diff --git "a/checkpoint-1373760/trainer_state.json" "b/checkpoint-1373760/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1373760/trainer_state.json"
@@ -0,0 +1,16267 @@
+{
+  "best_metric": 3.842909812927246,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/det-noun/transformer/2/checkpoints/checkpoint-915840",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 1373760,
+  "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": 10.9882,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8218,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1856,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9763,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8063,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.6959,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.5953,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5289,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4485,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.391,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3415,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3037,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2566,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 5.2067,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 5.1749,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 5.1257,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1181,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0752,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0436,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0214,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0081,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9773,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9527,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9316,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.917,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.8861,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8712,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8498,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8334,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8184,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.8039,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7914,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7869,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7668,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7555,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7409,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7348,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.708,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.711,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6772,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6781,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6747,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6724,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6405,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.648,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6344,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6229,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.631,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.5943,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95892359812181e-05,
+      "loss": 4.5944,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958085003370758e-05,
+      "loss": 4.6007,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957246408619706e-05,
+      "loss": 4.5933,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 4.5751,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5567,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 4.5646,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 4.5449,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 4.5697,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5201,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5397,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5302,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5188,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.524,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5049,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4896,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.492,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5061,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.4763,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4773,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4619,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.462,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 4.4781,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4751,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4487,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4611,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4685,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 4.452,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 4.432,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 4.4368,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 4.4311,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 4.4222,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 4.4197,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 4.4214,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.4215,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 4.4144,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.3944,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.3975,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.3982,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 4.3939,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926231505873767e-05,
+      "loss": 4.3923,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925394549003088e-05,
+      "loss": 4.3822,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924555954252036e-05,
+      "loss": 4.379,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923717359500984e-05,
+      "loss": 4.3712,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9228804026303046e-05,
+      "loss": 4.3801,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922041807879253e-05,
+      "loss": 4.3718,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921203213128201e-05,
+      "loss": 4.3722,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920364618377149e-05,
+      "loss": 4.3697,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919526023626097e-05,
+      "loss": 4.3552,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918687428875045e-05,
+      "loss": 4.3581,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917848834123993e-05,
+      "loss": 4.3535,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917010239372941e-05,
+      "loss": 4.3519,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.3536,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.3354,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914496093000158e-05,
+      "loss": 4.3427,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913657498249106e-05,
+      "loss": 4.3404,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912818903498054e-05,
+      "loss": 4.3342,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911980308747002e-05,
+      "loss": 4.3246,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911143351876323e-05,
+      "loss": 4.3243,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.3278,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3198,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.3193,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.3134,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3037,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3221,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052748264993326e-05,
+      "loss": 4.3011,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90443623174828e-05,
+      "loss": 4.3084,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.3105,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.3075,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.299,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.2954,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.3079,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994063011223415e-05,
+      "loss": 4.2956,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985677063712895e-05,
+      "loss": 4.2827,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977291116202375e-05,
+      "loss": 4.2828,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968905168691855e-05,
+      "loss": 4.2795,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.2795,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2728,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2773,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.266,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.2664,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2623,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2807,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2742,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2578,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.2573,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.2579,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886830655617308e-05,
+      "loss": 4.2638,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.2597,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515674187595e-05,
+      "loss": 4.2556,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2548,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.2512,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.2478,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2405,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.2465,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.2361,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.2286,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2416,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877611026996855e-05,
+      "loss": 4.236,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876772432245803e-05,
+      "loss": 4.2398,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875933837494751e-05,
+      "loss": 4.2285,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875096880624072e-05,
+      "loss": 4.2261,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.195034503936768,
+      "eval_runtime": 324.8443,
+      "eval_samples_per_second": 1174.689,
+      "eval_steps_per_second": 36.71,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87425828587302e-05,
+      "loss": 4.2165,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873419691121968e-05,
+      "loss": 4.2132,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872582734251289e-05,
+      "loss": 4.2287,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871744139500237e-05,
+      "loss": 4.2139,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870905544749185e-05,
+      "loss": 4.2222,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870066949998133e-05,
+      "loss": 4.2064,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869229993127454e-05,
+      "loss": 4.2152,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683913983764027e-05,
+      "loss": 4.2022,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675544415057236e-05,
+      "loss": 4.2092,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667158467546716e-05,
+      "loss": 4.2036,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658772520036196e-05,
+      "loss": 4.2115,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650386572525676e-05,
+      "loss": 4.2063,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642000625015155e-05,
+      "loss": 4.1904,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633614677504635e-05,
+      "loss": 4.1899,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862522872999411e-05,
+      "loss": 4.1864,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8616859161287324e-05,
+      "loss": 4.1843,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608473213776804e-05,
+      "loss": 4.1873,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860008726626628e-05,
+      "loss": 4.1816,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591701318755764e-05,
+      "loss": 4.1828,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583315371245244e-05,
+      "loss": 4.1992,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574929423734724e-05,
+      "loss": 4.1815,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566543476224204e-05,
+      "loss": 4.1821,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558157528713684e-05,
+      "loss": 4.1763,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854978796000689e-05,
+      "loss": 4.1898,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854140201249637e-05,
+      "loss": 4.1599,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853301606498585e-05,
+      "loss": 4.1773,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852463011747533e-05,
+      "loss": 4.1669,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851626054876854e-05,
+      "loss": 4.1652,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850787460125802e-05,
+      "loss": 4.1611,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84994886537475e-05,
+      "loss": 4.1565,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849111908504072e-05,
+      "loss": 4.1637,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84827331375302e-05,
+      "loss": 4.1728,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847434719001968e-05,
+      "loss": 4.1578,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846596124250916e-05,
+      "loss": 4.161,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845757529499864e-05,
+      "loss": 4.1508,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844918934748812e-05,
+      "loss": 4.1593,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844081977878133e-05,
+      "loss": 4.1479,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843243383127081e-05,
+      "loss": 4.1546,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842404788376029e-05,
+      "loss": 4.1369,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841566193624977e-05,
+      "loss": 4.1437,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840727598873925e-05,
+      "loss": 4.1415,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398906420032456e-05,
+      "loss": 4.1542,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8390520472521936e-05,
+      "loss": 4.1297,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382134525011416e-05,
+      "loss": 4.1413,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83737485775009e-05,
+      "loss": 4.141,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836536262999038e-05,
+      "loss": 4.13,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835697668247986e-05,
+      "loss": 4.1429,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834859073496934e-05,
+      "loss": 4.1183,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834022116626255e-05,
+      "loss": 4.1248,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833183521875203e-05,
+      "loss": 4.1365,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832344927124151e-05,
+      "loss": 4.1311,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831506332373099e-05,
+      "loss": 4.1205,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83066937550242e-05,
+      "loss": 4.118,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829830780751368e-05,
+      "loss": 4.1223,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828992186000316e-05,
+      "loss": 4.1112,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828153591249264e-05,
+      "loss": 4.1294,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827314996498212e-05,
+      "loss": 4.1031,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264780396275336e-05,
+      "loss": 4.1162,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256394448764816e-05,
+      "loss": 4.1161,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248008501254296e-05,
+      "loss": 4.1094,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823962255374377e-05,
+      "loss": 4.1189,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823123660623325e-05,
+      "loss": 4.1076,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822285065872273e-05,
+      "loss": 4.0964,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821446471121221e-05,
+      "loss": 4.0961,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820607876370169e-05,
+      "loss": 4.1206,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81977091949949e-05,
+      "loss": 4.098,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189339626288114e-05,
+      "loss": 4.0997,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180953678777594e-05,
+      "loss": 4.0919,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172567731267074e-05,
+      "loss": 4.0942,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164181783756554e-05,
+      "loss": 4.1056,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155795836246034e-05,
+      "loss": 4.1102,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147409888735514e-05,
+      "loss": 4.0945,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813904032002872e-05,
+      "loss": 4.1053,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81306543725182e-05,
+      "loss": 4.1091,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812226842500768e-05,
+      "loss": 4.1065,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811389885630089e-05,
+      "loss": 4.084,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810551290879037e-05,
+      "loss": 4.0984,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809712696127985e-05,
+      "loss": 4.0914,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808874101376933e-05,
+      "loss": 4.0899,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808035506625881e-05,
+      "loss": 4.0859,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807196911874829e-05,
+      "loss": 4.091,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806358317123777e-05,
+      "loss": 4.0981,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805519722372726e-05,
+      "loss": 4.0911,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804681127621674e-05,
+      "loss": 4.0724,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803844170750995e-05,
+      "loss": 4.0775,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803005575999943e-05,
+      "loss": 4.0802,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802166981248891e-05,
+      "loss": 4.0836,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801328386497839e-05,
+      "loss": 4.0832,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800489791746787e-05,
+      "loss": 4.0701,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996528348761076e-05,
+      "loss": 4.0799,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988142401250556e-05,
+      "loss": 4.0627,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979772832543765e-05,
+      "loss": 4.077,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971386885033245e-05,
+      "loss": 4.0772,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7963000937522725e-05,
+      "loss": 4.0741,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795461499001221e-05,
+      "loss": 4.0804,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794622904250169e-05,
+      "loss": 4.0657,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793784309499117e-05,
+      "loss": 4.0697,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792945714748065e-05,
+      "loss": 4.0711,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792107119997013e-05,
+      "loss": 4.0717,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791270163126334e-05,
+      "loss": 4.0678,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790431568375282e-05,
+      "loss": 4.0606,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78959297362423e-05,
+      "loss": 4.0616,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887543788731774e-05,
+      "loss": 4.0665,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787917422002499e-05,
+      "loss": 4.0659,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787078827251447e-05,
+      "loss": 4.0454,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786241870380768e-05,
+      "loss": 4.0602,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854032756297166e-05,
+      "loss": 4.0644,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845646808786646e-05,
+      "loss": 4.0577,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837260861276126e-05,
+      "loss": 4.0511,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828874913765606e-05,
+      "loss": 4.0526,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782048896625508e-05,
+      "loss": 4.044,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781210301874456e-05,
+      "loss": 4.067,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780371707123404e-05,
+      "loss": 4.0486,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779534750252725e-05,
+      "loss": 4.0553,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778696155501673e-05,
+      "loss": 4.0544,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777857560750621e-05,
+      "loss": 4.0627,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770206038799423e-05,
+      "loss": 4.0483,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761820091288903e-05,
+      "loss": 4.0506,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7753434143778383e-05,
+      "loss": 4.059,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774504819626786e-05,
+      "loss": 4.0502,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773666224875734e-05,
+      "loss": 4.0427,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772827630124682e-05,
+      "loss": 4.0434,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77198903537363e-05,
+      "loss": 4.0378,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771150440622578e-05,
+      "loss": 4.0412,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770313483751899e-05,
+      "loss": 4.0415,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769474889000847e-05,
+      "loss": 4.0409,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768636294249795e-05,
+      "loss": 4.0322,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767797699498743e-05,
+      "loss": 4.0349,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766959104747691e-05,
+      "loss": 4.0323,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766122147877012e-05,
+      "loss": 4.051,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76528355312596e-05,
+      "loss": 4.0504,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764444958374909e-05,
+      "loss": 4.0349,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763606363623857e-05,
+      "loss": 4.0248,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762769406753178e-05,
+      "loss": 4.0385,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761930812002126e-05,
+      "loss": 4.038,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761092217251074e-05,
+      "loss": 4.0387,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760253622500022e-05,
+      "loss": 4.0334,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7594166656293426e-05,
+      "loss": 4.0379,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7585780708782906e-05,
+      "loss": 4.03,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577394761272386e-05,
+      "loss": 4.0333,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569008813761866e-05,
+      "loss": 4.0285,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560639245055075e-05,
+      "loss": 4.0305,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552253297544555e-05,
+      "loss": 4.0247,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543867350034035e-05,
+      "loss": 4.0193,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753548140252352e-05,
+      "loss": 4.0328,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752711183381673e-05,
+      "loss": 4.0282,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751872588630621e-05,
+      "loss": 4.0317,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751033993879569e-05,
+      "loss": 4.0214,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750195399128517e-05,
+      "loss": 4.0212,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.026453971862793,
+      "eval_runtime": 326.8324,
+      "eval_samples_per_second": 1167.543,
+      "eval_steps_per_second": 36.487,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749356804377465e-05,
+      "loss": 4.0146,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748518209626413e-05,
+      "loss": 4.0118,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747679614875361e-05,
+      "loss": 4.026,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468410201243084e-05,
+      "loss": 4.0165,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460024253732563e-05,
+      "loss": 4.0251,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451638306222043e-05,
+      "loss": 4.0082,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744325235871152e-05,
+      "loss": 4.0191,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434866411201e-05,
+      "loss": 4.0066,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742648046369049e-05,
+      "loss": 4.0178,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741809451617997e-05,
+      "loss": 4.0117,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740970856866945e-05,
+      "loss": 4.0185,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740133899996266e-05,
+      "loss": 4.0143,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739296943125587e-05,
+      "loss": 4.0013,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738458348374535e-05,
+      "loss": 4.0058,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737619753623483e-05,
+      "loss": 3.9994,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736781158872431e-05,
+      "loss": 4.0017,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735942564121379e-05,
+      "loss": 4.0025,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735103969370327e-05,
+      "loss": 3.9974,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734265374619275e-05,
+      "loss": 3.9982,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733428417748596e-05,
+      "loss": 4.0195,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7325898229975444e-05,
+      "loss": 4.0033,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7317512282464924e-05,
+      "loss": 4.0033,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7309126334954404e-05,
+      "loss": 4.0017,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730075676624761e-05,
+      "loss": 4.0079,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729237081873709e-05,
+      "loss": 3.986,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728398487122657e-05,
+      "loss": 3.9999,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727559892371605e-05,
+      "loss": 3.9971,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726721297620553e-05,
+      "loss": 3.9915,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725884340749874e-05,
+      "loss": 3.9847,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725045745998822e-05,
+      "loss": 3.9842,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72420715124777e-05,
+      "loss": 3.9946,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723368556496718e-05,
+      "loss": 4.003,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72253159962604e-05,
+      "loss": 3.9912,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721693004874988e-05,
+      "loss": 3.9932,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208560480043086e-05,
+      "loss": 3.9854,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7200174532532566e-05,
+      "loss": 3.9896,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191788585022046e-05,
+      "loss": 3.9823,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183402637511526e-05,
+      "loss": 3.9943,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175016690001006e-05,
+      "loss": 3.9711,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7166630742490486e-05,
+      "loss": 3.9821,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158244794979966e-05,
+      "loss": 3.9778,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149858847469446e-05,
+      "loss": 3.9929,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141489278762655e-05,
+      "loss": 3.9728,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133103331252135e-05,
+      "loss": 3.9781,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124717383741615e-05,
+      "loss": 3.9845,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116331436231095e-05,
+      "loss": 3.9735,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710796186752431e-05,
+      "loss": 3.9847,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709957592001379e-05,
+      "loss": 3.9645,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709118997250327e-05,
+      "loss": 3.9705,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082804024992744e-05,
+      "loss": 3.9808,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707443445628596e-05,
+      "loss": 3.9724,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706604850877544e-05,
+      "loss": 3.9687,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705766256126492e-05,
+      "loss": 3.9654,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704927661375439e-05,
+      "loss": 3.9736,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704090704504761e-05,
+      "loss": 3.9566,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703252109753709e-05,
+      "loss": 3.9733,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702413515002657e-05,
+      "loss": 3.9529,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701574920251605e-05,
+      "loss": 3.9692,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007379633809265e-05,
+      "loss": 3.9717,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998993686298745e-05,
+      "loss": 3.9596,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699060773878822e-05,
+      "loss": 3.9676,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69822217912777e-05,
+      "loss": 3.9679,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973852222570914e-05,
+      "loss": 3.9467,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965466275060394e-05,
+      "loss": 3.9547,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695708032754987e-05,
+      "loss": 3.9673,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694869438003935e-05,
+      "loss": 3.9598,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694032481133256e-05,
+      "loss": 3.957,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693193886382204e-05,
+      "loss": 3.946,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692355291631152e-05,
+      "loss": 3.951,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915166968801e-05,
+      "loss": 3.9653,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690679740009422e-05,
+      "loss": 3.9666,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689841145258369e-05,
+      "loss": 3.9506,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689002550507317e-05,
+      "loss": 3.9664,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688163955756265e-05,
+      "loss": 3.9678,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687326998885587e-05,
+      "loss": 3.9725,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686488404134534e-05,
+      "loss": 3.94,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685649809383482e-05,
+      "loss": 3.9644,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848128525128036e-05,
+      "loss": 3.9534,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683974257761751e-05,
+      "loss": 3.9506,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831356630106996e-05,
+      "loss": 3.9534,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822970682596476e-05,
