.0475,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628545437749736e-05,
+      "loss": 4.0511,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627708480879058e-05,
+      "loss": 4.053,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626869886128006e-05,
+      "loss": 4.0459,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626031291376953e-05,
+      "loss": 4.0615,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.082852840423584,
+      "eval_runtime": 328.4034,
+      "eval_samples_per_second": 1161.958,
+      "eval_steps_per_second": 36.312,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.625192696625901e-05,
+      "loss": 4.0502,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.624354101874849e-05,
+      "loss": 4.0424,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.623515507123797e-05,
+      "loss": 4.0545,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622676912372745e-05,
+      "loss": 4.0481,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621838317621693e-05,
+      "loss": 4.0637,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620999722870641e-05,
+      "loss": 4.053,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620161128119589e-05,
+      "loss": 4.0505,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.619322533368537e-05,
+      "loss": 4.0567,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.618485576497858e-05,
+      "loss": 4.0477,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6176486196271795e-05,
+      "loss": 4.0456,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6168100248761275e-05,
+      "loss": 4.0632,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6159714301250754e-05,
+      "loss": 4.0517,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6151328353740234e-05,
+      "loss": 4.0624,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6142942406229714e-05,
+      "loss": 4.0404,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6134556458719194e-05,
+      "loss": 4.0411,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6126170511208674e-05,
+      "loss": 4.0374,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611780094250188e-05,
+      "loss": 4.0532,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610941499499136e-05,
+      "loss": 4.0496,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610102904748084e-05,
+      "loss": 4.0425,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609264309997032e-05,
+      "loss": 4.0559,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60842571524598e-05,
+      "loss": 4.0679,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607587120494928e-05,
+      "loss": 4.0549,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606748525743876e-05,
+      "loss": 4.0536,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605909930992824e-05,
+      "loss": 4.0541,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605071336241772e-05,
+      "loss": 4.057,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60423274149072e-05,
+      "loss": 4.0503,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603395784620042e-05,
+      "loss": 4.0419,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602557189868989e-05,
+      "loss": 4.0538,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601718595117937e-05,
+      "loss": 4.0396,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600881638247259e-05,
+      "loss": 4.0404,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600043043496206e-05,
+      "loss": 4.052,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599204448745154e-05,
+      "loss": 4.0466,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598365853994102e-05,
+      "loss": 4.0538,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597528897123424e-05,
+      "loss": 4.0553,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596690302372372e-05,
+      "loss": 4.0415,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59585170762132e-05,
+      "loss": 4.0492,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595013112870268e-05,
+      "loss": 4.0474,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.594174518119216e-05,
+      "loss": 4.0328,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5933359233681637e-05,
+      "loss": 4.0491,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5924973286171117e-05,
+      "loss": 4.0408,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5916587338660597e-05,
+      "loss": 4.0418,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5908201391150076e-05,
+      "loss": 4.0473,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5899815443639556e-05,
+      "loss": 4.0574,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5891429496129036e-05,
+      "loss": 4.0397,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5883043548618516e-05,
+      "loss": 4.0503,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5874673979911725e-05,
+      "loss": 4.045,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5866288032401205e-05,
+      "loss": 4.0434,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5857918463694414e-05,
+      "loss": 4.0471,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58495325161839e-05,
+      "loss": 4.0293,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584114656867338e-05,
+      "loss": 4.0495,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583276062116286e-05,
+      "loss": 4.0432,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582437467365234e-05,
+      "loss": 4.0499,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581598872614182e-05,
+      "loss": 4.0343,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58076027786313e-05,
+      "loss": 4.0293,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5799216831120774e-05,
+      "loss": 4.0347,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579084726241399e-05,
+      "loss": 4.0308,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.578246131490347e-05,
+      "loss": 4.0557,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577407536739295e-05,
+      "loss": 4.0351,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.576568941988242e-05,
+      "loss": 4.0576,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57573034723719e-05,
+      "loss": 4.0456,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574891752486138e-05,
+      "loss": 4.0262,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574053157735087e-05,
+      "loss": 4.0412,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573214562984035e-05,
+      "loss": 4.0346,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572375968232983e-05,
+      "loss": 4.0297,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571539011362304e-05,
+      "loss": 4.042,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570700416611252e-05,
+      "loss": 4.0461,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569863459740573e-05,
+      "loss": 4.0321,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569024864989521e-05,
+      "loss": 4.0459,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568186270238469e-05,
+      "loss": 4.0206,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567347675487417e-05,
+      "loss": 4.0359,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566509080736365e-05,
+      "loss": 4.0394,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565670485985313e-05,
+      "loss": 4.0475,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564833529114634e-05,