+      "loss": 3.957,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681460111388969e-05,
+      "loss": 3.9604,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806215166379165e-05,
+      "loss": 3.9588,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6797829218868645e-05,
+      "loss": 3.9362,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789443271358125e-05,
+      "loss": 3.9447,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678107370265134e-05,
+      "loss": 3.9479,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772687755140814e-05,
+      "loss": 3.9514,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764301807630294e-05,
+      "loss": 3.9513,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755915860119774e-05,
+      "loss": 3.9376,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674754629141298e-05,
+      "loss": 3.9505,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673916034390246e-05,
+      "loss": 3.9338,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673077439639194e-05,
+      "loss": 3.9465,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672238844888143e-05,
+      "loss": 3.946,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671401888017464e-05,
+      "loss": 3.9458,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670563293266412e-05,
+      "loss": 3.9509,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66972469851536e-05,
+      "loss": 3.9461,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668886103764308e-05,
+      "loss": 3.9359,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668049146893629e-05,
+      "loss": 3.9485,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667210552142577e-05,
+      "loss": 3.9465,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663735952718984e-05,
+      "loss": 3.9429,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665535000520846e-05,
+      "loss": 3.9328,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664696405769794e-05,
+      "loss": 3.9365,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663857811018742e-05,
+      "loss": 3.9409,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66301921626769e-05,
+      "loss": 3.9435,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621806215166383e-05,
+      "loss": 3.9198,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661343664645959e-05,
+      "loss": 3.9354,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660505069894907e-05,
+      "loss": 3.9456,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659666475143855e-05,
+      "loss": 3.9339,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658827880392803e-05,
+      "loss": 3.9276,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657989285641751e-05,
+      "loss": 3.9321,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657150690890699e-05,
+      "loss": 3.9254,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656312096139647e-05,
+      "loss": 3.9451,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655475139268968e-05,
+      "loss": 3.9265,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654636544517916e-05,
+      "loss": 3.9373,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653797949766864e-05,
+      "loss": 3.9364,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652959355015812e-05,
+      "loss": 3.9398,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652122398145134e-05,
+      "loss": 3.9302,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651283803394082e-05,
+      "loss": 3.937,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65044520864303e-05,
+      "loss": 3.9396,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649606613891978e-05,
+      "loss": 3.9308,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648768019140926e-05,
+      "loss": 3.9225,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647929424389874e-05,
+      "loss": 3.928,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647090829638822e-05,
+      "loss": 3.9219,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462538727681426e-05,
+      "loss": 3.9303,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454152780170906e-05,
+      "loss": 3.9245,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445766832660386e-05,
+      "loss": 3.9287,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437380885149866e-05,
+      "loss": 3.9163,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429011316443075e-05,
+      "loss": 3.9198,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420625368932555e-05,
+      "loss": 3.9179,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412239421422035e-05,
+      "loss": 3.9366,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640385347391152e-05,
+      "loss": 3.936,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639548390520473e-05,
+      "loss": 3.9266,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638709795769421e-05,
+      "loss": 3.9123,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637871201018369e-05,
+      "loss": 3.9255,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637032606267317e-05,
+      "loss": 3.9286,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636194011516265e-05,
+      "loss": 3.9255,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353554167652124e-05,
+      "loss": 3.9215,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345168220141604e-05,
+      "loss": 3.9275,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336782272631084e-05,
+      "loss": 3.9214,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632841270392429e-05,
+      "loss": 3.9296,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632004313521751e-05,
+      "loss": 3.9138,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631165718770699e-05,
+      "loss": 3.9243,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303271240196475e-05,
+      "loss": 3.9142,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629488529268595e-05,
+      "loss": 3.9065,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628649934517543e-05,
+      "loss": 3.9232,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6278129776468644e-05,
+      "loss": 3.9243,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626974382895812e-05,
+      "loss": 3.921,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62613578814476e-05,
+      "loss": 3.9152,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625297193393708e-05,
+      "loss": 3.9102,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.952582836151123,
+      "eval_runtime": 303.347,
+      "eval_samples_per_second": 1257.936,
+      "eval_steps_per_second": 39.311,
+      "step": 228960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.624458598642656e-05,
+      "loss": 3.9129,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.623620003891604e-05,
+      "loss": 3.9046,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.622781409140552e-05,
+      "loss": 3.9206,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6219428143895e-05,
+      "loss": 3.9078,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621104219638448e-05,
+      "loss": 3.9253,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.620265624887396e-05,
+      "loss": 3.9,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.619427030136344e-05,
+      "loss": 3.9175,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6185884353852924e-05,
+      "loss": 3.9011,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6177498406342404e-05,
+      "loss": 3.914,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6169112458831884e-05,
+      "loss": 3.9097,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6160726511321364e-05,
+      "loss": 3.9083,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.615235694261457e-05,
+      "loss": 3.9153,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.614398737390778e-05,
+      "loss": 3.8967,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.613560142639726e-05,
+      "loss": 3.9029,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.612721547888674e-05,
+      "loss": 3.9,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611882953137622e-05,
+      "loss": 3.9016,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.61104435838657e-05,
+      "loss": 3.9029,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.610205763635518e-05,
+      "loss": 3.8984,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.609368806764839e-05,
+      "loss": 3.8977,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.608530212013788e-05,
+      "loss": 3.9205,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.607691617262736e-05,
+      "loss": 3.903,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606853022511684e-05,
+      "loss": 3.9036,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606014427760631e-05,
+      "loss": 3.9036,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.605175833009579e-05,
+      "loss": 3.9082,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.604337238258527e-05,
+      "loss": 3.8878,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.603498643507475e-05,
+      "loss": 3.9085,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.602661686636796e-05,
+      "loss": 3.8957,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.601823091885744e-05,
+      "loss": 3.8938,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.600984497134692e-05,
+      "loss": 3.8848,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600147540264013e-05,
+      "loss": 3.891,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5993089455129615e-05,
+      "loss": 3.8942,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5984703507619095e-05,
+      "loss": 3.9082,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5976317560108575e-05,
+      "loss": 3.8992,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5967931612598055e-05,
+      "loss": 3.8968,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5959562043891264e-05,
+      "loss": 3.886,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5951176096380744e-05,
+      "loss": 3.8978,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5942790148870224e-05,
+      "loss": 3.8839,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5934404201359704e-05,
+      "loss": 3.9014,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5926018253849184e-05,
+      "loss": 3.8771,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.591764868514239e-05,
+      "loss": 3.8876,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590926273763187e-05,
+      "loss": 3.8852,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590087679012135e-05,
+      "loss": 3.8995,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.589249084261083e-05,
+      "loss": 3.8847,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.588410489510031e-05,
+      "loss": 3.8833,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58757189475898e-05,
+      "loss": 3.8879,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.586733300007928e-05,
+      "loss": 3.8851,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585894705256876e-05,
+      "loss": 3.891,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585057748386197e-05,
+      "loss": 3.8774,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584219153635145e-05,
+      "loss": 3.8733,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.583380558884093e-05,
+      "loss": 3.8898,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.582543602013414e-05,
+      "loss": 3.8808,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.581705007262362e-05,
+      "loss": 3.8786,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58086641251131e-05,
+      "loss": 3.8777,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580029455640631e-05,
+      "loss": 3.8822,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.579190860889579e-05,
+      "loss": 3.8648,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5783522661385267e-05,
+      "loss": 3.8851,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.577513671387475e-05,
+      "loss": 3.8641,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.576675076636423e-05,
+      "loss": 3.8817,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.575836481885371e-05,
+      "loss": 3.8787,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.574997887134319e-05,
+      "loss": 3.8792,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.57416093026364e-05,
+      "loss": 3.8732,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.573322335512588e-05,
+      "loss": 3.8808,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.572483740761536e-05,
+      "loss": 3.8575,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.571645146010484e-05,
+      "loss": 3.8701,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.570808189139805e-05,
+      "loss": 3.8798,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569971232269126e-05,
+      "loss": 3.8754,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569132637518074e-05,
+      "loss": 3.8654,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.568294042767022e-05,
+      "loss": 3.865,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.567455448015971e-05,
+      "loss": 3.8633,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5666184911452916e-05,
+      "loss": 3.874,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5657798963942396e-05,
+      "loss": 3.8818,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5649413016431876e-05,
+      "loss": 3.8659,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5641027068921356e-05,
+      "loss": 3.8827,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5632641121410836e-05,
+      "loss": 3.8828,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5624255173900316e-05,
+      "loss": 3.8904,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5615885605193525e-05,
+      "loss": 3.855,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5607499657683005e-05,
+      "loss": 3.876,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5599113710172485e-05,
+      "loss": 3.8762,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5590727762661965e-05,
+      "loss": 3.861,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5582341815151445e-05,
+      "loss": 3.8703,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5573955867640925e-05,
+      "loss": 3.8757,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.556558629893414e-05,
+      "loss": 3.878,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5557200351423614e-05,
+      "loss": 3.869,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5548814403913094e-05,
+      "loss": 3.8598,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5540428456402574e-05,
+      "loss": 3.8594,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5532042508892054e-05,
+      "loss": 3.863,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.552367294018526e-05,
+      "loss": 3.8713,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.551528699267474e-05,
+      "loss": 3.8716,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.550690104516422e-05,
+      "loss": 3.8549,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54985150976537e-05,
+      "loss": 3.8675,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549012915014318e-05,
+      "loss": 3.8525,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548174320263266e-05,
+      "loss": 3.862,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.547337363392588e-05,
+      "loss": 3.8651,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546498768641536e-05,
+      "loss": 3.8615,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.545660173890484e-05,
+      "loss": 3.8705,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544821579139432e-05,
+      "loss": 3.865,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54398298438838e-05,
+      "loss": 3.8575,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543144389637328e-05,
+      "loss": 3.865,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.542305794886276e-05,
+      "loss": 3.8646,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.541468838015597e-05,
+      "loss": 3.8663,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.540630243264545e-05,
+      "loss": 3.8505,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.539791648513493e-05,
+      "loss": 3.8576,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538953053762441e-05,
+      "loss": 3.8606,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5381160968917616e-05,
+      "loss": 3.8666,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5372775021407096e-05,
+      "loss": 3.8452,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5364389073896576e-05,
+      "loss": 3.8522,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.535600312638606e-05,
+      "loss": 3.8665,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.534761717887554e-05,
+      "loss": 3.8573,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533924761016875e-05,
+      "loss": 3.8481,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533086166265823e-05,
+      "loss": 3.8504,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.532247571514771e-05,
+      "loss": 3.8543,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.531408976763719e-05,
+      "loss": 3.863,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.530570382012667e-05,
+      "loss": 3.8535,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.529733425141988e-05,
+      "loss": 3.8536,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528894830390936e-05,
+      "loss": 3.8609,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528056235639884e-05,
+      "loss": 3.8632,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.527217640888832e-05,
+      "loss": 3.8606,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.52637904613778e-05,
+      "loss": 3.8523,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5255420892671017e-05,
+      "loss": 3.8664,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5247034945160496e-05,
+      "loss": 3.8556,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5238648997649976e-05,
+      "loss": 3.8444,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523026305013945e-05,
+      "loss": 3.8522,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.522187710262893e-05,
+      "loss": 3.8477,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5213507533922145e-05,
+      "loss": 3.8557,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5205121586411625e-05,
+      "loss": 3.847,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51967356389011e-05,
+      "loss": 3.8556,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.518834969139058e-05,
+      "loss": 3.8424,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517996374388006e-05,
+      "loss": 3.8446,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5171594175173274e-05,
+      "loss": 3.8446,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5163208227662754e-05,
+      "loss": 3.8645,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5154822280152234e-05,
+      "loss": 3.8565,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5146436332641714e-05,
+      "loss": 3.8562,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.513806676393492e-05,
+      "loss": 3.8381,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51296808164244e-05,
+      "loss": 3.8531,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512129486891388e-05,
+      "loss": 3.8574,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.511290892140336e-05,
+      "loss": 3.8483,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.510452297389284e-05,
+      "loss": 3.8486,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.509615340518605e-05,
+      "loss": 3.8533,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.508776745767553e-05,
+      "loss": 3.8485,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507938151016501e-05,
+      "loss": 3.8574,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507099556265449e-05,
+      "loss": 3.8376,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.506260961514397e-05,
+      "loss": 3.8519,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.505424004643719e-05,
+      "loss": 3.8413,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.504585409892667e-05,
+      "loss": 3.8393,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503746815141615e-05,
+      "loss": 3.8487,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502908220390563e-05,
+      "loss": 3.8536,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502071263519884e-05,
+      "loss": 3.8479,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.501232668768832e-05,
+      "loss": 3.8392,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.50039407401778e-05,
+      "loss": 3.8427,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.911499261856079,
+      "eval_runtime": 309.2488,
+      "eval_samples_per_second": 1233.929,
+      "eval_steps_per_second": 38.561,
+      "step": 305280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.499555479266728e-05,
+      "loss": 3.8433,
+      "step": 305664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4987185223960486e-05,
+      "loss": 3.8367,
+      "step": 306176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4978799276449966e-05,
+      "loss": 3.8438,
+      "step": 306688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4970413328939446e-05,
+      "loss": 3.8396,
+      "step": 307200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4962027381428926e-05,
+      "loss": 3.8533,
+      "step": 307712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.495365781272214e-05,
+      "loss": 3.8351,
+      "step": 308224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.494527186521162e-05,
+      "loss": 3.8415,
+      "step": 308736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.49368859177011e-05,
+      "loss": 3.8315,
+      "step": 309248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492849997019058e-05,
+      "loss": 3.8419,
+      "step": 309760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492013040148379e-05,
+      "loss": 3.8426,
+      "step": 310272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491174445397327e-05,
+      "loss": 3.8385,
+      "step": 310784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.490335850646275e-05,
+      "loss": 3.8418,
+      "step": 311296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.489497255895223e-05,
+      "loss": 3.8291,
+      "step": 311808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.488658661144171e-05,
+      "loss": 3.8338,
+      "step": 312320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.487821704273492e-05,
+      "loss": 3.8334,
+      "step": 312832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.48698310952244e-05,
+      "loss": 3.8311,
+      "step": 313344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486144514771388e-05,
+      "loss": 3.8333,
+      "step": 313856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.485305920020336e-05,
+      "loss": 3.8318,
+      "step": 314368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4844689631496575e-05,
+      "loss": 3.8297,
+      "step": 314880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4836303683986055e-05,
+      "loss": 3.8475,
+      "step": 315392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4827917736475535e-05,
+      "loss": 3.8381,
+      "step": 315904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4819531788965015e-05,
+      "loss": 3.8346,
+      "step": 316416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4811145841454495e-05,
+      "loss": 3.8375,
+      "step": 316928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4802776272747704e-05,
+      "loss": 3.8385,
+      "step": 317440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4794390325237184e-05,
+      "loss": 3.8202,
+      "step": 317952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4786004377726664e-05,
+      "loss": 3.8385,
+      "step": 318464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4777618430216144e-05,
+      "loss": 3.8265,
+      "step": 318976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4769232482705624e-05,
+      "loss": 3.8231,
+      "step": 319488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476086291399883e-05,
+      "loss": 3.8176,
+      "step": 320000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.475247696648831e-05,
+      "loss": 3.8249,
+      "step": 320512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.474409101897779e-05,
+      "loss": 3.8272,
+      "step": 321024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.473572145027101e-05,
+      "loss": 3.8427,
+      "step": 321536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.472733550276049e-05,
+      "loss": 3.8324,
+      "step": 322048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471894955524997e-05,
+      "loss": 3.8289,
+      "step": 322560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471056360773945e-05,
+      "loss": 3.8203,
+      "step": 323072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.470217766022893e-05,
+      "loss": 3.831,
+      "step": 323584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.469380809152214e-05,
+      "loss": 3.8261,
+      "step": 324096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.468542214401162e-05,
+      "loss": 3.8307,
+      "step": 324608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.46770361965011e-05,
+      "loss": 3.8112,
+      "step": 325120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466866662779431e-05,
+      "loss": 3.8201,
+      "step": 325632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466028068028379e-05,
+      "loss": 3.8188,
+      "step": 326144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.465189473277327e-05,
+      "loss": 3.8348,
+      "step": 326656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.464350878526275e-05,
+      "loss": 3.8225,
+      "step": 327168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.463512283775223e-05,
+      "loss": 3.8189,
+      "step": 327680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.462675326904544e-05,
+      "loss": 3.824,
+      "step": 328192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.461836732153492e-05,
+      "loss": 3.8195,
+      "step": 328704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.46099813740244e-05,
+      "loss": 3.8254,
+      "step": 329216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.460159542651388e-05,
+      "loss": 3.8152,
+      "step": 329728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.459322585780709e-05,
+      "loss": 3.8097,
+      "step": 330240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.458483991029657e-05,
+      "loss": 3.8273,
+      "step": 330752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.457645396278605e-05,
+      "loss": 3.8143,
+      "step": 331264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4568068015275525e-05,
+      "loss": 3.8169,
+      "step": 331776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.455969844656874e-05,
+      "loss": 3.8116,
+      "step": 332288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.455131249905822e-05,
+      "loss": 3.8208,
+      "step": 332800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.45429265515477e-05,
+      "loss": 3.8021,
+      "step": 333312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.453454060403718e-05,
+      "loss": 3.8183,
+      "step": 333824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4526171035330396e-05,
+      "loss": 3.8049,
+      "step": 334336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4517785087819876e-05,
+      "loss": 3.8148,
+      "step": 334848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4509399140309356e-05,
+      "loss": 3.8186,
+      "step": 335360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450101319279883e-05,
+      "loss": 3.8147,
+      "step": 335872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4492643624092045e-05,
+      "loss": 3.8107,
+      "step": 336384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4484257676581525e-05,
+      "loss": 3.8217,
+      "step": 336896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4475871729071e-05,
+      "loss": 3.7982,
+      "step": 337408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.446748578156048e-05,
+      "loss": 3.804,
+      "step": 337920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4459116212853694e-05,
+      "loss": 3.8196,
+      "step": 338432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4450730265343174e-05,
+      "loss": 3.8113,
+      "step": 338944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4442344317832654e-05,
+      "loss": 3.8074,
+      "step": 339456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4433958370322134e-05,
+      "loss": 3.801,
+      "step": 339968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.442558880161535e-05,
+      "loss": 3.8011,
+      "step": 340480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.441720285410482e-05,
+      "loss": 3.81,
+      "step": 340992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.44088169065943e-05,
+      "loss": 3.8203,
+      "step": 341504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.440043095908378e-05,
+      "loss": 3.8086,
+      "step": 342016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4392061390377e-05,
+      "loss": 3.8192,
+      "step": 342528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.438367544286647e-05,
+      "loss": 3.818,
+      "step": 343040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.437528949535595e-05,
+      "loss": 3.8332,
+      "step": 343552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.436690354784543e-05,
+      "loss": 3.7972,
+      "step": 344064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.435853397913865e-05,
+      "loss": 3.8116,
+      "step": 344576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.435014803162813e-05,
+      "loss": 3.8196,
+      "step": 345088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434176208411761e-05,
+      "loss": 3.7959,
+      "step": 345600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.433337613660709e-05,
+      "loss": 3.814,
+      "step": 346112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.43250065679003e-05,
+      "loss": 3.8128,
+      "step": 346624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.431662062038978e-05,
+      "loss": 3.8158,
+      "step": 347136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4308234672879257e-05,
+      "loss": 3.8108,
+      "step": 347648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4299848725368737e-05,
+      "loss": 3.798,
+      "step": 348160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4291479156661946e-05,
+      "loss": 3.7993,
+      "step": 348672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4283093209151426e-05,
+      "loss": 3.8028,
+      "step": 349184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4274707261640906e-05,
+      "loss": 3.811,
+      "step": 349696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4266321314130386e-05,
+      "loss": 3.8123,
+      "step": 350208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.42579517454236e-05,
+      "loss": 3.7961,
+      "step": 350720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424956579791308e-05,
+      "loss": 3.8051,
+      "step": 351232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424117985040256e-05,
+      "loss": 3.7938,
+      "step": 351744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.423279390289204e-05,
+      "loss": 3.8014,
+      "step": 352256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.422442433418525e-05,
+      "loss": 3.8068,
+      "step": 352768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.421603838667473e-05,
+      "loss": 3.8071,
+      "step": 353280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.420765243916421e-05,
+      "loss": 3.8135,
+      "step": 353792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.419926649165369e-05,
+      "loss": 3.8043,
+      "step": 354304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.41908969229469e-05,
+      "loss": 3.8008,
+      "step": 354816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.418251097543638e-05,
+      "loss": 3.8049,
+      "step": 355328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.417412502792586e-05,
+      "loss": 3.8041,
+      "step": 355840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.416573908041534e-05,
+      "loss": 3.8104,
+      "step": 356352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4157369511708555e-05,
+      "loss": 3.7939,
+      "step": 356864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4148983564198035e-05,
+      "loss": 3.8019,
+      "step": 357376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4140597616687515e-05,
+      "loss": 3.801,
+      "step": 357888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4132211669176995e-05,
+      "loss": 3.8096,
+      "step": 358400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4123842100470204e-05,
+      "loss": 3.7887,
+      "step": 358912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4115456152959684e-05,
+      "loss": 3.7956,
+      "step": 359424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4107070205449164e-05,
+      "loss": 3.8094,
+      "step": 359936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4098684257938644e-05,
+      "loss": 3.7996,
+      "step": 360448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.409033106803559e-05,
+      "loss": 3.7894,
+      "step": 360960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.408194512052507e-05,
+      "loss": 3.7995,
+      "step": 361472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.407355917301454e-05,
+      "loss": 3.7916,
+      "step": 361984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.406517322550402e-05,
+      "loss": 3.8101,
+      "step": 362496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.405678727799351e-05,
+      "loss": 3.7957,
+      "step": 363008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.404840133048299e-05,
+      "loss": 3.7924,
+      "step": 363520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.404001538297247e-05,
+      "loss": 3.8052,
+      "step": 364032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403162943546195e-05,
+      "loss": 3.8106,
+      "step": 364544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.402325986675516e-05,
+      "loss": 3.802,
+      "step": 365056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.401487391924464e-05,
+      "loss": 3.7975,
+      "step": 365568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.400648797173412e-05,
+      "loss": 3.8051,
+      "step": 366080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.39981020242236e-05,
+      "loss": 3.7978,
+      "step": 366592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398973245551681e-05,
+      "loss": 3.7902,
+      "step": 367104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3981346508006287e-05,
+      "loss": 3.8021,
+      "step": 367616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3972960560495767e-05,
+      "loss": 3.7892,
+      "step": 368128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3964574612985247e-05,
+      "loss": 3.7986,
+      "step": 368640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3956205044278456e-05,
+      "loss": 3.7915,
+      "step": 369152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.394781909676794e-05,
+      "loss": 3.798,
+      "step": 369664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.393943314925742e-05,
+      "loss": 3.7867,
+      "step": 370176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.39310472017469e-05,
+      "loss": 3.787,
+      "step": 370688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.392267763304011e-05,
+      "loss": 3.794,
+      "step": 371200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.391429168552959e-05,
+      "loss": 3.8037,
+      "step": 371712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.390590573801907e-05,
+      "loss": 3.8026,
+      "step": 372224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.389751979050855e-05,
+      "loss": 3.8016,
+      "step": 372736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388915022180176e-05,
+      "loss": 3.7812,
+      "step": 373248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388076427429124e-05,
+      "loss": 3.8011,
+      "step": 373760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.387237832678072e-05,
+      "loss": 3.7996,
+      "step": 374272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38639923792702e-05,
+      "loss": 3.7976,
+      "step": 374784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.385562281056341e-05,
+      "loss": 3.7888,
+      "step": 375296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3847236863052896e-05,
+      "loss": 3.803,
+      "step": 375808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3838850915542376e-05,
+      "loss": 3.7904,
+      "step": 376320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3830464968031856e-05,
+      "loss": 3.804,
+      "step": 376832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3822095399325065e-05,
+      "loss": 3.7799,
+      "step": 377344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3813709451814545e-05,
+      "loss": 3.803,
+      "step": 377856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3805323504304025e-05,
+      "loss": 3.7835,
+      "step": 378368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3796937556793505e-05,
+      "loss": 3.7865,
+      "step": 378880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3788567988086714e-05,
+      "loss": 3.7973,
+      "step": 379392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3780182040576194e-05,
+      "loss": 3.7954,
+      "step": 379904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3771796093065674e-05,
+      "loss": 3.7979,
+      "step": 380416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.376342652435888e-05,
+      "loss": 3.781,
+      "step": 380928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.375504057684836e-05,
+      "loss": 3.795,
+      "step": 381440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8860723972320557,
+      "eval_runtime": 307.2775,
+      "eval_samples_per_second": 1241.845,
+      "eval_steps_per_second": 38.809,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374665462933785e-05,
+      "loss": 3.7855,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373826868182733e-05,
+      "loss": 3.7835,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372988273431681e-05,
+      "loss": 3.7927,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372149678680629e-05,
+      "loss": 3.7856,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371311083929577e-05,
+      "loss": 3.7986,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370472489178525e-05,
+      "loss": 3.7817,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369633894427473e-05,
+      "loss": 3.7828,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36879529967642e-05,
+      "loss": 3.783,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367958342805742e-05,
+      "loss": 3.7888,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36711974805469e-05,
+      "loss": 3.7901,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366281153303637e-05,
+      "loss": 3.7853,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365442558552585e-05,
+      "loss": 3.793,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364605601681907e-05,
+      "loss": 3.776,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363767006930855e-05,
+      "loss": 3.7807,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362928412179803e-05,
+      "loss": 3.7833,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362089817428751e-05,
+      "loss": 3.775,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361252860558072e-05,
+      "loss": 3.7827,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36041426580702e-05,
+      "loss": 3.7775,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3595756710559676e-05,
+      "loss": 3.7758,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3587370763049156e-05,
+      "loss": 3.798,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357900119434237e-05,
+      "loss": 3.7848,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3570615246831845e-05,
+      "loss": 3.789,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3562229299321325e-05,
+      "loss": 3.783,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3553843351810805e-05,
+      "loss": 3.7869,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354547378310402e-05,
+      "loss": 3.7684,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35370878355935e-05,
+      "loss": 3.7891,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352870188808298e-05,
+      "loss": 3.7776,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35203323193762e-05,
+      "loss": 3.7687,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351194637186568e-05,
+      "loss": 3.7681,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350356042435515e-05,
+      "loss": 3.7716,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349517447684463e-05,
+      "loss": 3.7789,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3486804908137846e-05,
+      "loss": 3.7885,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347841896062732e-05,
+      "loss": 3.787,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34700330131168e-05,
+      "loss": 3.7782,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346164706560628e-05,
+      "loss": 3.773,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3453277496899495e-05,
+      "loss": 3.7782,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3444891549388975e-05,
+      "loss": 3.775,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3436505601878455e-05,
+      "loss": 3.7799,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3428119654367935e-05,
+      "loss": 3.7656,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341975008566115e-05,
+      "loss": 3.7658,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341138051695436e-05,
+      "loss": 3.7706,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.340299456944384e-05,
+      "loss": 3.7829,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339460862193332e-05,
+      "loss": 3.7754,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338622267442279e-05,
+      "loss": 3.7686,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337783672691227e-05,
+      "loss": 3.7763,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336945077940175e-05,
+      "loss": 3.7663,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336106483189123e-05,
+      "loss": 3.7774,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335267888438071e-05,
+      "loss": 3.766,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334430931567393e-05,
+      "loss": 3.7571,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333592336816341e-05,
+      "loss": 3.7764,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332753742065289e-05,
+      "loss": 3.7689,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331915147314237e-05,
+      "loss": 3.7672,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331078190443558e-05,
+      "loss": 3.7679,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330239595692506e-05,
+      "loss": 3.7679,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329401000941454e-05,
+      "loss": 3.7554,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328562406190402e-05,
+      "loss": 3.7664,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3277254493197226e-05,
+      "loss": 3.7585,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3268868545686706e-05,
+      "loss": 3.7652,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3260482598176186e-05,
+      "loss": 3.7726,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3252096650665666e-05,
+      "loss": 3.7626,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324372708195888e-05,
+      "loss": 3.7634,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323534113444836e-05,
+      "loss": 3.7736,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322695518693784e-05,
+      "loss": 3.7534,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321856923942732e-05,
+      "loss": 3.7556,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321019967072053e-05,
+      "loss": 3.7661,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320181372321001e-05,
+      "loss": 3.7673,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319342777569949e-05,
+      "loss": 3.7564,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318504182818897e-05,
+      "loss": 3.7579,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317667225948218e-05,
+      "loss": 3.7524,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316828631197166e-05,
+      "loss": 3.7623,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315990036446114e-05,
+      "loss": 3.7693,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315151441695062e-05,
+      "loss": 3.763,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3143144848243836e-05,
+      "loss": 3.7716,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3134758900733316e-05,
+      "loss": 3.7653,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3126372953222796e-05,
+      "loss": 3.7916,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3117987005712276e-05,
+      "loss": 3.7495,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109617437005485e-05,
+      "loss": 3.7635,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3101231489494965e-05,
+      "loss": 3.7721,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3092845541984445e-05,
+      "loss": 3.7492,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3084459594473925e-05,
+      "loss": 3.7669,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3076090025767134e-05,
+      "loss": 3.7674,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067704078256614e-05,
+      "loss": 3.7669,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3059318130746094e-05,
+      "loss": 3.7634,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3050932183235574e-05,
+      "loss": 3.7516,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304256261452879e-05,
+      "loss": 3.7606,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3034193045822e-05,
+      "loss": 3.7516,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.302580709831148e-05,
+      "loss": 3.7648,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.301742115080096e-05,
+      "loss": 3.7684,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300903520329044e-05,
+      "loss": 3.7461,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300064925577992e-05,
+      "loss": 3.762,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29922633082694e-05,
+      "loss": 3.7513,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298387736075888e-05,
+      "loss": 3.7529,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297549141324836e-05,
+      "loss": 3.7563,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296712184454157e-05,
+      "loss": 3.768,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295873589703105e-05,
+      "loss": 3.7691,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295034994952053e-05,
+      "loss": 3.7539,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294196400201001e-05,
+      "loss": 3.7535,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293359443330322e-05,
+      "loss": 3.7611,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29252084857927e-05,
+      "loss": 3.7553,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291682253828218e-05,
+      "loss": 3.767,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290843659077166e-05,
+      "loss": 3.7489,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290006702206487e-05,
+      "loss": 3.7535,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289168107455435e-05,
+      "loss": 3.7572,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288329512704383e-05,
+      "loss": 3.7645,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287490917953331e-05,
+      "loss": 3.7432,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286653961082652e-05,
+      "loss": 3.7505,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2858153663316e-05,
+      "loss": 3.7602,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284976771580548e-05,
+      "loss": 3.759,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28413981470987e-05,
+      "loss": 3.7422,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283301219958818e-05,
+      "loss": 3.7594,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282462625207766e-05,
+      "loss": 3.7448,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281624030456714e-05,
+      "loss": 3.7601,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2807854357056617e-05,
+      "loss": 3.7554,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2799468409546097e-05,
+      "loss": 3.7493,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2791082462035576e-05,
+      "loss": 3.7601,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278269651452505e-05,
+      "loss": 3.7676,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2774326945818266e-05,
+      "loss": 3.7565,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2765940998307746e-05,
+      "loss": 3.7534,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2757555050797225e-05,
+      "loss": 3.761,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27491691032867e-05,
+      "loss": 3.7548,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2740799534579915e-05,
+      "loss": 3.7467,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2732413587069394e-05,
+      "loss": 3.7535,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2724027639558874e-05,
+      "loss": 3.7465,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715641692048354e-05,
+      "loss": 3.7541,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270727212334157e-05,
+      "loss": 3.7529,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269888617583105e-05,
+      "loss": 3.7523,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2690500228320523e-05,
+      "loss": 3.7467,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268211428081e-05,
+      "loss": 3.7416,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267374471210322e-05,
+      "loss": 3.7524,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26653587645927e-05,
+      "loss": 3.7586,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265697281708217e-05,
+      "loss": 3.7565,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264858686957165e-05,
+      "loss": 3.7603,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264021730086487e-05,
+      "loss": 3.735,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263183135335435e-05,
+      "loss": 3.761,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262344540584383e-05,
+      "loss": 3.7544,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261505945833331e-05,
+      "loss": 3.7504,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2606689889626524e-05,
+      "loss": 3.7513,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2598303942116e-05,
+      "loss": 3.7546,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258991799460548e-05,
+      "loss": 3.7487,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258154842589869e-05,
+      "loss": 3.7635,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257316247838817e-05,
+      "loss": 3.7362,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2564776530877646e-05,
+      "loss": 3.7562,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255640696217086e-05,
+      "loss": 3.7399,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254802101466034e-05,
+      "loss": 3.7448,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253963506714982e-05,
+      "loss": 3.7532,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25312491196393e-05,
+      "loss": 3.7488,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252286317212878e-05,
+      "loss": 3.7615,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251447722461826e-05,
+      "loss": 3.7331,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250609127710774e-05,
+      "loss": 3.7548,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8695645332336426,
+      "eval_runtime": 304.0955,
+      "eval_samples_per_second": 1254.84,
+      "eval_steps_per_second": 39.215,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.249772170840095e-05,
+      "loss": 3.7482,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248935213969417e-05,
+      "loss": 3.7382,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248096619218364e-05,
+      "loss": 3.7502,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.247258024467312e-05,
+      "loss": 3.7449,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.24641942971626e-05,
+      "loss": 3.753,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245580834965208e-05,
+      "loss": 3.7413,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244742240214156e-05,
+      "loss": 3.7399,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243903645463104e-05,
+      "loss": 3.7424,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243065050712052e-05,
+      "loss": 3.7445,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2422280938413735e-05,
+      "loss": 3.7448,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2413894990903215e-05,
+      "loss": 3.7461,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2405509043392695e-05,
+      "loss": 3.7454,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2397139474685904e-05,
+      "loss": 3.7399,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2388753527175384e-05,
+      "loss": 3.7385,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2380367579664864e-05,
+      "loss": 3.7424,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2371981632154344e-05,
+      "loss": 3.7349,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2363595684643824e-05,
+      "loss": 3.7403,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2355209737133304e-05,
+      "loss": 3.7355,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2346823789622784e-05,
+      "loss": 3.7357,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2338437842112264e-05,
+      "loss": 3.7543,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233006827340547e-05,
+      "loss": 3.7458,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232168232589496e-05,
+      "loss": 3.7461,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.231329637838444e-05,
+      "loss": 3.7406,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.230491043087392e-05,
+      "loss": 3.7453,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.229654086216713e-05,
+      "loss": 3.7284,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.228815491465661e-05,
+      "loss": 3.7496,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227976896714609e-05,
+      "loss": 3.7367,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227138301963557e-05,
+      "loss": 3.729,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.226301345092878e-05,
+      "loss": 3.7294,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.225462750341826e-05,
+      "loss": 3.7322,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.224624155590774e-05,
+      "loss": 3.7368,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.223787198720095e-05,
+      "loss": 3.748,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222948603969043e-05,
+      "loss": 3.7449,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2221100092179914e-05,
+      "loss": 3.7386,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2212714144669394e-05,
+      "loss": 3.7277,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.22043445759626e-05,
+      "loss": 3.7401,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.219595862845208e-05,
+      "loss": 3.7376,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.218757268094156e-05,
+      "loss": 3.7324,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217918673343104e-05,
+      "loss": 3.7299,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217081716472425e-05,
+      "loss": 3.7248,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.216243121721373e-05,
+      "loss": 3.7325,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.215406164850694e-05,
+      "loss": 3.741,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.214567570099642e-05,
+      "loss": 3.7359,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.21372897534859e-05,
+      "loss": 3.7267,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212890380597538e-05,
+      "loss": 3.7358,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212051785846487e-05,
+      "loss": 3.7269,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.211213191095435e-05,
+      "loss": 3.7388,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.210374596344383e-05,
+      "loss": 3.7259,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2095376394737036e-05,
+      "loss": 3.7141,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2086990447226516e-05,
+      "loss": 3.7372,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2078604499715996e-05,
+      "loss": 3.7306,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2070218552205476e-05,
+      "loss": 3.7306,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2061848983498685e-05,
+      "loss": 3.726,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2053463035988165e-05,
+      "loss": 3.7253,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2045077088477645e-05,
+      "loss": 3.7194,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2036691140967125e-05,
+      "loss": 3.7249,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2028321572260334e-05,
+      "loss": 3.7214,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.201993562474982e-05,
+      "loss": 3.7278,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.20115496772393e-05,
+      "loss": 3.7316,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.200316372972878e-05,
+      "loss": 3.7244,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199479416102199e-05,
+      "loss": 3.7209,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.198640821351147e-05,
+      "loss": 3.7419,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197802226600095e-05,
+      "loss": 3.7103,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196963631849042e-05,
+      "loss": 3.7157,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196126674978364e-05,
+      "loss": 3.7264,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195288080227312e-05,
+      "loss": 3.7322,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19444948547626e-05,
+      "loss": 3.7173,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.193610890725207e-05,
+      "loss": 3.7177,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1927755717349024e-05,
+      "loss": 3.7137,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1919369769838504e-05,
+      "loss": 3.7227,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1910983822327984e-05,
+      "loss": 3.7342,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1902597874817464e-05,
+      "loss": 3.7228,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1894211927306944e-05,
+      "loss": 3.7287,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1885825979796424e-05,
+      "loss": 3.7317,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18774400322859e-05,
+      "loss": 3.7485,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186905408477538e-05,
+      "loss": 3.7152,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186070089487232e-05,
+      "loss": 3.7268,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18523149473618e-05,
+      "loss": 3.7304,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.184392899985128e-05,
+      "loss": 3.7116,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.183554305234076e-05,
+      "loss": 3.7266,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.182715710483024e-05,
+      "loss": 3.7295,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181877115731972e-05,
+      "loss": 3.7298,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181040158861294e-05,
+      "loss": 3.7277,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180201564110242e-05,
+      "loss": 3.7131,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.17936296935919e-05,
+      "loss": 3.7189,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.178524374608137e-05,
+      "loss": 3.7157,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.177685779857085e-05,
+      "loss": 3.7238,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.176847185106033e-05,
+      "loss": 3.7289,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.176008590354981e-05,
+      "loss": 3.7109,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.175171633484302e-05,
+      "loss": 3.7235,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.17433303873325e-05,
+      "loss": 3.7145,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.173494443982198e-05,
+      "loss": 3.7117,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1726558492311466e-05,
+      "loss": 3.7225,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1718188923604675e-05,
+      "loss": 3.7245,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1709802976094155e-05,
+      "loss": 3.7328,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1701417028583635e-05,
+      "loss": 3.7206,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1693031081073115e-05,
+      "loss": 3.7151,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1684661512366324e-05,
+      "loss": 3.7228,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1676275564855804e-05,
+      "loss": 3.7181,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1667889617345284e-05,
+      "loss": 3.7255,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1659503669834764e-05,
+      "loss": 3.7144,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165113410112797e-05,
+      "loss": 3.7149,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.164274815361745e-05,
+      "loss": 3.7183,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.163437858491067e-05,
+      "loss": 3.7316,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.162599263740015e-05,
+      "loss": 3.7037,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.161760668988963e-05,
+      "loss": 3.7179,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160922074237911e-05,
+      "loss": 3.719,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160083479486859e-05,
+      "loss": 3.7236,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.15924652261618e-05,
+      "loss": 3.705,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.158407927865128e-05,
+      "loss": 3.7206,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.157569333114076e-05,
+      "loss": 3.7162,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.156730738363024e-05,
+      "loss": 3.7173,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155892143611972e-05,
+      "loss": 3.7238,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.15505354886092e-05,
+      "loss": 3.7112,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154214954109868e-05,
+      "loss": 3.7182,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.153376359358816e-05,
+      "loss": 3.732,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.152539402488137e-05,
+      "loss": 3.7217,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151700807737085e-05,
+      "loss": 3.7127,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150863850866406e-05,
+      "loss": 3.7272,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150025256115354e-05,
+      "loss": 3.7174,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.149186661364302e-05,
+      "loss": 3.7052,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14834806661325e-05,
+      "loss": 3.7209,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.147509471862198e-05,
+      "loss": 3.709,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.146670877111146e-05,