+      "loss": 4.0435,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5639949343635823e-05,
+      "loss": 4.0386,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5631563396125303e-05,
+      "loss": 4.0524,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562317744861478e-05,
+      "loss": 4.0388,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561479150110426e-05,
+      "loss": 4.0482,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.560640555359374e-05,
+      "loss": 4.0321,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559801960608322e-05,
+      "loss": 4.0348,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558965003737643e-05,
+      "loss": 4.0383,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558128046866964e-05,
+      "loss": 4.0429,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557289452115912e-05,
+      "loss": 4.0416,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55645085736486e-05,
+      "loss": 4.0491,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555612262613808e-05,
+      "loss": 4.0301,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554773667862756e-05,
+      "loss": 4.0191,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553935073111704e-05,
+      "loss": 4.0394,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553096478360652e-05,
+      "loss": 4.0449,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5522578836096e-05,
+      "loss": 4.0352,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551419288858549e-05,
+      "loss": 4.0424,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55058233198787e-05,
+      "loss": 4.0365,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549743737236818e-05,
+      "loss": 4.0344,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548905142485766e-05,
+      "loss": 4.0361,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548066547734714e-05,
+      "loss": 4.0337,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.547227952983661e-05,
+      "loss": 4.0457,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.546389358232609e-05,
+      "loss": 4.0482,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.545550763481557e-05,
+      "loss": 4.0499,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.544712168730505e-05,
+      "loss": 4.0281,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543875211859826e-05,
+      "loss": 4.0389,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543036617108774e-05,
+      "loss": 4.0323,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5421980223577226e-05,
+      "loss": 4.0397,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5413594276066706e-05,
+      "loss": 4.0432,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5405208328556186e-05,
+      "loss": 4.0327,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5396838759849395e-05,
+      "loss": 4.0381,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5388452812338875e-05,
+      "loss": 4.0293,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5380066864828355e-05,
+      "loss": 4.0342,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5371680917317835e-05,
+      "loss": 4.0336,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5363311348611044e-05,
+      "loss": 4.0285,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5354925401100524e-05,
+      "loss": 4.0373,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5346539453590004e-05,
+      "loss": 4.0373,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5338153506079483e-05,
+      "loss": 4.0312,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5329767558568963e-05,
+      "loss": 4.0344,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532138161105844e-05,
+      "loss": 4.035,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531299566354792e-05,
+      "loss": 4.038,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.530460971603741e-05,
+      "loss": 4.0343,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.529624014733062e-05,
+      "loss": 4.0344,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52878541998201e-05,
+      "loss": 4.0468,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527946825230958e-05,
+      "loss": 4.0452,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527109868360279e-05,
+      "loss": 4.032,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526271273609227e-05,
+      "loss": 4.0375,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.525432678858175e-05,
+      "loss": 4.0278,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524594084107123e-05,
+      "loss": 4.0417,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523757127236444e-05,
+      "loss": 4.0347,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522918532485392e-05,
+      "loss": 4.0258,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52207993773434e-05,
+      "loss": 4.0427,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521241342983288e-05,
+      "loss": 4.0403,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5204027482322364e-05,
+      "loss": 4.0352,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5195641534811844e-05,
+      "loss": 4.0336,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518727196610505e-05,
+      "loss": 4.0272,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517888601859453e-05,
+      "loss": 4.0242,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517050007108401e-05,
+      "loss": 4.0423,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516211412357349e-05,
+      "loss": 4.0377,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515372817606297e-05,
+      "loss": 4.0445,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5145342228552446e-05,
+      "loss": 4.0174,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5136956281041926e-05,
+      "loss": 4.0414,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512858671233514e-05,
+      "loss": 4.0346,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512020076482462e-05,
+      "loss": 4.044,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51118148173141e-05,
+      "loss": 4.0313,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.510342886980358e-05,
+      "loss": 4.0291,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.509504292229306e-05,
+      "loss": 4.0364,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.508665697478254e-05,
+      "loss": 4.049,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507827102727202e-05,
+      "loss": 4.0291,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.50698850797615e-05,
+      "loss": 4.0411,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506151551105471e-05,
+      "loss": 4.0305,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505312956354419e-05,
+      "loss": 4.0376,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504474361603367e-05,
+      "loss": 4.0308,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.503635766852315e-05,
+      "loss": 4.0307,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502797172101263e-05,
+      "loss": 4.0321,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501961853110957e-05,
+      "loss": 4.0345,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501123258359905e-05,
+      "loss": 4.0398,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0708794593811035,
+      "eval_runtime": 313.3913,
+      "eval_samples_per_second": 1217.618,
+      "eval_steps_per_second": 38.051,
+      "step": 915840
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 3.763162807120353e+17,
+  "trial_name": null,
+  "trial_params": null
+}