+      "loss": 3.714,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.145832282360094e-05,
+      "loss": 3.7166,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144993687609042e-05,
+      "loss": 3.7158,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144156730738363e-05,
+      "loss": 3.7123,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.143318135987311e-05,
+      "loss": 3.7084,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.142479541236259e-05,
+      "loss": 3.7108,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.141640946485207e-05,
+      "loss": 3.7251,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.140803989614529e-05,
+      "loss": 3.7161,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139965394863477e-05,
+      "loss": 3.7225,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1391284379927976e-05,
+      "loss": 3.6986,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1382898432417456e-05,
+      "loss": 3.7297,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1374512484906936e-05,
+      "loss": 3.7165,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1366126537396416e-05,
+      "loss": 3.7174,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1357740589885896e-05,
+      "loss": 3.7167,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1349354642375376e-05,
+      "loss": 3.7166,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1340968694864856e-05,
+      "loss": 3.7098,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.133258274735433e-05,
+      "loss": 3.7305,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1324213178647545e-05,
+      "loss": 3.702,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1315827231137025e-05,
+      "loss": 3.7176,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.130745766243024e-05,
+      "loss": 3.7085,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129907171491972e-05,
+      "loss": 3.7097,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.12906857674092e-05,
+      "loss": 3.7125,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128229981989868e-05,
+      "loss": 3.7107,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1273913872388154e-05,
+      "loss": 3.7263,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1265527924877634e-05,
+      "loss": 3.6995,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.125715835617085e-05,
+      "loss": 3.7195,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8594038486480713,
+      "eval_runtime": 303.7587,
+      "eval_samples_per_second": 1256.231,
+      "eval_steps_per_second": 39.258,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124877240866032e-05,
+      "loss": 3.7142,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.12403864611498e-05,
+      "loss": 3.7027,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123200051363928e-05,
+      "loss": 3.7143,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.122361456612876e-05,
+      "loss": 3.7112,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.121524499742198e-05,
+      "loss": 3.717,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.120685904991146e-05,
+      "loss": 3.7056,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119847310240094e-05,
+      "loss": 3.7084,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119008715489042e-05,
+      "loss": 3.7058,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11817012073799e-05,
+      "loss": 3.7138,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117333163867311e-05,
+      "loss": 3.7108,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.116494569116259e-05,
+      "loss": 3.7136,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115655974365207e-05,
+      "loss": 3.7071,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114817379614155e-05,
+      "loss": 3.7052,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113978784863103e-05,
+      "loss": 3.7028,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1131418279924236e-05,
+      "loss": 3.7096,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1123032332413716e-05,
+      "loss": 3.6986,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1114646384903196e-05,
+      "loss": 3.7039,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110626043739268e-05,
+      "loss": 3.7033,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109787448988216e-05,
+      "loss": 3.7031,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108948854237164e-05,
+      "loss": 3.714,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108110259486112e-05,
+      "loss": 3.7146,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10727166473506e-05,
+      "loss": 3.7142,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106434707864381e-05,
+      "loss": 3.7078,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.105596113113329e-05,
+      "loss": 3.7089,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.104757518362277e-05,
+      "loss": 3.6952,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103918923611225e-05,
+      "loss": 3.7167,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103081966740546e-05,
+      "loss": 3.7007,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.102243371989494e-05,
+      "loss": 3.6945,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101406415118815e-05,
+      "loss": 3.6951,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1005678203677637e-05,
+      "loss": 3.6996,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0997292256167117e-05,
+      "loss": 3.7013,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0988906308656597e-05,
+      "loss": 3.7146,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0980520361146076e-05,
+      "loss": 3.7111,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0972134413635556e-05,
+      "loss": 3.7014,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0963748466125036e-05,
+      "loss": 3.6976,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0955362518614516e-05,
+      "loss": 3.7086,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0946992949907725e-05,
+      "loss": 3.7068,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0938607002397205e-05,
+      "loss": 3.6993,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0930221054886685e-05,
+      "loss": 3.6969,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092183510737616e-05,
+      "loss": 3.6927,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0913465538669374e-05,
+      "loss": 3.6965,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0905095969962583e-05,
+      "loss": 3.7092,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.089671002245207e-05,
+      "loss": 3.7022,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.088832407494155e-05,
+      "loss": 3.6954,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087993812743103e-05,
+      "loss": 3.7028,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087155217992051e-05,
+      "loss": 3.6926,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.086316623240999e-05,
+      "loss": 3.706,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085478028489946e-05,
+      "loss": 3.6958,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084641071619268e-05,
+      "loss": 3.6781,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083802476868216e-05,
+      "loss": 3.7044,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082965519997537e-05,
+      "loss": 3.6959,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082126925246485e-05,
+      "loss": 3.7,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081288330495433e-05,
+      "loss": 3.692,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080449735744381e-05,
+      "loss": 3.6924,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079611140993329e-05,
+      "loss": 3.6852,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078772546242277e-05,
+      "loss": 3.691,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077933951491225e-05,
+      "loss": 3.691,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0770969946205464e-05,
+      "loss": 3.6934,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076258399869494e-05,
+      "loss": 3.7059,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075419805118442e-05,
+      "loss": 3.6852,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.07458121036739e-05,
+      "loss": 3.6915,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0737442534967106e-05,
+      "loss": 3.7046,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0729056587456586e-05,
+      "loss": 3.6795,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0720670639946066e-05,
+      "loss": 3.682,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0712284692435546e-05,
+      "loss": 3.6969,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.070391512372876e-05,
+      "loss": 3.6979,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069552917621824e-05,
+      "loss": 3.6838,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068714322870772e-05,
+      "loss": 3.6878,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067877366000094e-05,
+      "loss": 3.6783,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067038771249041e-05,
+      "loss": 3.6933,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066200176497989e-05,
+      "loss": 3.6963,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065361581746937e-05,
+      "loss": 3.6912,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064522986995885e-05,
+      "loss": 3.6943,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063684392244833e-05,
+      "loss": 3.6998,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062847435374154e-05,
+      "loss": 3.714,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062008840623102e-05,
+      "loss": 3.6848,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.06117024587205e-05,
+      "loss": 3.6943,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.060331651120998e-05,
+      "loss": 3.6987,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.059493056369946e-05,
+      "loss": 3.6819,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0586560994992675e-05,
+      "loss": 3.6923,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0578175047482155e-05,
+      "loss": 3.6981,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0569789099971635e-05,
+      "loss": 3.6988,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0561403152461115e-05,
+      "loss": 3.6993,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0553017204950595e-05,
+      "loss": 3.6814,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0544631257440075e-05,
+      "loss": 3.6833,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0536261688733284e-05,
+      "loss": 3.686,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0527875741222764e-05,
+      "loss": 3.6934,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0519489793712244e-05,
+      "loss": 3.6957,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0511103846201724e-05,
+      "loss": 3.6825,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0502717898691204e-05,
+      "loss": 3.6893,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049434832998441e-05,
+      "loss": 3.684,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04859623824739e-05,
+      "loss": 3.6812,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.047757643496338e-05,
+      "loss": 3.6885,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046919048745286e-05,
+      "loss": 3.6959,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046082091874607e-05,
+      "loss": 3.7006,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.045243497123555e-05,
+      "loss": 3.6914,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044404902372503e-05,
+      "loss": 3.6798,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.043566307621451e-05,
+      "loss": 3.6931,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.042727712870399e-05,
+      "loss": 3.6885,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041889118119347e-05,
+      "loss": 3.6907,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041050523368294e-05,
+      "loss": 3.6846,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040211928617242e-05,
+      "loss": 3.6841,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039374971746564e-05,
+      "loss": 3.6844,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0385380148758853e-05,
+      "loss": 3.6989,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0376994201248333e-05,
+      "loss": 3.6747,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036860825373781e-05,
+      "loss": 3.687,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036022230622729e-05,
+      "loss": 3.684,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0351836358716766e-05,
+      "loss": 3.6912,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034346679000998e-05,
+      "loss": 3.6782,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033508084249946e-05,
+      "loss": 3.69,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032669489498894e-05,
+      "loss": 3.6786,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0318308947478415e-05,
+      "loss": 3.6891,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030993937877163e-05,
+      "loss": 3.6921,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030155343126111e-05,
+      "loss": 3.6787,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.029316748375059e-05,
+      "loss": 3.687,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028478153624007e-05,
+      "loss": 3.7015,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027641196753329e-05,
+      "loss": 3.6892,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026802602002277e-05,
+      "loss": 3.6837,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025964007251224e-05,
+      "loss": 3.691,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025125412500172e-05,
+      "loss": 3.6909,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0242884556294936e-05,
+      "loss": 3.6793,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0234498608784416e-05,
+      "loss": 3.6881,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.022611266127389e-05,
+      "loss": 3.6749,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.021772671376337e-05,
+      "loss": 3.6861,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020934076625285e-05,
+      "loss": 3.6814,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020095481874233e-05,
+      "loss": 3.6852,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0192585250035545e-05,
+      "loss": 3.6833,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0184199302525025e-05,
+      "loss": 3.6807,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0175813355014505e-05,
+      "loss": 3.6795,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0167427407503985e-05,
+      "loss": 3.6959,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0159057838797194e-05,
+      "loss": 3.6886,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0150671891286674e-05,
+      "loss": 3.6924,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0142285943776154e-05,
+      "loss": 3.6695,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0133899996265634e-05,
+      "loss": 3.6985,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0125514048755114e-05,
+      "loss": 3.6838,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011714448004832e-05,
+      "loss": 3.6858,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01087585325378e-05,
+      "loss": 3.6901,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010037258502728e-05,
+      "loss": 3.6835,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009198663751676e-05,
+      "loss": 3.6803,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008360069000624e-05,
+      "loss": 3.7012,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007521474249572e-05,
+      "loss": 3.6721,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006684517378894e-05,
+      "loss": 3.6892,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005845922627842e-05,
+      "loss": 3.679,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00500732787679e-05,
+      "loss": 3.681,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004168733125738e-05,
+      "loss": 3.6801,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.003330138374686e-05,
+      "loss": 3.6817,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.002491543623634e-05,
+      "loss": 3.6933,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.001652948872582e-05,
+      "loss": 3.6746,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000815992001903e-05,
+      "loss": 3.6856,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8519585132598877,
+      "eval_runtime": 315.2808,
+      "eval_samples_per_second": 1210.321,
+      "eval_steps_per_second": 37.823,
+      "step": 610560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999977397250851e-05,
+      "loss": 3.6817,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999138802499799e-05,
+      "loss": 3.6744,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.998300207748747e-05,
+      "loss": 3.679,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9974632508780676e-05,
+      "loss": 3.6787,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.996624656127016e-05,
+      "loss": 3.6906,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.995786061375964e-05,
+      "loss": 3.673,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994947466624912e-05,
+      "loss": 3.6822,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.99410887187386e-05,
+      "loss": 3.6776,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.993271915003181e-05,
+      "loss": 3.6798,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.992433320252129e-05,
+      "loss": 3.6864,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.991594725501077e-05,
+      "loss": 3.6805,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.990756130750025e-05,
+      "loss": 3.6775,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989919173879346e-05,
+      "loss": 3.681,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989080579128294e-05,
+      "loss": 3.6665,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.988241984377242e-05,
+      "loss": 3.6774,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98740338962619e-05,
+      "loss": 3.6718,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.986564794875138e-05,
+      "loss": 3.6758,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.985726200124086e-05,
+      "loss": 3.6739,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984887605373034e-05,
+      "loss": 3.6684,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984050648502355e-05,
+      "loss": 3.6855,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983212053751303e-05,
+      "loss": 3.6888,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.982373459000251e-05,
+      "loss": 3.6867,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.981534864249199e-05,
+      "loss": 3.6755,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98069790737852e-05,
+      "loss": 3.6761,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.979859312627468e-05,
+      "loss": 3.6704,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.979020717876416e-05,
+      "loss": 3.684,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.978182123125364e-05,
+      "loss": 3.6746,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.977343528374312e-05,
+      "loss": 3.6721,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9765065715036334e-05,
+      "loss": 3.6647,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9756679767525814e-05,
+      "loss": 3.669,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9748293820015294e-05,
+      "loss": 3.6673,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9739907872504774e-05,
+      "loss": 3.6866,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9731521924994254e-05,
+      "loss": 3.6809,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.972315235628746e-05,
+      "loss": 3.6731,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.971476640877694e-05,
+      "loss": 3.6687,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.970638046126642e-05,
+      "loss": 3.6777,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96979945137559e-05,
+      "loss": 3.6761,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968960856624538e-05,
+      "loss": 3.6738,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968123899753859e-05,
+      "loss": 3.671,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.967285305002807e-05,
+      "loss": 3.6614,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.966446710251755e-05,
+      "loss": 3.6638,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.965608115500703e-05,
+      "loss": 3.6786,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.964769520749652e-05,
+      "loss": 3.6774,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963932563878973e-05,
+      "loss": 3.6645,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963093969127921e-05,
+      "loss": 3.6741,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.962255374376869e-05,
+      "loss": 3.665,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.961416779625817e-05,
+      "loss": 3.6756,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.960579822755138e-05,
+      "loss": 3.6685,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.959741228004086e-05,
+      "loss": 3.6455,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958902633253034e-05,
+      "loss": 3.6802,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958064038501982e-05,
+      "loss": 3.6655,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.95722544375093e-05,
+      "loss": 3.673,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9563884868802506e-05,
+      "loss": 3.6655,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9555498921291986e-05,
+      "loss": 3.6629,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.954711297378147e-05,
+      "loss": 3.6535,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953872702627095e-05,
+      "loss": 3.6624,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953034107876043e-05,
+      "loss": 3.6631,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952195513124991e-05,
+      "loss": 3.6616,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9513569183739386e-05,
+      "loss": 3.6762,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.95051996150326e-05,
+      "loss": 3.659,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.949681366752208e-05,
+      "loss": 3.6631,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948842772001156e-05,
+      "loss": 3.6784,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9480041772501035e-05,
+      "loss": 3.6526,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947167220379425e-05,
+      "loss": 3.6531,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.946328625628373e-05,
+      "loss": 3.6688,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.945490030877321e-05,
+      "loss": 3.6686,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.944651436126269e-05,
+      "loss": 3.6569,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9438144792555906e-05,
+      "loss": 3.6598,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9429758845045386e-05,
+      "loss": 3.6551,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942137289753486e-05,
+      "loss": 3.662,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.941298695002434e-05,
+      "loss": 3.6663,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9404617381317555e-05,
+      "loss": 3.6624,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.939623143380703e-05,
+      "loss": 3.671,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.938784548629651e-05,
+      "loss": 3.6704,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937945953878599e-05,
+      "loss": 3.6866,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9371089970079204e-05,
+      "loss": 3.6546,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.936270402256868e-05,
+      "loss": 3.6724,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9354318075058164e-05,
+      "loss": 3.6683,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9345932127547644e-05,
+      "loss": 3.6559,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9337546180037124e-05,
+      "loss": 3.6607,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932917661133033e-05,
+      "loss": 3.6746,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932079066381981e-05,
+      "loss": 3.6679,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.931240471630929e-05,
+      "loss": 3.6716,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.930401876879877e-05,
+      "loss": 3.6529,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.929563282128825e-05,
+      "loss": 3.6542,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.928726325258146e-05,
+      "loss": 3.6593,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.927887730507094e-05,
+      "loss": 3.6675,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.927049135756042e-05,
+      "loss": 3.6676,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.92621054100499e-05,
+      "loss": 3.6587,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.925373584134312e-05,
+      "loss": 3.654,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.92453498938326e-05,
+      "loss": 3.6616,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.923696394632208e-05,
+      "loss": 3.6524,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.922857799881156e-05,
+      "loss": 3.6602,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9220208430104767e-05,
+      "loss": 3.669,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9211822482594247e-05,
+      "loss": 3.6713,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9203436535083726e-05,
+      "loss": 3.663,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9195050587573206e-05,
+      "loss": 3.6554,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9186681018866416e-05,
+      "loss": 3.6658,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9178295071355896e-05,
+      "loss": 3.6601,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9169909123845375e-05,
+      "loss": 3.6653,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9161523176334855e-05,
+      "loss": 3.6588,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9153137228824335e-05,
+      "loss": 3.654,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9144751281313815e-05,
+      "loss": 3.6659,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.913638171260703e-05,
+      "loss": 3.6669,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.912799576509651e-05,
+      "loss": 3.6485,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911960981758599e-05,
+      "loss": 3.6595,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911122387007547e-05,
+      "loss": 3.6517,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.910283792256495e-05,
+      "loss": 3.6706,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.909446835385816e-05,
+      "loss": 3.6476,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.908608240634764e-05,
+      "loss": 3.6639,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.907769645883712e-05,
+      "loss": 3.6526,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.90693105113266e-05,
+      "loss": 3.6607,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906094094261981e-05,
+      "loss": 3.6634,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.905255499510929e-05,
+      "loss": 3.6558,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.904416904759877e-05,
+      "loss": 3.6587,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9035783100088256e-05,
+      "loss": 3.6728,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9027413531381465e-05,
+      "loss": 3.6604,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9019027583870945e-05,
+      "loss": 3.6617,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9010641636360425e-05,
+      "loss": 3.6637,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9002255688849905e-05,
+      "loss": 3.6614,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8993886120143114e-05,
+      "loss": 3.6586,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8985500172632594e-05,
+      "loss": 3.6588,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8977114225122074e-05,
+      "loss": 3.6466,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8968728277611554e-05,
+      "loss": 3.6603,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8960342330101034e-05,
+      "loss": 3.653,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.895197276139424e-05,
+      "loss": 3.6584,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.894358681388372e-05,
+      "loss": 3.6561,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.89352008663732e-05,
+      "loss": 3.656,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.892681491886269e-05,
+      "loss": 3.6526,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.89184453501559e-05,
+      "loss": 3.67,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.891005940264538e-05,
+      "loss": 3.6575,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.890167345513486e-05,
+      "loss": 3.6691,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.889328750762434e-05,
+      "loss": 3.641,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.888490156011381e-05,
+      "loss": 3.6719,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.887653199140703e-05,
+      "loss": 3.6545,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.886814604389651e-05,
+      "loss": 3.6606,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885976009638599e-05,
+      "loss": 3.6626,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885137414887546e-05,
+      "loss": 3.6548,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8843004580168676e-05,
+      "loss": 3.6582,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8834618632658156e-05,
+      "loss": 3.6735,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.882623268514764e-05,
+      "loss": 3.6419,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8817846737637116e-05,
+      "loss": 3.6654,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880947716893033e-05,
+      "loss": 3.6567,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880109122141981e-05,
+      "loss": 3.6479,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8792705273909285e-05,
+      "loss": 3.6563,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8784319326398765e-05,
+      "loss": 3.6547,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8775933378888245e-05,
+      "loss": 3.6676,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.876756381018146e-05,
+      "loss": 3.6505,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8759177862670934e-05,
+      "loss": 3.6564,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.847994089126587,
+      "eval_runtime": 323.6415,
+      "eval_samples_per_second": 1179.054,
+      "eval_steps_per_second": 36.846,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8750791915160414e-05,
+      "loss": 3.6583,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8742405967649894e-05,
+      "loss": 3.6491,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.873402002013938e-05,
+      "loss": 3.6527,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.872563407262886e-05,
+      "loss": 3.657,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.871726450392207e-05,
+      "loss": 3.6641,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870887855641155e-05,
+      "loss": 3.6455,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870049260890103e-05,
+      "loss": 3.6565,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869210666139051e-05,
+      "loss": 3.6518,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.868372071387999e-05,
+      "loss": 3.6548,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.86753511451732e-05,
+      "loss": 3.6612,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.866696519766268e-05,
+      "loss": 3.6527,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865857925015216e-05,
+      "loss": 3.6517,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865019330264164e-05,
+      "loss": 3.6599,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.864182373393485e-05,
+      "loss": 3.6378,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8633437786424334e-05,
+      "loss": 3.6504,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8625051838913814e-05,
+      "loss": 3.6424,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8616665891403294e-05,
+      "loss": 3.6513,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8608279943892774e-05,
+      "loss": 3.6471,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8599910375185983e-05,
+      "loss": 3.6447,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8591524427675463e-05,
+      "loss": 3.6603,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.858313848016494e-05,
+      "loss": 3.661,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.857475253265442e-05,
+      "loss": 3.6609,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.856638296394763e-05,
+      "loss": 3.6517,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.855799701643711e-05,
+      "loss": 3.6493,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854961106892659e-05,
+      "loss": 3.644,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854122512141607e-05,
+      "loss": 3.6578,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.853285555270929e-05,
+      "loss": 3.652,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.852446960519877e-05,
+      "loss": 3.6491,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.851608365768825e-05,
+      "loss": 3.6395,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.850769771017773e-05,
+      "loss": 3.6424,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849932814147094e-05,
+      "loss": 3.6409,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849094219396042e-05,
+      "loss": 3.6596,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84825562464499e-05,
+      "loss": 3.6596,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.847417029893938e-05,
+      "loss": 3.6472,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.846578435142886e-05,
+      "loss": 3.6451,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8457414782722066e-05,
+      "loss": 3.648,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8449028835211546e-05,
+      "loss": 3.6546,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8440642887701026e-05,
+      "loss": 3.648,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8432256940190506e-05,
+      "loss": 3.6436,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842388737148372e-05,
+      "loss": 3.6351,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84155014239732e-05,
+      "loss": 3.643,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.840711547646268e-05,
+      "loss": 3.6491,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839872952895216e-05,
+      "loss": 3.6542,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839035996024537e-05,
+      "loss": 3.6433,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838197401273485e-05,
+      "loss": 3.646,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837358806522433e-05,
+      "loss": 3.6447,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836520211771381e-05,
+      "loss": 3.6454,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835683254900702e-05,
+      "loss": 3.6441,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.83484466014965e-05,
+      "loss": 3.6196,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834006065398598e-05,
+      "loss": 3.6555,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833167470647546e-05,
+      "loss": 3.6421,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.832328875896494e-05,
+      "loss": 3.649,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8314919190258155e-05,
+      "loss": 3.6419,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8306533242747635e-05,
+      "loss": 3.6378,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8298147295237115e-05,
+      "loss": 3.6313,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8289761347726595e-05,
+      "loss": 3.6345,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8281391779019804e-05,
+      "loss": 3.6442,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8273005831509284e-05,
+      "loss": 3.6294,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8264619883998764e-05,
+      "loss": 3.6509,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8256233936488244e-05,
+      "loss": 3.6374,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.824786436778145e-05,
+      "loss": 3.6341,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823947842027093e-05,
+      "loss": 3.6573,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823109247276041e-05,
+      "loss": 3.6302,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.822270652524989e-05,
+      "loss": 3.6266,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821432057773937e-05,
+      "loss": 3.6459,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.820595100903259e-05,
+      "loss": 3.6444,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.819756506152207e-05,
+      "loss": 3.6325,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818917911401154e-05,
+      "loss": 3.6352,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818079316650102e-05,
+      "loss": 3.6237,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817242359779424e-05,
+      "loss": 3.6401,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.816403765028371e-05,
+      "loss": 3.641,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.815565170277319e-05,
+      "loss": 3.6432,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.814726575526267e-05,
+      "loss": 3.6435,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813889618655589e-05,
+      "loss": 3.6436,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813051023904537e-05,
+      "loss": 3.6608,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812212429153485e-05,
+      "loss": 3.6334,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.811373834402433e-05,
+      "loss": 3.65,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810536877531754e-05,
+      "loss": 3.6399,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8096982827807016e-05,
+      "loss": 3.634,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8088596880296496e-05,
+      "loss": 3.6339,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8080210932785976e-05,
+      "loss": 3.6525,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8071841364079185e-05,
+      "loss": 3.6425,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8063455416568665e-05,
+      "loss": 3.6477,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8055069469058145e-05,
+      "loss": 3.6291,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8046683521547625e-05,
+      "loss": 3.6314,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8038297574037105e-05,
+      "loss": 3.6352,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802992800533032e-05,
+      "loss": 3.6409,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.80215420578198e-05,
+      "loss": 3.6424,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.801315611030928e-05,
+      "loss": 3.6387,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.800477016279876e-05,
+      "loss": 3.627,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799640059409197e-05,
+      "loss": 3.6393,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.798801464658145e-05,
+      "loss": 3.631,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797962869907093e-05,
+      "loss": 3.6347,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797124275156041e-05,
+      "loss": 3.6438,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796287318285362e-05,
+      "loss": 3.6453,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79544872353431e-05,
+      "loss": 3.638,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794610128783258e-05,
+      "loss": 3.6334,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793771534032206e-05,
+      "loss": 3.636,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7929345771615274e-05,
+      "loss": 3.6346,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7920959824104754e-05,
+      "loss": 3.6431,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7912573876594234e-05,
+      "loss": 3.6347,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7904187929083714e-05,
+      "loss": 3.6338,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7895801981573194e-05,
+      "loss": 3.6378,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78874324128664e-05,
+      "loss": 3.6445,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787904646535588e-05,
+      "loss": 3.6257,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787066051784536e-05,
+      "loss": 3.6368,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786227457033484e-05,
+      "loss": 3.6271,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.785390500162805e-05,
+      "loss": 3.6464,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784551905411753e-05,
+      "loss": 3.6254,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.783713310660701e-05,
+      "loss": 3.637,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782874715909649e-05,
+      "loss": 3.6266,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782037759038971e-05,
+      "loss": 3.6416,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781199164287919e-05,
+      "loss": 3.6384,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.780360569536867e-05,
+      "loss": 3.6294,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.779521974785815e-05,
+      "loss": 3.6361,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778685017915136e-05,
+      "loss": 3.6483,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777846423164084e-05,
+      "loss": 3.6354,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777007828413032e-05,
+      "loss": 3.6411,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.77616923366198e-05,
+      "loss": 3.637,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7753322767913006e-05,
+      "loss": 3.6362,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7744936820402486e-05,
+      "loss": 3.6389,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7736550872891966e-05,
+      "loss": 3.6299,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7728164925381446e-05,
+      "loss": 3.6259,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771979535667466e-05,
+      "loss": 3.6368,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771140940916414e-05,
+      "loss": 3.6343,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.770302346165362e-05,
+      "loss": 3.6341,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76946375141431e-05,
+      "loss": 3.6334,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768626794543631e-05,
+      "loss": 3.6313,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.767788199792579e-05,
+      "loss": 3.6281,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766949605041527e-05,
+      "loss": 3.6463,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766111010290475e-05,
+      "loss": 3.6342,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.765274053419796e-05,
+      "loss": 3.644,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.764435458668744e-05,
+      "loss": 3.6176,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.763596863917692e-05,
+      "loss": 3.6461,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76275826916664e-05,
+      "loss": 3.6357,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7619213122959615e-05,
+      "loss": 3.636,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7610827175449095e-05,
+      "loss": 3.6346,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7602441227938575e-05,
+      "loss": 3.6307,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7594055280428055e-05,
+      "loss": 3.6355,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7585685711721264e-05,
+      "loss": 3.6484,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7577299764210744e-05,
+      "loss": 3.6223,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7568913816700224e-05,
+      "loss": 3.6373,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7560527869189704e-05,
+      "loss": 3.6313,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755215830048291e-05,
+      "loss": 3.6247,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754377235297239e-05,
+      "loss": 3.6345,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753538640546187e-05,
+      "loss": 3.6324,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.752700045795135e-05,
+      "loss": 3.64,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751863088924457e-05,
+      "loss": 3.6353,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751024494173405e-05,
+      "loss": 3.6303,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8446507453918457,
+      "eval_runtime": 303.3125,
+      "eval_samples_per_second": 1258.079,
+      "eval_steps_per_second": 39.316,
+      "step": 763200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.750185899422353e-05,
+      "loss": 3.6298,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.749347304671301e-05,
+      "loss": 3.6256,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.748508709920249e-05,
+      "loss": 3.6289,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.74767175304957e-05,
+      "loss": 3.6347,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.746833158298518e-05,
+      "loss": 3.6421,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.745994563547466e-05,
+      "loss": 3.623,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.745155968796414e-05,
+      "loss": 3.6317,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.744319011925735e-05,
+      "loss": 3.6307,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.743480417174683e-05,
+      "loss": 3.6286,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.742641822423631e-05,
+      "loss": 3.6335,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7418032276725787e-05,
+      "loss": 3.6299,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7409662708019e-05,
+      "loss": 3.6313,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740129313931221e-05,
+      "loss": 3.6358,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.739290719180169e-05,
+      "loss": 3.6159,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.738452124429117e-05,
+      "loss": 3.6267,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.737613529678065e-05,
+      "loss": 3.6167,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.736774934927013e-05,
+      "loss": 3.6287,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735936340175961e-05,
+      "loss": 3.625,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735097745424909e-05,
+      "loss": 3.6226,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7342591506738564e-05,
+      "loss": 3.6361,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7334205559228044e-05,
+      "loss": 3.6351,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.732583599052126e-05,
+      "loss": 3.6364,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.731745004301074e-05,
+      "loss": 3.6315,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730906409550022e-05,
+      "loss": 3.6276,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.73006781479897e-05,
+      "loss": 3.6254,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7292308579282916e-05,
+      "loss": 3.6299,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.728392263177239e-05,
+      "loss": 3.6273,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.727553668426187e-05,
+      "loss": 3.6292,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.726715073675135e-05,
+      "loss": 3.6174,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7258781168044565e-05,
+      "loss": 3.6177,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.725039522053404e-05,
+      "loss": 3.6204,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724200927302352e-05,
+      "loss": 3.6301,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7233623325513e-05,
+      "loss": 3.6381,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7225253756806214e-05,
+      "loss": 3.6258,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7216867809295694e-05,
+      "loss": 3.6223,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7208481861785174e-05,
+      "loss": 3.6212,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7200095914274654e-05,
+      "loss": 3.6355,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719172634556786e-05,
+      "loss": 3.6277,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.718334039805734e-05,
+      "loss": 3.6231,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.717495445054682e-05,
+      "loss": 3.6118,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.71665685030363e-05,
+      "loss": 3.6227,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.715819893432951e-05,
+      "loss": 3.6234,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714981298681899e-05,
+      "loss": 3.6301,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714142703930847e-05,
+      "loss": 3.6241,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.713304109179795e-05,
+      "loss": 3.6216,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.712467152309117e-05,
+      "loss": 3.621,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.711628557558065e-05,
+      "loss": 3.6252,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.710789962807013e-05,
+      "loss": 3.6231,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709951368055961e-05,
+      "loss": 3.5975,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7091144111852817e-05,
+      "loss": 3.6287,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7082758164342297e-05,
+      "loss": 3.6188,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7074372216831777e-05,
+      "loss": 3.6292,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7065986269321256e-05,
+      "loss": 3.6184,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7057616700614466e-05,
+      "loss": 3.6174,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7049230753103946e-05,
+      "loss": 3.6072,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7040844805593425e-05,
+      "loss": 3.6126,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7032458858082905e-05,
+      "loss": 3.6241,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.702408928937612e-05,
+      "loss": 3.6053,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.70157033418656e-05,
+      "loss": 3.6317,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.700731739435508e-05,
+      "loss": 3.6142,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.699893144684456e-05,
+      "loss": 3.6099,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.699054549933404e-05,
+      "loss": 3.6361,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.698217593062725e-05,
+      "loss": 3.6119,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.697378998311673e-05,
+      "loss": 3.6034,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.696540403560621e-05,
+      "loss": 3.6206,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.695703446689942e-05,
+      "loss": 3.6219,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69486485193889e-05,
+      "loss": 3.609,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694026257187838e-05,
+      "loss": 3.616,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693187662436786e-05,
+      "loss": 3.6036,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.692349067685734e-05,
+      "loss": 3.6141,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6915121108150555e-05,
+      "loss": 3.6211,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6906735160640035e-05,
+      "loss": 3.6193,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6898349213129515e-05,
+      "loss": 3.6201,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6889963265618995e-05,
+      "loss": 3.6226,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6881593696912204e-05,
+      "loss": 3.638,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6873207749401684e-05,
+      "loss": 3.6159,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6864821801891164e-05,
+      "loss": 3.6239,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6856435854380644e-05,
+      "loss": 3.6161,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.684806628567385e-05,
+      "loss": 3.6138,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683968033816333e-05,
+      "loss": 3.6093,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683129439065281e-05,
+      "loss": 3.6329,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.682290844314229e-05,
+      "loss": 3.6211,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.681453887443551e-05,
+      "loss": 3.6281,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.680615292692499e-05,
+      "loss": 3.6024,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.679776697941447e-05,
+      "loss": 3.611,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678938103190395e-05,
+      "loss": 3.6143,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678101146319716e-05,
+      "loss": 3.6188,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677262551568664e-05,
+      "loss": 3.6177,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.676423956817612e-05,
+      "loss": 3.6187,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.67558536206656e-05,
+      "loss": 3.6049,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6747484051958807e-05,
+      "loss": 3.616,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6739098104448286e-05,
+      "loss": 3.612,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6730712156937766e-05,
+      "loss": 3.612,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6722326209427246e-05,
+      "loss": 3.6228,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.671395664072046e-05,
+      "loss": 3.625,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.670557069320994e-05,
+      "loss": 3.6154,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.669718474569942e-05,
+      "loss": 3.6112,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.66887987981889e-05,
+      "loss": 3.6161,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668042922948211e-05,
+      "loss": 3.6107,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667204328197159e-05,
+      "loss": 3.6207,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.666365733446107e-05,
+      "loss": 3.6113,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.665527138695055e-05,
+      "loss": 3.6165,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664690181824376e-05,
+      "loss": 3.615,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663851587073324e-05,
+      "loss": 3.6153,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663012992322272e-05,
+      "loss": 3.6127,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.66217439757122e-05,
+      "loss": 3.6123,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6613374407005416e-05,
+      "loss": 3.6046,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6604988459494896e-05,
+      "loss": 3.6232,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6596602511984376e-05,
+      "loss": 3.6099,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6588216564473856e-05,
+      "loss": 3.6109,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6579846995767065e-05,
+      "loss": 3.6023,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6571461048256545e-05,
+      "loss": 3.6172,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6563075100746025e-05,
+      "loss": 3.6221,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6554689153235505e-05,
+      "loss": 3.6107,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6546319584528714e-05,
+      "loss": 3.6154,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6537933637018194e-05,
+      "loss": 3.6269,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6529547689507674e-05,
+      "loss": 3.6116,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6521161741997154e-05,
+      "loss": 3.6195,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.651279217329037e-05,
+      "loss": 3.6169,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.650440622577985e-05,
+      "loss": 3.6158,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.649602027826933e-05,
+      "loss": 3.6188,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.648763433075881e-05,
+      "loss": 3.6067,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.647926476205202e-05,
+      "loss": 3.6063,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.64708788145415e-05,
+      "loss": 3.6146,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646249286703098e-05,
+      "loss": 3.6144,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.645410691952046e-05,
+      "loss": 3.615,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.644573735081367e-05,
+      "loss": 3.6081,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.643735140330315e-05,
+      "loss": 3.6114,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642896545579263e-05,
+      "loss": 3.6052,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642057950828211e-05,
+      "loss": 3.6263,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641220993957532e-05,
+      "loss": 3.6116,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.64038239920648e-05,
+      "loss": 3.625,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.639543804455428e-05,
+      "loss": 3.5981,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.638705209704376e-05,
+      "loss": 3.6207,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6378666149533236e-05,
+      "loss": 3.6168,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637029658082645e-05,
+      "loss": 3.6128,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636191063331593e-05,
+      "loss": 3.6171,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.635352468580541e-05,
+      "loss": 3.6089,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6345138738294885e-05,
+      "loss": 3.6136,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63367691695881e-05,
+      "loss": 3.6296,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.632838322207758e-05,
+      "loss": 3.6029,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631999727456706e-05,
+      "loss": 3.6166,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631161132705654e-05,
+      "loss": 3.6082,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.630324175834976e-05,
+      "loss": 3.6062,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.629485581083924e-05,
+      "loss": 3.612,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.628646986332871e-05,
+      "loss": 3.61,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.627808391581819e-05,
+      "loss": 3.621,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6269714347111406e-05,
+      "loss": 3.6151,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6261328399600886e-05,
+      "loss": 3.6105,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8437488079071045,
+      "eval_runtime": 304.1626,
+      "eval_samples_per_second": 1254.563,
+      "eval_steps_per_second": 39.206,
+      "step": 839520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.625294245209036e-05,
+      "loss": 3.6103,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.624455650457984e-05,
+      "loss": 3.6035,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.623617055706932e-05,
+      "loss": 3.6082,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.62277846095588e-05,
+      "loss": 3.6131,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6219415040852015e-05,
+      "loss": 3.6208,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6211029093341495e-05,
+      "loss": 3.6038,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6202643145830975e-05,
+      "loss": 3.6138,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6194257198320455e-05,
+      "loss": 3.6093,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6185887629613664e-05,
+      "loss": 3.6068,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6177501682103144e-05,
+      "loss": 3.6078,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6169115734592624e-05,
+      "loss": 3.6135,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6160729787082104e-05,
+      "loss": 3.6086,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.615236021837531e-05,
+      "loss": 3.6142,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.614397427086479e-05,
+      "loss": 3.5978,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.613558832335427e-05,
+      "loss": 3.6037,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.612720237584375e-05,
+      "loss": 3.5937,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611881642833323e-05,
+      "loss": 3.6135,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611043048082271e-05,
+      "loss": 3.6041,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.61020445333122e-05,
+      "loss": 3.6008,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.609365858580168e-05,
+      "loss": 3.6135,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.608528901709489e-05,
+      "loss": 3.6152,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.607690306958437e-05,
+      "loss": 3.6132,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606851712207385e-05,
+      "loss": 3.6157,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606013117456333e-05,
+      "loss": 3.6047,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605176160585654e-05,
+      "loss": 3.6086,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.604337565834602e-05,
+      "loss": 3.6124,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.60349897108355e-05,
+      "loss": 3.6002,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.602660376332498e-05,
+      "loss": 3.6103,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6018234194618186e-05,
+      "loss": 3.6,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6009848247107666e-05,
+      "loss": 3.5952,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.600146229959715e-05,
+      "loss": 3.6016,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.599307635208663e-05,
+      "loss": 3.6114,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.598469040457611e-05,
+      "loss": 3.6132,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.597632083586932e-05,
+      "loss": 3.6112,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59679348883588e-05,
+      "loss": 3.5999,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595954894084828e-05,
+      "loss": 3.6017,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595116299333776e-05,
+      "loss": 3.6155,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.594279342463097e-05,
+      "loss": 3.603,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.593440747712045e-05,
+      "loss": 3.6035,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.592602152960993e-05,
+      "loss": 3.594,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.591763558209941e-05,
+      "loss": 3.6008,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590926601339262e-05,
+      "loss": 3.6006,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5900880065882106e-05,
+      "loss": 3.6097,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5892494118371586e-05,
+      "loss": 3.6032,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5884124549664796e-05,
+      "loss": 3.5984,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5875738602154276e-05,
+      "loss": 3.6034,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5867352654643755e-05,
+      "loss": 3.6036,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5858966707133235e-05,
+      "loss": 3.6044,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5850597138426445e-05,
+      "loss": 3.5807,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5842211190915924e-05,
+      "loss": 3.6076,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5833825243405404e-05,
+      "loss": 3.5964,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5825439295894884e-05,
+      "loss": 3.6097,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5817069727188093e-05,
+      "loss": 3.5986,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5808683779677573e-05,
+      "loss": 3.5933,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580029783216706e-05,
+      "loss": 3.5879,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579191188465654e-05,
+      "loss": 3.5921,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.578354231594975e-05,
+      "loss": 3.6026,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.577515636843923e-05,
+      "loss": 3.5831,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.576677042092871e-05,
+      "loss": 3.6126,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.575838447341819e-05,
+      "loss": 3.5958,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.57500149047114e-05,
+      "loss": 3.5887,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574162895720088e-05,
+      "loss": 3.6126,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.573324300969036e-05,
+      "loss": 3.592,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.572485706217984e-05,
+      "loss": 3.5864,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.571648749347305e-05,
+      "loss": 3.6005,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.570810154596253e-05,
+      "loss": 3.6014,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569971559845201e-05,
+      "loss": 3.5861,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5691329650941494e-05,
+      "loss": 3.5989,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.56829600822347e-05,
+      "loss": 3.582,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.567457413472418e-05,
+      "loss": 3.5936,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.566618818721366e-05,
+      "loss": 3.6044,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.565780223970314e-05,
+      "loss": 3.5982,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564943267099635e-05,
+      "loss": 3.6,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564104672348583e-05,
+      "loss": 3.6019,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563266077597531e-05,
+      "loss": 3.62,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5624274828464785e-05,
+      "loss": 3.596,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5615905259758e-05,
+      "loss": 3.6047,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.560751931224748e-05,
+      "loss": 3.5944,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559913336473696e-05,
+      "loss": 3.5929,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559074741722644e-05,
+      "loss": 3.593,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5582377848519657e-05,
+      "loss": 3.6098,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5573991901009137e-05,
+      "loss": 3.6013,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.556560595349861e-05,
+      "loss": 3.608,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.555722000598809e-05,
+      "loss": 3.5882,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5548850437281306e-05,
+      "loss": 3.5879,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5540464489770785e-05,
+      "loss": 3.5928,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553207854226026e-05,
+      "loss": 3.5978,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.552369259474974e-05,
+      "loss": 3.5994,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5515323026042954e-05,
+      "loss": 3.5998,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5506937078532434e-05,
+      "loss": 3.5891,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5498551131021914e-05,
+      "loss": 3.5931,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5490165183511394e-05,
+      "loss": 3.5916,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548179561480461e-05,
+      "loss": 3.5903,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5473409667294083e-05,
+      "loss": 3.6031,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.546502371978356e-05,
+      "loss": 3.6072,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.545663777227304e-05,
+      "loss": 3.5995,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.544826820356626e-05,
+      "loss": 3.5889,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543988225605573e-05,
+      "loss": 3.5984,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543149630854521e-05,
+      "loss": 3.5912,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.542311036103469e-05,
+      "loss": 3.5995,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.541474079232791e-05,
+      "loss": 3.5962,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.540635484481739e-05,
+      "loss": 3.5916,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.539796889730687e-05,
+      "loss": 3.5977,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538958294979635e-05,
+      "loss": 3.5922,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538121338108956e-05,
+      "loss": 3.5963,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.537282743357904e-05,
+      "loss": 3.591,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.536444148606852e-05,
+      "loss": 3.5873,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5356055538558e-05,
+      "loss": 3.6013,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5347685969851206e-05,
+      "loss": 3.5943,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5339300022340686e-05,
+      "loss": 3.59,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5330914074830166e-05,
+      "loss": 3.583,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5322528127319646e-05,
+      "loss": 3.5963,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.531415855861286e-05,
+      "loss": 3.6037,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.530577261110234e-05,
+      "loss": 3.5883,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.529738666359182e-05,
+      "loss": 3.5985,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52890007160813e-05,
+      "loss": 3.6042,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528063114737451e-05,
+      "loss": 3.5963,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527224519986399e-05,
+      "loss": 3.5956,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.526385925235347e-05,
+      "loss": 3.601,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.525547330484295e-05,
+      "loss": 3.5947,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.524710373613616e-05,
+      "loss": 3.5969,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523871778862564e-05,
+      "loss": 3.5874,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523033184111512e-05,
+      "loss": 3.5876,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52219458936046e-05,
+      "loss": 3.5957,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5213576324897815e-05,
+      "loss": 3.5948,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5205190377387295e-05,
+      "loss": 3.593,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5196804429876775e-05,
+      "loss": 3.5934,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5188418482366255e-05,
+      "loss": 3.5864,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5180048913659464e-05,
+      "loss": 3.5851,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5171662966148944e-05,
+      "loss": 3.6059,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5163277018638424e-05,
+      "loss": 3.5907,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5154891071127904e-05,
+      "loss": 3.6085,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5146521502421113e-05,
+      "loss": 3.5799,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.513813555491059e-05,
+      "loss": 3.602,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512974960740007e-05,
+      "loss": 3.6001,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512136365988955e-05,
+      "loss": 3.5926,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.511299409118277e-05,
+      "loss": 3.5901,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.510460814367225e-05,
+      "loss": 3.5925,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.509622219616173e-05,
+      "loss": 3.5972,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.508783624865121e-05,
+      "loss": 3.6031,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.507946667994442e-05,
+      "loss": 3.588,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.50710807324339e-05,
+      "loss": 3.5923,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506269478492338e-05,
+      "loss": 3.5927,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.505430883741286e-05,
+      "loss": 3.5882,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.504593926870607e-05,
+      "loss": 3.5887,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.503755332119555e-05,
+      "loss": 3.5958,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.502916737368503e-05,
+      "loss": 3.5967,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.502078142617451e-05,
+      "loss": 3.5962,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.501241185746772e-05,
+      "loss": 3.5889,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.842909812927246,
+      "eval_runtime": 343.3193,
+      "eval_samples_per_second": 1111.475,
+      "eval_steps_per_second": 34.734,
+      "step": 915840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.50040259099572e-05,
+      "loss": 3.5914,
+      "step": 915968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.499563996244668e-05,
+      "loss": 3.5813,
+      "step": 916480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.498725401493616e-05,
+      "loss": 3.5928,
+      "step": 916992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.497888444622937e-05,
+      "loss": 3.5939,
+      "step": 917504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.497049849871885e-05,
+      "loss": 3.5957,
+      "step": 918016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.496211255120833e-05,
+      "loss": 3.5892,
+      "step": 918528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.495372660369781e-05,
+      "loss": 3.5947,
+      "step": 919040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.494535703499102e-05,
+      "loss": 3.5911,
+      "step": 919552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.49369710874805e-05,
+      "loss": 3.5867,
+      "step": 920064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.492858513996998e-05,
+      "loss": 3.5911,
+      "step": 920576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.492019919245946e-05,
+      "loss": 3.5962,
+      "step": 921088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4911829623752676e-05,
+      "loss": 3.5886,
+      "step": 921600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4903460055045886e-05,
+      "loss": 3.6009,
+      "step": 922112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4895074107535366e-05,
+      "loss": 3.5735,
+      "step": 922624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4886688160024845e-05,
+      "loss": 3.5823,
+      "step": 923136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4878302212514325e-05,
+      "loss": 3.5774,
+      "step": 923648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4869916265003805e-05,
+      "loss": 3.5945,
+      "step": 924160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4861530317493285e-05,
+      "loss": 3.5841,
+      "step": 924672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4853144369982765e-05,
+      "loss": 3.5818,
+      "step": 925184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4844758422472245e-05,
+      "loss": 3.5913,
+      "step": 925696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4836388853765454e-05,
+      "loss": 3.5986,
+      "step": 926208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4828002906254934e-05,
+      "loss": 3.595,
+      "step": 926720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4819616958744414e-05,
+      "loss": 3.5929,
+      "step": 927232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4811231011233894e-05,
+      "loss": 3.5866,
+      "step": 927744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.480286144252711e-05,
+      "loss": 3.5955,
+      "step": 928256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.479447549501659e-05,
+      "loss": 3.5917,
+      "step": 928768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.478608954750607e-05,
+      "loss": 3.5802,
+      "step": 929280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.477770359999555e-05,
+      "loss": 3.5903,
+      "step": 929792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.476933403128876e-05,
+      "loss": 3.5813,
+      "step": 930304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.476094808377824e-05,
+      "loss": 3.5769,
+      "step": 930816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.475256213626772e-05,
+      "loss": 3.585,
+      "step": 931328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.47441761887572e-05,
+      "loss": 3.5896,
+      "step": 931840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.473580662005041e-05,
+      "loss": 3.5939,
+      "step": 932352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.472742067253989e-05,
+      "loss": 3.5943,
+      "step": 932864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.471903472502937e-05,
+      "loss": 3.5813,
+      "step": 933376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.471064877751885e-05,
+      "loss": 3.5815,
+      "step": 933888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4702279208812064e-05,
+      "loss": 3.5955,
+      "step": 934400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4693893261301544e-05,
+      "loss": 3.583,
+      "step": 934912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4685507313791024e-05,
+      "loss": 3.5861,
+      "step": 935424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4677121366280504e-05,
+      "loss": 3.579,
+      "step": 935936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.466875179757371e-05,
+      "loss": 3.5772,
+      "step": 936448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.466038222886692e-05,
+      "loss": 3.5825,
+      "step": 936960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.46519962813564e-05,
+      "loss": 3.5925,
+      "step": 937472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.464361033384588e-05,
+      "loss": 3.5842,
+      "step": 937984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.463522438633536e-05,
+      "loss": 3.581,
+      "step": 938496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.462683843882484e-05,
+      "loss": 3.585,
+      "step": 939008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.461845249131432e-05,
+      "loss": 3.5841,
+      "step": 939520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.46100665438038e-05,
+      "loss": 3.5876,
+      "step": 940032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.460168059629328e-05,
+      "loss": 3.5654,
+      "step": 940544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.45933110275865e-05,
+      "loss": 3.5888,
+      "step": 941056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.458492508007598e-05,
+      "loss": 3.578,
+      "step": 941568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.457653913256546e-05,
+      "loss": 3.5912,
+      "step": 942080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.456815318505494e-05,
+      "loss": 3.5799,
+      "step": 942592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4559783616348146e-05,
+      "loss": 3.5762,
+      "step": 943104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4551397668837626e-05,
+      "loss": 3.5687,
+      "step": 943616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4543011721327106e-05,
+      "loss": 3.5722,
+      "step": 944128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.453462577381658e-05,
+      "loss": 3.584,
+      "step": 944640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4526256205109795e-05,
+      "loss": 3.5634,
+      "step": 945152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4517870257599275e-05,
+      "loss": 3.5933,
+      "step": 945664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4509484310088755e-05,
+      "loss": 3.5805,
+      "step": 946176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4501098362578235e-05,
+      "loss": 3.5667,
+      "step": 946688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.449272879387145e-05,
+      "loss": 3.5967,
+      "step": 947200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.448434284636093e-05,
+      "loss": 3.5708,
+      "step": 947712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.447595689885041e-05,
+      "loss": 3.567,
+      "step": 948224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4467570951339884e-05,
+      "loss": 3.5796,
+      "step": 948736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.44592013826331e-05,
+      "loss": 3.5836,
+      "step": 949248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.445081543512258e-05,
+      "loss": 3.5713,
+      "step": 949760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.444242948761205e-05,
+      "loss": 3.5816,
+      "step": 950272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.443404354010153e-05,
+      "loss": 3.5622,
+      "step": 950784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.442567397139475e-05,
+      "loss": 3.5763,
+      "step": 951296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.441728802388423e-05,
+      "loss": 3.5827,
+      "step": 951808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.440890207637371e-05,
+      "loss": 3.5826,
+      "step": 952320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.440051612886319e-05,
+      "loss": 3.5793,
+      "step": 952832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4392146560156405e-05,
+      "loss": 3.5849,
+      "step": 953344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4383760612645885e-05,
+      "loss": 3.5991,
+      "step": 953856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.437537466513536e-05,
+      "loss": 3.577,
+      "step": 954368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.436698871762484e-05,
+      "loss": 3.5932,
+      "step": 954880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4358619148918054e-05,
+      "loss": 3.5756,
+      "step": 955392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.435023320140753e-05,
+      "loss": 3.5777,
+      "step": 955904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.434184725389701e-05,
+      "loss": 3.5693,
+      "step": 956416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.433346130638649e-05,
+      "loss": 3.5898,
+      "step": 956928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.43250917376797e-05,
+      "loss": 3.5844,
+      "step": 957440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4316705790169176e-05,
+      "loss": 3.5906,
+      "step": 957952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.430831984265866e-05,
+      "loss": 3.572,
+      "step": 958464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.429993389514814e-05,
+      "loss": 3.571,
+      "step": 958976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.429156432644136e-05,
+      "loss": 3.5715,
+      "step": 959488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.428317837893083e-05,
+      "loss": 3.5855,
+      "step": 960000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.427479243142031e-05,
+      "loss": 3.5783,
+      "step": 960512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.426640648390979e-05,
+      "loss": 3.5806,
+      "step": 961024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4258036915203e-05,
+      "loss": 3.5751,
+      "step": 961536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424965096769248e-05,
+      "loss": 3.5741,
+      "step": 962048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424126502018196e-05,
+      "loss": 3.5738,
+      "step": 962560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.423287907267144e-05,
+      "loss": 3.5726,
+      "step": 963072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.422450950396465e-05,
+      "loss": 3.5821,
+      "step": 963584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.421612355645413e-05,
+      "loss": 3.5916,
+      "step": 964096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4207737608943616e-05,
+      "loss": 3.5814,
+      "step": 964608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4199351661433096e-05,
+      "loss": 3.5723,
+      "step": 965120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4190982092726305e-05,
+      "loss": 3.5773,
+      "step": 965632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4182596145215785e-05,
+      "loss": 3.5756,
+      "step": 966144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4174210197705265e-05,
+      "loss": 3.581,
+      "step": 966656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4165824250194745e-05,
+      "loss": 3.577,
+      "step": 967168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4157454681487954e-05,
+      "loss": 3.5722,
+      "step": 967680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4149068733977434e-05,
+      "loss": 3.5799,
+      "step": 968192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4140682786466914e-05,
+      "loss": 3.5779,
+      "step": 968704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4132296838956394e-05,
+      "loss": 3.5771,
+      "step": 969216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.41239272702496e-05,
+      "loss": 3.5727,
+      "step": 969728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.411554132273908e-05,
+      "loss": 3.5711,
+      "step": 970240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.410715537522857e-05,
+      "loss": 3.579,
+      "step": 970752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.409878580652178e-05,
+      "loss": 3.5769,
+      "step": 971264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.409039985901126e-05,
+      "loss": 3.5718,
+      "step": 971776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.408201391150074e-05,
+      "loss": 3.5659,
+      "step": 972288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.407362796399022e-05,
+      "loss": 3.5749,
+      "step": 972800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.406525839528343e-05,
+      "loss": 3.5892,
+      "step": 973312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.405687244777291e-05,
+      "loss": 3.5738,
+      "step": 973824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.404848650026239e-05,
+      "loss": 3.5776,
+      "step": 974336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.404010055275187e-05,
+      "loss": 3.585,
+      "step": 974848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.403173098404508e-05,
+      "loss": 3.5795,
+      "step": 975360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.402334503653456e-05,
+      "loss": 3.5784,
+      "step": 975872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.401495908902404e-05,
+      "loss": 3.5813,
+      "step": 976384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4006573141513524e-05,
+      "loss": 3.5781,
+      "step": 976896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.399820357280673e-05,
+      "loss": 3.583,
+      "step": 977408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.398981762529621e-05,
+      "loss": 3.5659,
+      "step": 977920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.398143167778569e-05,
+      "loss": 3.5722,
+      "step": 978432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.397304573027517e-05,
+      "loss": 3.5738,
+      "step": 978944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.396467616156838e-05,
+      "loss": 3.5769,
+      "step": 979456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.395629021405786e-05,
+      "loss": 3.5784,
+      "step": 979968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.394790426654734e-05,
+      "loss": 3.5727,
+      "step": 980480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393951831903682e-05,
+      "loss": 3.5636,
+      "step": 980992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393114875033003e-05,
+      "loss": 3.5693,
+      "step": 981504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.392276280281951e-05,
+      "loss": 3.5864,
+      "step": 982016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.391437685530899e-05,
+      "loss": 3.5785,
+      "step": 982528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.390599090779848e-05,
+      "loss": 3.5856,
+      "step": 983040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3897621339091686e-05,
+      "loss": 3.5677,
+      "step": 983552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3889235391581166e-05,
+      "loss": 3.5793,
+      "step": 984064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3880849444070646e-05,
+      "loss": 3.5849,
+      "step": 984576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3872463496560126e-05,
+      "loss": 3.5736,
+      "step": 985088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3864093927853335e-05,
+      "loss": 3.5727,
+      "step": 985600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3855707980342815e-05,
+      "loss": 3.576,
+      "step": 986112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3847322032832295e-05,
+      "loss": 3.5811,
+      "step": 986624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3838936085321775e-05,
+      "loss": 3.5846,
+      "step": 987136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3830566516614984e-05,
+      "loss": 3.571,
+      "step": 987648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3822180569104464e-05,
+      "loss": 3.5711,
+      "step": 988160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3813794621593944e-05,
+      "loss": 3.5746,
+      "step": 988672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3805408674083424e-05,
+      "loss": 3.5734,
+      "step": 989184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.379703910537664e-05,
+      "loss": 3.5681,
+      "step": 989696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.378865315786612e-05,
+      "loss": 3.5818,
+      "step": 990208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.37802672103556e-05,
+      "loss": 3.5736,
+      "step": 990720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.377188126284508e-05,
+      "loss": 3.5827,
+      "step": 991232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.376351169413829e-05,
+      "loss": 3.5707,
+      "step": 991744
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.843370199203491,
+      "eval_runtime": 344.5223,
+      "eval_samples_per_second": 1107.595,
+      "eval_steps_per_second": 34.613,
+      "step": 992160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.375512574662777e-05,
+      "loss": 3.5889,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.374673979911725e-05,
+      "loss": 3.5633,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.373835385160673e-05,
+      "loss": 3.5725,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372996790409621e-05,
+      "loss": 3.5805,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372159833538942e-05,
+      "loss": 3.5752,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.37132123878789e-05,
+      "loss": 3.5742,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.370482644036838e-05,
+      "loss": 3.5743,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3696440492857865e-05,
+      "loss": 3.5705,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3688070924151074e-05,
+      "loss": 3.5686,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3679684976640554e-05,
+      "loss": 3.5775,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3671299029130034e-05,
+      "loss": 3.5728,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3662913081619513e-05,
+      "loss": 3.5755,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.365454351291272e-05,
+      "loss": 3.5823,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.36461575654022e-05,
+      "loss": 3.5573,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.363777161789168e-05,
+      "loss": 3.5672,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.362938567038116e-05,
+      "loss": 3.5593,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.362099972287064e-05,
+      "loss": 3.5799,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.361261377536012e-05,
+      "loss": 3.5637,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.36042278278496e-05,
+      "loss": 3.5665,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.359584188033908e-05,
+      "loss": 3.571,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.35874723116323e-05,
+      "loss": 3.5823,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357908636412178e-05,
+      "loss": 3.5763,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357070041661126e-05,
+      "loss": 3.5785,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.356231446910073e-05,
+      "loss": 3.5675,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.355394490039395e-05,
+      "loss": 3.579,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.354555895288343e-05,
+      "loss": 3.568,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.35371730053729e-05,
+      "loss": 3.5679,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.352878705786238e-05,
+      "loss": 3.5727,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3520417489155596e-05,
+      "loss": 3.5647,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3512031541645076e-05,
+      "loss": 3.5602,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3503645594134556e-05,
+      "loss": 3.566,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3495259646624036e-05,
+      "loss": 3.5715,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.348689007791725e-05,
+      "loss": 3.5793,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347850413040673e-05,
+      "loss": 3.5751,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3470118182896205e-05,
+      "loss": 3.5644,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3461732235385685e-05,
+      "loss": 3.5637,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.34533626666789e-05,
+      "loss": 3.5832,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3444976719168374e-05,
+      "loss": 3.5593,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3436590771657854e-05,
+      "loss": 3.5734,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3428204824147334e-05,
+      "loss": 3.5636,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341983525544055e-05,
+      "loss": 3.5566,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341146568673376e-05,
+      "loss": 3.5659,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.340307973922324e-05,
+      "loss": 3.5759,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3394693791712726e-05,
+      "loss": 3.5678,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3386307844202205e-05,
+      "loss": 3.5643,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.337792189669168e-05,
+      "loss": 3.5683,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336953594918116e-05,
+      "loss": 3.5658,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336115000167064e-05,
+      "loss": 3.5703,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.335276405416012e-05,
+      "loss": 3.5476,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.334439448545333e-05,
+      "loss": 3.5666,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.333600853794281e-05,
+      "loss": 3.5614,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.332762259043229e-05,
+      "loss": 3.5769,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.331923664292177e-05,
+      "loss": 3.5587,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3310867074214977e-05,
+      "loss": 3.5633,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.330248112670446e-05,
+      "loss": 3.5451,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.329409517919394e-05,
+      "loss": 3.5575,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.328570923168342e-05,
+      "loss": 3.5704,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.327733966297663e-05,
+      "loss": 3.5462,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.326895371546611e-05,
+      "loss": 3.575,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.326056776795559e-05,
+      "loss": 3.5626,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325218182044507e-05,
+      "loss": 3.5546,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.324381225173828e-05,
+      "loss": 3.5767,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.323542630422776e-05,
+      "loss": 3.552,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.322704035671724e-05,
+      "loss": 3.5512,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321865440920672e-05,
+      "loss": 3.5593,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321028484049993e-05,
+      "loss": 3.5722,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320189889298942e-05,
+      "loss": 3.5545,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.31935129454789e-05,
+      "loss": 3.5613,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.318512699796838e-05,
+      "loss": 3.5455,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.317677380806532e-05,
+      "loss": 3.5585,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3168387860554795e-05,
+      "loss": 3.5633,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3160001913044275e-05,
+      "loss": 3.5686,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3151615965533755e-05,
+      "loss": 3.563,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3143230018023235e-05,
+      "loss": 3.5695,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3134844070512715e-05,
+      "loss": 3.5797,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3126458123002195e-05,
+      "loss": 3.5619,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3118072175491675e-05,
+      "loss": 3.5683,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3109702606784884e-05,
+      "loss": 3.5627,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310131665927437e-05,
+      "loss": 3.5627,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309293071176385e-05,
+      "loss": 3.5547,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.308454476425333e-05,
+      "loss": 3.568,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.307617519554654e-05,
+      "loss": 3.5652,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.306778924803602e-05,
+      "loss": 3.577,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30594033005255e-05,
+      "loss": 3.5596,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.305101735301498e-05,
+      "loss": 3.5508,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304264778430819e-05,
+      "loss": 3.554,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.303426183679767e-05,
+      "loss": 3.5664,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.302587588928715e-05,
+      "loss": 3.5563,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.301748994177663e-05,
+      "loss": 3.5645,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300912037306984e-05,
+      "loss": 3.5597,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3000734425559324e-05,
+      "loss": 3.5546,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2992348478048804e-05,
+      "loss": 3.558,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2983962530538284e-05,
+      "loss": 3.5579,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.297559296183149e-05,
+      "loss": 3.5631,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.296720701432097e-05,
+      "loss": 3.5712,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.295882106681045e-05,
+      "loss": 3.5679,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.295043511929993e-05,
+      "loss": 3.5503,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.294206555059314e-05,
+      "loss": 3.5626,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.293367960308262e-05,
+      "loss": 3.5576,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.29252936555721e-05,
+      "loss": 3.5647,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.291692408686531e-05,
+      "loss": 3.5627,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.290853813935479e-05,
+      "loss": 3.5543,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.290015219184427e-05,
+      "loss": 3.5617,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289176624433376e-05,
+      "loss": 3.5591,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.288338029682324e-05,
+      "loss": 3.5592,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.287501072811645e-05,
+      "loss": 3.5588,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.286662478060593e-05,
+      "loss": 3.5534,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.285823883309541e-05,
+      "loss": 3.5638,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284985288558489e-05,
+      "loss": 3.5622,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2841483316878096e-05,
+      "loss": 3.554,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2833097369367576e-05,
+      "loss": 3.551,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2824711421857056e-05,
+      "loss": 3.5572,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2816341853150265e-05,
+      "loss": 3.5724,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2807955905639745e-05,
+      "loss": 3.5554,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2799569958129225e-05,
+      "loss": 3.5622,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279118401061871e-05,
+      "loss": 3.5642,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.278281444191192e-05,
+      "loss": 3.5677,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.27744284944014e-05,
+      "loss": 3.5612,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.276604254689088e-05,
+      "loss": 3.5626,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.275765659938036e-05,
+      "loss": 3.564,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.274928703067357e-05,
+      "loss": 3.5649,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.274090108316305e-05,
+      "loss": 3.5493,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.273251513565253e-05,
+      "loss": 3.5555,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.272412918814201e-05,
+      "loss": 3.5514,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.271575961943522e-05,
+      "loss": 3.5637,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.27073736719247e-05,
+      "loss": 3.5616,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.269898772441418e-05,
+      "loss": 3.5551,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2690601776903665e-05,
+      "loss": 3.5478,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2682232208196874e-05,
+      "loss": 3.5565,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2673846260686354e-05,
+      "loss": 3.5692,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2665460313175834e-05,
+      "loss": 3.5621,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2657074365665314e-05,
+      "loss": 3.5663,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264870479695852e-05,
+      "loss": 3.5514,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2640318849448e-05,
+      "loss": 3.5642,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263193290193748e-05,
+      "loss": 3.5665,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.262354695442696e-05,
+      "loss": 3.5597,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.261517738572017e-05,
+      "loss": 3.5553,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.260679143820965e-05,
+      "loss": 3.5587,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.259840549069913e-05,
+      "loss": 3.5642,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.259001954318862e-05,
+      "loss": 3.5672,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258164997448183e-05,
+      "loss": 3.5559,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.257326402697131e-05,
+      "loss": 3.5531,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.256487807946079e-05,
+      "loss": 3.5611,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.255649213195027e-05,
+      "loss": 3.5567,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.254812256324348e-05,
+      "loss": 3.5481,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253973661573296e-05,
+      "loss": 3.5653,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253135066822244e-05,
+      "loss": 3.5562,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.252296472071192e-05,
+      "loss": 3.569,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2514595152005126e-05,
+      "loss": 3.5487,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8440396785736084,
+      "eval_runtime": 304.6108,
+      "eval_samples_per_second": 1252.717,
+      "eval_steps_per_second": 39.148,
+      "step": 1068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2506209204494606e-05,
+      "loss": 3.5729,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2497823256984086e-05,
+      "loss": 3.5508,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.248943730947357e-05,
+      "loss": 3.5547,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.248105136196305e-05,
+      "loss": 3.5628,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247268179325626e-05,
+      "loss": 3.5598,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.246429584574574e-05,
+      "loss": 3.5588,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.245590989823522e-05,
+      "loss": 3.5546,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.24475239507247e-05,
+      "loss": 3.5538,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.243915438201791e-05,
+      "loss": 3.5538,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.243076843450739e-05,
+      "loss": 3.559,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242238248699687e-05,
+      "loss": 3.5585,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2413996539486344e-05,
+      "loss": 3.5576,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2405643349583296e-05,
+      "loss": 3.5665,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2397257402072775e-05,
+      "loss": 3.5428,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2388871454562255e-05,
+      "loss": 3.5495,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2380485507051735e-05,
+      "loss": 3.5423,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2372099559541215e-05,
+      "loss": 3.564,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2363713612030695e-05,
+      "loss": 3.5461,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2355327664520175e-05,
+      "loss": 3.5509,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.234694171700965e-05,
+      "loss": 3.5593,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2338572148302864e-05,
+      "loss": 3.5638,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2330186200792344e-05,
+      "loss": 3.5625,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232180025328182e-05,
+      "loss": 3.5578,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.23134143057713e-05,
+      "loss": 3.5545,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.230504473706451e-05,
+      "loss": 3.5663,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.229665878955399e-05,
+      "loss": 3.5485,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.228827284204347e-05,
+      "loss": 3.5549,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227988689453295e-05,
+      "loss": 3.5576,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227151732582617e-05,
+      "loss": 3.5466,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.226313137831564e-05,
+      "loss": 3.5421,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.225474543080512e-05,
+      "loss": 3.5479,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.22463594832946e-05,
+      "loss": 3.5545,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223798991458782e-05,
+      "loss": 3.5668,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222960396707729e-05,
+      "loss": 3.5598,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222121801956677e-05,
+      "loss": 3.5488,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221283207205625e-05,
+      "loss": 3.5508,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220446250334947e-05,
+      "loss": 3.5637,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.219607655583895e-05,
+      "loss": 3.5439,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.218769060832843e-05,
+      "loss": 3.5595,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.217930466081791e-05,
+      "loss": 3.5413,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2170935092111116e-05,
+      "loss": 3.543,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216256552340433e-05,
+      "loss": 3.5507,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.215417957589381e-05,
+      "loss": 3.5639,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.214579362838329e-05,
+      "loss": 3.5465,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2137407680872765e-05,
+      "loss": 3.5485,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2129021733362245e-05,
+      "loss": 3.5542,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2120635785851725e-05,
+      "loss": 3.5486,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2112249838341205e-05,
+      "loss": 3.5523,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2103863890830685e-05,
+      "loss": 3.5332,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20954943221239e-05,
+      "loss": 3.553,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.208710837461338e-05,
+      "loss": 3.5486,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207872242710286e-05,
+      "loss": 3.5598,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207033647959234e-05,
+      "loss": 3.5398,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206196691088555e-05,
+      "loss": 3.543,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.205358096337503e-05,
+      "loss": 3.54,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.204519501586451e-05,
+      "loss": 3.536,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.203680906835399e-05,
+      "loss": 3.5572,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20284394996472e-05,
+      "loss": 3.5295,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.202005355213668e-05,
+      "loss": 3.555,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201166760462616e-05,
+      "loss": 3.549,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.200328165711564e-05,
+      "loss": 3.541,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1994912088408854e-05,
+      "loss": 3.5585,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1986526140898334e-05,
+      "loss": 3.5371,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1978140193387814e-05,
+      "loss": 3.5374,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1969754245877294e-05,
+      "loss": 3.5373,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.19613846771705e-05,
+      "loss": 3.5584,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.195299872965998e-05,
+      "loss": 3.5395,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.194461278214946e-05,
+      "loss": 3.5383,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.193622683463894e-05,
+      "loss": 3.5356,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.192787364473589e-05,
+      "loss": 3.5397,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191948769722536e-05,
+      "loss": 3.55,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191110174971484e-05,
+      "loss": 3.552,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190271580220433e-05,
+      "loss": 3.545,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.189432985469381e-05,
+      "loss": 3.556,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.188594390718329e-05,
+      "loss": 3.5619,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.187755795967277e-05,
+      "loss": 3.5465,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.186917201216225e-05,
+      "loss": 3.5513,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1860818822259186e-05,
+      "loss": 3.5489,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1852432874748666e-05,
+      "loss": 3.5457,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1844046927238146e-05,
+      "loss": 3.5423,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1835660979727626e-05,
+      "loss": 3.5503,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1827275032217106e-05,
+      "loss": 3.5509,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1818889084706586e-05,
+      "loss": 3.5593,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1810503137196066e-05,
+      "loss": 3.5467,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1802117189685546e-05,
+      "loss": 3.5328,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.179374762097876e-05,
+      "loss": 3.5368,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.178536167346824e-05,
+      "loss": 3.5499,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.177697572595772e-05,
+      "loss": 3.5427,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.17685897784472e-05,
+      "loss": 3.5491,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.176022020974041e-05,
+      "loss": 3.5465,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.175183426222989e-05,
+      "loss": 3.5399,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.174344831471937e-05,
+      "loss": 3.5377,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.173506236720885e-05,
+      "loss": 3.5448,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.172669279850206e-05,
+      "loss": 3.5493,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.171830685099154e-05,
+      "loss": 3.5559,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170992090348102e-05,
+      "loss": 3.5497,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.17015349559705e-05,
+      "loss": 3.5338,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1693165387263715e-05,
+      "loss": 3.5477,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1684779439753195e-05,
+      "loss": 3.5475,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1676393492242675e-05,
+      "loss": 3.5461,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1668023923535884e-05,
+      "loss": 3.5477,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1659637976025364e-05,
+      "loss": 3.5364,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1651252028514844e-05,
+      "loss": 3.5459,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1642866081004324e-05,
+      "loss": 3.5431,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.163449651229753e-05,
+      "loss": 3.5444,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.162611056478701e-05,
+      "loss": 3.5393,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.161772461727649e-05,
+      "loss": 3.5418,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160933866976597e-05,
+      "loss": 3.5479,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160096910105919e-05,
+      "loss": 3.5485,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159258315354867e-05,
+      "loss": 3.5383,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.158419720603815e-05,
+      "loss": 3.5352,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157581125852763e-05,
+      "loss": 3.5382,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.156744168982084e-05,
+      "loss": 3.5578,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.155905574231032e-05,
+      "loss": 3.5413,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.15506697947998e-05,
+      "loss": 3.5427,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154228384728928e-05,
+      "loss": 3.5504,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.153391427858249e-05,
+      "loss": 3.5529,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.152552833107197e-05,
+      "loss": 3.5469,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.151714238356145e-05,
+      "loss": 3.549,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150875643605093e-05,
+      "loss": 3.5494,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150038686734414e-05,
+      "loss": 3.5471,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.149200091983362e-05,
+      "loss": 3.535,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.14836149723231e-05,
+      "loss": 3.5402,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.147522902481258e-05,
+      "loss": 3.5411,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.146685945610579e-05,
+      "loss": 3.5435,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.145847350859527e-05,
+      "loss": 3.5433,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.145008756108475e-05,
+      "loss": 3.5399,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144170161357423e-05,
+      "loss": 3.5364,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.143333204486744e-05,
+      "loss": 3.5439,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.142494609735692e-05,
+      "loss": 3.5433,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.14165601498464e-05,
+      "loss": 3.5514,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.140817420233588e-05,
+      "loss": 3.5526,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1399804633629096e-05,
+      "loss": 3.5354,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1391418686118576e-05,
+      "loss": 3.5446,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1383032738608056e-05,
+      "loss": 3.5485,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1374646791097536e-05,
+      "loss": 3.5466,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1366277222390745e-05,
+      "loss": 3.5391,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1357891274880225e-05,
+      "loss": 3.5464,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1349505327369705e-05,
+      "loss": 3.5483,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1341119379859185e-05,
+      "loss": 3.547,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1332749811152394e-05,
+      "loss": 3.5426,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1324363863641874e-05,
+      "loss": 3.5402,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.131599429493508e-05,
+      "loss": 3.5453,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.130760834742456e-05,
+      "loss": 3.5399,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.129922239991405e-05,
+      "loss": 3.5342,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.129083645240353e-05,
+      "loss": 3.5505,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128245050489301e-05,
+      "loss": 3.5413,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.127406455738249e-05,
+      "loss": 3.5538,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.126567860987197e-05,
+      "loss": 3.5351,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8453166484832764,
+      "eval_runtime": 309.1111,
+      "eval_samples_per_second": 1234.479,
+      "eval_steps_per_second": 38.578,
+      "step": 1144800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.125729266236144e-05,
+      "loss": 3.5422,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.124892309365466e-05,
+      "loss": 3.5378,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.124053714614414e-05,
+      "loss": 3.5383,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123215119863361e-05,
+      "loss": 3.5485,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.122376525112309e-05,
+      "loss": 3.543,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.121539568241631e-05,
+      "loss": 3.5452,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.120700973490579e-05,
+      "loss": 3.5377,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.119862378739527e-05,
+      "loss": 3.5377,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.119023783988475e-05,
+      "loss": 3.5421,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1181868271177964e-05,
+      "loss": 3.5419,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1173482323667443e-05,
+      "loss": 3.5387,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.116509637615692e-05,
+      "loss": 3.5499,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.115672680745013e-05,
+      "loss": 3.5505,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.114834085993961e-05,
+      "loss": 3.5283,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1139954912429086e-05,
+      "loss": 3.5294,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1131568964918566e-05,
+      "loss": 3.5287,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1123183017408046e-05,
+      "loss": 3.5446,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1114797069897526e-05,
+      "loss": 3.5354,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1106411122387005e-05,
+      "loss": 3.5352,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1098025174876485e-05,
+      "loss": 3.5384,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.10896556061697e-05,
+      "loss": 3.5484,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.108126965865918e-05,
+      "loss": 3.546,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107288371114866e-05,
+      "loss": 3.5415,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.106449776363814e-05,
+      "loss": 3.5397,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.105612819493135e-05,
+      "loss": 3.5537,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.104774224742083e-05,
+      "loss": 3.5301,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.103935629991031e-05,
+      "loss": 3.5385,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.103097035239979e-05,
+      "loss": 3.5436,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1022600783693e-05,
+      "loss": 3.5295,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.101421483618248e-05,
+      "loss": 3.5303,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.100582888867196e-05,
+      "loss": 3.5298,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.099744294116144e-05,
+      "loss": 3.54,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0989073372454655e-05,
+      "loss": 3.5511,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0980687424944135e-05,
+      "loss": 3.5399,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0972301477433615e-05,
+      "loss": 3.536,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0963915529923095e-05,
+      "loss": 3.5339,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0955545961216304e-05,
+      "loss": 3.5473,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0947160013705784e-05,
+      "loss": 3.5307,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0938774066195264e-05,
+      "loss": 3.5418,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0930388118684744e-05,
+      "loss": 3.5283,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092201854997795e-05,
+      "loss": 3.53,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.091364898127116e-05,
+      "loss": 3.5323,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.090526303376064e-05,
+      "loss": 3.5488,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.089687708625013e-05,
+      "loss": 3.5272,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.088849113873961e-05,
+      "loss": 3.5381,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.088010519122909e-05,
+      "loss": 3.5398,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087171924371857e-05,
+      "loss": 3.5357,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.086333329620805e-05,
+      "loss": 3.5384,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.085494734869753e-05,
+      "loss": 3.5159,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.084657777999074e-05,
+      "loss": 3.537,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.083819183248022e-05,
+      "loss": 3.538,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.08298058849697e-05,
+      "loss": 3.54,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.082141993745918e-05,
+      "loss": 3.5275,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0813050368752387e-05,
+      "loss": 3.5248,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0804664421241866e-05,
+      "loss": 3.5257,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0796278473731346e-05,
+      "loss": 3.5187,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0787892526220826e-05,
+      "loss": 3.5439,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.077952295751404e-05,
+      "loss": 3.5151,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.077113701000352e-05,
+      "loss": 3.5386,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0762751062493e-05,
+      "loss": 3.5356,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.075436511498248e-05,
+      "loss": 3.5255,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.074599554627569e-05,
+      "loss": 3.5421,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.073760959876517e-05,
+      "loss": 3.526,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.072922365125465e-05,
+      "loss": 3.524,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.072083770374413e-05,
+      "loss": 3.5228,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071246813503734e-05,
+      "loss": 3.5452,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.070408218752682e-05,
+      "loss": 3.5242,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.06956962400163e-05,
+      "loss": 3.5246,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.068731029250578e-05,
+      "loss": 3.5209,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0678957102602725e-05,
+      "loss": 3.5272,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0670571155092205e-05,
+      "loss": 3.5356,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0662185207581685e-05,
+      "loss": 3.538,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0653799260071165e-05,
+      "loss": 3.5269,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0645413312560645e-05,
+      "loss": 3.5402,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0637027365050125e-05,
+      "loss": 3.5473,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0628641417539605e-05,
+      "loss": 3.5373,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0620255470029085e-05,
+      "loss": 3.5322,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0611885901322294e-05,
+      "loss": 3.5362,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0603499953811774e-05,
+      "loss": 3.5327,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0595114006301254e-05,
+      "loss": 3.5293,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0586728058790734e-05,
+      "loss": 3.5296,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.057835849008395e-05,
+      "loss": 3.5403,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056997254257343e-05,
+      "loss": 3.5425,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056158659506291e-05,
+      "loss": 3.5343,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055320064755239e-05,
+      "loss": 3.515,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.05448310788456e-05,
+      "loss": 3.5227,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.053644513133508e-05,
+      "loss": 3.5347,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.052805918382456e-05,
+      "loss": 3.5287,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.051967323631404e-05,
+      "loss": 3.5356,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.051130366760725e-05,
+      "loss": 3.5303,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.050291772009673e-05,
+      "loss": 3.5243,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.049453177258621e-05,
+      "loss": 3.522,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.048614582507569e-05,
+      "loss": 3.5332,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.04777762563689e-05,
+      "loss": 3.5337,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.046939030885838e-05,
+      "loss": 3.5406,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.046100436134786e-05,
+      "loss": 3.5366,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0452618413837343e-05,
+      "loss": 3.516,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0444248845130552e-05,
+      "loss": 3.536,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0435862897620032e-05,
+      "loss": 3.5298,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0427476950109512e-05,
+      "loss": 3.5325,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.041910738140272e-05,
+      "loss": 3.5329,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0410721433892205e-05,
+      "loss": 3.524,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0402335486381685e-05,
+      "loss": 3.5274,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0393949538871164e-05,
+      "loss": 3.5305,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0385579970164374e-05,
+      "loss": 3.5329,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0377194022653854e-05,
+      "loss": 3.5238,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0368808075143333e-05,
+      "loss": 3.5284,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0360422127632813e-05,
+      "loss": 3.5327,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0352052558926026e-05,
+      "loss": 3.5357,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0343666611415506e-05,
+      "loss": 3.5205,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0335280663904986e-05,
+      "loss": 3.5226,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.032689471639446e-05,
+      "loss": 3.5209,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0318525147687675e-05,
+      "loss": 3.5429,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0310139200177158e-05,
+      "loss": 3.5271,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0301753252666638e-05,
+      "loss": 3.5288,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.029336730515611e-05,
+      "loss": 3.5351,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0284997736449327e-05,
+      "loss": 3.539,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0276611788938807e-05,
+      "loss": 3.5292,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0268225841428287e-05,
+      "loss": 3.533,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0259839893917764e-05,
+      "loss": 3.5368,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.025147032521098e-05,
+      "loss": 3.529,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.024308437770046e-05,
+      "loss": 3.5218,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0234698430189933e-05,
+      "loss": 3.5263,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0226312482679413e-05,
+      "loss": 3.5224,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.021794291397263e-05,
+      "loss": 3.5291,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0209556966462112e-05,
+      "loss": 3.5272,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0201171018951585e-05,
+      "loss": 3.5294,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0192785071441065e-05,
+      "loss": 3.523,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.018441550273428e-05,
+      "loss": 3.5255,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.017602955522376e-05,
+      "loss": 3.5284,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0167643607713237e-05,
+      "loss": 3.5393,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0159257660202717e-05,
+      "loss": 3.5395,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0150888091495933e-05,
+      "loss": 3.5272,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0142502143985406e-05,
+      "loss": 3.5253,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0134116196474886e-05,
+      "loss": 3.5303,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0125730248964366e-05,
+      "loss": 3.5317,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0117360680257582e-05,
+      "loss": 3.5296,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.010897473274706e-05,
+      "loss": 3.5314,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.010058878523654e-05,
+      "loss": 3.5359,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.009220283772602e-05,
+      "loss": 3.5314,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0083833269019235e-05,
+      "loss": 3.5297,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.007544732150871e-05,
+      "loss": 3.523,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0067077752801927e-05,
+      "loss": 3.5295,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0058691805291407e-05,
+      "loss": 3.5258,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.005030585778088e-05,
+      "loss": 3.5191,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.004191991027036e-05,
+      "loss": 3.5393,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.003353396275984e-05,
+      "loss": 3.5311,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.002514801524932e-05,
+      "loss": 3.5338,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0016762067738803e-05,
+      "loss": 3.5222,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.0008376120228283e-05,
+      "loss": 3.5265,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.843841552734375,
+      "eval_runtime": 307.3634,
+      "eval_samples_per_second": 1241.498,
+      "eval_steps_per_second": 38.798,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9999990172717763e-05,
+      "loss": 3.5263,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9991620604010972e-05,
+      "loss": 3.519,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9983234656500452e-05,
+      "loss": 3.5352,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9974848708989932e-05,
+      "loss": 3.5288,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9966462761479412e-05,
+      "loss": 3.53,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9958076813968892e-05,
+      "loss": 3.5217,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9949690866458376e-05,
+      "loss": 3.5287,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9941304918947855e-05,
+      "loss": 3.5212,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9932935350241065e-05,
+      "loss": 3.5307,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9924549402730545e-05,
+      "loss": 3.5266,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9916163455220025e-05,
+      "loss": 3.5343,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9907777507709504e-05,
+      "loss": 3.5321,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9899407939002717e-05,
+      "loss": 3.5187,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9891021991492197e-05,
+      "loss": 3.5187,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9882636043981677e-05,
+      "loss": 3.5126,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9874250096471157e-05,
+      "loss": 3.5282,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9865864148960637e-05,
+      "loss": 3.5217,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9857478201450117e-05,
+      "loss": 3.5256,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9849092253939597e-05,
+      "loss": 3.5186,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.984072268523281e-05,
+      "loss": 3.5393,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.983233673772229e-05,
+      "loss": 3.5293,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.982395079021177e-05,
+      "loss": 3.5276,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9815564842701242e-05,
+      "loss": 3.5251,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9807195273994458e-05,
+      "loss": 3.5366,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9798809326483938e-05,
+      "loss": 3.5168,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.979042337897342e-05,
+      "loss": 3.5279,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9782037431462895e-05,
+      "loss": 3.5237,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.977366786275611e-05,
+      "loss": 3.5175,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.976528191524559e-05,
+      "loss": 3.5156,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9756895967735067e-05,
+      "loss": 3.5151,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9748510020224547e-05,
+      "loss": 3.527,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9740140451517763e-05,
+      "loss": 3.5342,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9731754504007243e-05,
+      "loss": 3.5266,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9723368556496716e-05,
+      "loss": 3.5233,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9714982608986196e-05,
+      "loss": 3.5209,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9706613040279412e-05,
+      "loss": 3.5301,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9698227092768892e-05,
+      "loss": 3.52,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.968984114525837e-05,
+      "loss": 3.5283,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9681455197747848e-05,
+      "loss": 3.5105,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9673085629041064e-05,
+      "loss": 3.5201,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9664716060334273e-05,
+      "loss": 3.5208,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9656330112823753e-05,
+      "loss": 3.5324,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9647944165313237e-05,
+      "loss": 3.5171,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9639558217802716e-05,
+      "loss": 3.5218,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.963117227029219e-05,
+      "loss": 3.5245,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962278632278167e-05,
+      "loss": 3.5183,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.961440037527115e-05,
+      "loss": 3.5274,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.960601442776063e-05,
+      "loss": 3.5001,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9597644859053842e-05,
+      "loss": 3.5206,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9589258911543322e-05,
+      "loss": 3.525,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9580872964032802e-05,
+      "loss": 3.5295,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9572487016522282e-05,
+      "loss": 3.5151,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.956411744781549e-05,
+      "loss": 3.506,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9555731500304974e-05,
+      "loss": 3.5157,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9547345552794454e-05,
+      "loss": 3.5035,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9538959605283934e-05,
+      "loss": 3.5315,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9530590036577143e-05,
+      "loss": 3.5027,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9522204089066623e-05,
+      "loss": 3.5243,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9513818141556103e-05,
+      "loss": 3.5204,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9505432194045583e-05,
+      "loss": 3.5125,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9497062625338796e-05,
+      "loss": 3.5266,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9488676677828276e-05,
+      "loss": 3.511,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9480290730317756e-05,
+      "loss": 3.5081,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9471904782807236e-05,
+      "loss": 3.5078,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9463535214100445e-05,
+      "loss": 3.529,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9455149266589928e-05,
+      "loss": 3.512,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9446763319079408e-05,
+      "loss": 3.5132,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9438377371568888e-05,
+      "loss": 3.5055,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9430024181665833e-05,
+      "loss": 3.5111,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9421638234155313e-05,
+      "loss": 3.5231,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.941325228664479e-05,
+      "loss": 3.5239,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.940486633913427e-05,
+      "loss": 3.5129,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.939648039162375e-05,
+      "loss": 3.5263,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.938809444411323e-05,
+      "loss": 3.5323,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.937970849660271e-05,
+      "loss": 3.5243,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.937132254909219e-05,
+      "loss": 3.5142,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9362969359189134e-05,
+      "loss": 3.5232,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.935458341167861e-05,
+      "loss": 3.5194,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.934619746416809e-05,
+      "loss": 3.5139,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.933781151665757e-05,
+      "loss": 3.518,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.932942556914705e-05,
+      "loss": 3.5221,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.932103962163653e-05,
+      "loss": 3.5282,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.931265367412601e-05,
+      "loss": 3.5216,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.930426772661549e-05,
+      "loss": 3.5029,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9295898157908703e-05,
+      "loss": 3.511,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9287512210398183e-05,
+      "loss": 3.5166,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9279126262887663e-05,
+      "loss": 3.5198,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9270740315377143e-05,
+      "loss": 3.5191,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9262370746670352e-05,
+      "loss": 3.5182,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9253984799159835e-05,
+      "loss": 3.5134,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9245598851649315e-05,
+      "loss": 3.5029,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9237212904138795e-05,
+      "loss": 3.5169,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9228843335432004e-05,
+      "loss": 3.5191,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9220457387921484e-05,
+      "loss": 3.5235,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9212071440410964e-05,
+      "loss": 3.524,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9203685492900444e-05,
+      "loss": 3.5064,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9195315924193657e-05,
+      "loss": 3.5171,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9186929976683137e-05,
+      "loss": 3.5192,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9178544029172617e-05,
+      "loss": 3.5173,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9170174460465826e-05,
+      "loss": 3.5195,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9161788512955306e-05,
+      "loss": 3.5125,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.915340256544479e-05,
+      "loss": 3.5128,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.914501661793427e-05,
+      "loss": 3.5182,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9136647049227478e-05,
+      "loss": 3.5187,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9128261101716958e-05,
+      "loss": 3.5106,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9119875154206438e-05,
+      "loss": 3.5114,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9111489206695918e-05,
+      "loss": 3.5239,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.910311963798913e-05,
+      "loss": 3.5157,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.909473369047861e-05,
+      "loss": 3.5107,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.908634774296809e-05,
+      "loss": 3.5093,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.907796179545757e-05,
+      "loss": 3.5085,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.906959222675078e-05,
+      "loss": 3.5307,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.906120627924026e-05,
+      "loss": 3.5115,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.905282033172974e-05,
+      "loss": 3.517,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9044450763022952e-05,
+      "loss": 3.5184,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9036064815512432e-05,
+      "loss": 3.526,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.902767886800191e-05,
+      "loss": 3.5117,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.901929292049139e-05,
+      "loss": 3.517,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.90109233517846e-05,
+      "loss": 3.526,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9002537404274084e-05,
+      "loss": 3.5147,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8994151456763564e-05,
+      "loss": 3.5069,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8985765509253044e-05,
+      "loss": 3.5118,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8977379561742524e-05,
+      "loss": 3.5115,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8968993614232004e-05,
+      "loss": 3.5154,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8960607666721484e-05,
+      "loss": 3.5127,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8952221719210964e-05,
+      "loss": 3.5175,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8943852150504176e-05,
+      "loss": 3.508,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8935466202993656e-05,
+      "loss": 3.5107,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8927080255483136e-05,
+      "loss": 3.5117,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8918710686776345e-05,
+      "loss": 3.5296,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8910324739265825e-05,
+      "loss": 3.5255,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8901938791755305e-05,
+      "loss": 3.512,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8893552844244785e-05,
+      "loss": 3.5099,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.888516689673427e-05,
+      "loss": 3.5158,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.887678094922374e-05,
+      "loss": 3.5207,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.886839500171322e-05,
+      "loss": 3.5175,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.88600090542027e-05,
+      "loss": 3.515,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8851639485495917e-05,
+      "loss": 3.5214,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8843253537985394e-05,
+      "loss": 3.5149,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8834867590474874e-05,
+      "loss": 3.5183,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8826481642964354e-05,
+      "loss": 3.5115,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.88181284530613e-05,
+      "loss": 3.5185,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.880974250555078e-05,
+      "loss": 3.5103,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.880135655804026e-05,
+      "loss": 3.503,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.879297061052974e-05,
+      "loss": 3.5236,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8784584663019215e-05,
+      "loss": 3.5185,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8776198715508695e-05,
+      "loss": 3.5195,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8767812767998175e-05,
+      "loss": 3.5073,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8759426820487655e-05,
+      "loss": 3.5122,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8459231853485107,
+      "eval_runtime": 304.0081,
+      "eval_samples_per_second": 1255.2,
+      "eval_steps_per_second": 39.226,
+      "step": 1297440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8751057251780868e-05,
+      "loss": 3.5138,
+      "step": 1297920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8742687683074084e-05,
+      "loss": 3.5097,
+      "step": 1298432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8734301735563564e-05,
+      "loss": 3.5219,
+      "step": 1298944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8725915788053037e-05,
+      "loss": 3.5101,
+      "step": 1299456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8717529840542517e-05,
+      "loss": 3.5201,
+      "step": 1299968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8709143893031997e-05,
+      "loss": 3.5065,
+      "step": 1300480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8700757945521477e-05,
+      "loss": 3.5165,
+      "step": 1300992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.869237199801096e-05,
+      "loss": 3.5068,
+      "step": 1301504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.868400242930417e-05,
+      "loss": 3.5164,
+      "step": 1302016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.867561648179365e-05,
+      "loss": 3.515,
+      "step": 1302528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.866723053428313e-05,
+      "loss": 3.52,
+      "step": 1303040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8658860965576338e-05,
+      "loss": 3.5176,
+      "step": 1303552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.865047501806582e-05,
+      "loss": 3.5044,
+      "step": 1304064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.86420890705553e-05,
+      "loss": 3.5048,
+      "step": 1304576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.863370312304478e-05,
+      "loss": 3.5076,
+      "step": 1305088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.862531717553426e-05,
+      "loss": 3.505,
+      "step": 1305600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.861693122802374e-05,
+      "loss": 3.51,
+      "step": 1306112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.860854528051322e-05,
+      "loss": 3.5106,
+      "step": 1306624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.86001593330027e-05,
+      "loss": 3.5048,
+      "step": 1307136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8591789764295914e-05,
+      "loss": 3.5276,
+      "step": 1307648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8583403816785394e-05,
+      "loss": 3.5153,
+      "step": 1308160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8575017869274874e-05,
+      "loss": 3.515,
+      "step": 1308672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8566631921764354e-05,
+      "loss": 3.5107,
+      "step": 1309184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8558262353057563e-05,
+      "loss": 3.5261,
+      "step": 1309696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8549876405547043e-05,
+      "loss": 3.4952,
+      "step": 1310208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8541490458036523e-05,
+      "loss": 3.5179,
+      "step": 1310720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8533104510526006e-05,
+      "loss": 3.5066,
+      "step": 1311232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8524734941819215e-05,
+      "loss": 3.509,
+      "step": 1311744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8516348994308695e-05,
+      "loss": 3.5019,
+      "step": 1312256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8507963046798175e-05,
+      "loss": 3.5006,
+      "step": 1312768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8499577099287655e-05,
+      "loss": 3.5147,
+      "step": 1313280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8491207530580867e-05,
+      "loss": 3.5226,
+      "step": 1313792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8482821583070347e-05,
+      "loss": 3.5106,
+      "step": 1314304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8474435635559827e-05,
+      "loss": 3.5138,
+      "step": 1314816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8466049688049307e-05,
+      "loss": 3.5032,
+      "step": 1315328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8457680119342516e-05,
+      "loss": 3.5151,
+      "step": 1315840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8449294171831996e-05,
+      "loss": 3.5068,
+      "step": 1316352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8440908224321476e-05,
+      "loss": 3.5149,
+      "step": 1316864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8432522276810956e-05,
+      "loss": 3.4997,
+      "step": 1317376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.842415270810417e-05,
+      "loss": 3.5077,
+      "step": 1317888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8415783139397378e-05,
+      "loss": 3.5044,
+      "step": 1318400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8407397191886858e-05,
+      "loss": 3.52,
+      "step": 1318912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8399011244376338e-05,
+      "loss": 3.5011,
+      "step": 1319424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.839062529686582e-05,
+      "loss": 3.5075,
+      "step": 1319936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.83822393493553e-05,
+      "loss": 3.5122,
+      "step": 1320448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.837385340184478e-05,
+      "loss": 3.5036,
+      "step": 1320960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.836546745433426e-05,
+      "loss": 3.5123,
+      "step": 1321472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.835708150682374e-05,
+      "loss": 3.4902,
+      "step": 1321984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.834871193811695e-05,
+      "loss": 3.5054,
+      "step": 1322496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.834032599060643e-05,
+      "loss": 3.5123,
+      "step": 1323008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.833194004309591e-05,
+      "loss": 3.5074,
+      "step": 1323520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8323554095585393e-05,
+      "loss": 3.5017,
+      "step": 1324032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8315184526878602e-05,
+      "loss": 3.499,
+      "step": 1324544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8306798579368082e-05,
+      "loss": 3.5014,
+      "step": 1325056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8298412631857562e-05,
+      "loss": 3.4921,
+      "step": 1325568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8290026684347042e-05,
+      "loss": 3.5136,
+      "step": 1326080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8281657115640255e-05,
+      "loss": 3.4927,
+      "step": 1326592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8273271168129735e-05,
+      "loss": 3.5093,
+      "step": 1327104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8264885220619215e-05,
+      "loss": 3.5064,
+      "step": 1327616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8256499273108694e-05,
+      "loss": 3.5026,
+      "step": 1328128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8248129704401904e-05,
+      "loss": 3.5106,
+      "step": 1328640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8239743756891384e-05,
+      "loss": 3.5034,
+      "step": 1329152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8231357809380863e-05,
+      "loss": 3.492,
+      "step": 1329664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8222971861870347e-05,
+      "loss": 3.4906,
+      "step": 1330176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8214602293163556e-05,
+      "loss": 3.5177,
+      "step": 1330688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8206216345653036e-05,
+      "loss": 3.501,
+      "step": 1331200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8197830398142516e-05,
+      "loss": 3.4971,
+      "step": 1331712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8189444450631996e-05,
+      "loss": 3.4932,
+      "step": 1332224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8181091260728937e-05,
+      "loss": 3.5002,
+      "step": 1332736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8172705313218417e-05,
+      "loss": 3.5053,
+      "step": 1333248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8164319365707897e-05,
+      "loss": 3.5106,
+      "step": 1333760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8155933418197377e-05,
+      "loss": 3.5004,
+      "step": 1334272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8147547470686857e-05,
+      "loss": 3.5127,
+      "step": 1334784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8139161523176337e-05,
+      "loss": 3.5152,
+      "step": 1335296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8130775575665817e-05,
+      "loss": 3.5159,
+      "step": 1335808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8122389628155294e-05,
+      "loss": 3.4979,
+      "step": 1336320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.811403643825224e-05,
+      "loss": 3.513,
+      "step": 1336832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.810565049074172e-05,
+      "loss": 3.5036,
+      "step": 1337344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.80972645432312e-05,
+      "loss": 3.5041,
+      "step": 1337856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.808887859572068e-05,
+      "loss": 3.5035,
+      "step": 1338368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8080492648210162e-05,
+      "loss": 3.508,
+      "step": 1338880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8072106700699642e-05,
+      "loss": 3.5154,
+      "step": 1339392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8063720753189115e-05,
+      "loss": 3.5062,
+      "step": 1339904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8055334805678595e-05,
+      "loss": 3.4889,
+      "step": 1340416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.804696523697181e-05,
+      "loss": 3.4976,
+      "step": 1340928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.803857928946129e-05,
+      "loss": 3.5011,
+      "step": 1341440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8030193341950767e-05,
+      "loss": 3.5093,
+      "step": 1341952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8021807394440247e-05,
+      "loss": 3.5064,
+      "step": 1342464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8013437825733463e-05,
+      "loss": 3.502,
+      "step": 1342976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8005051878222943e-05,
+      "loss": 3.5038,
+      "step": 1343488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7996665930712416e-05,
+      "loss": 3.4874,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.79882799832019e-05,
+      "loss": 3.5029,
+      "step": 1344512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7979910414495116e-05,
+      "loss": 3.5056,
+      "step": 1345024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.797152446698459e-05,
+      "loss": 3.506,
+      "step": 1345536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.796313851947407e-05,
+      "loss": 3.5093,
+      "step": 1346048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.795475257196355e-05,
+      "loss": 3.4948,
+      "step": 1346560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7946383003256765e-05,
+      "loss": 3.5026,
+      "step": 1347072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.793799705574624e-05,
+      "loss": 3.5079,
+      "step": 1347584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.792961110823572e-05,
+      "loss": 3.5062,
+      "step": 1348096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7921241539528937e-05,
+      "loss": 3.5009,
+      "step": 1348608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7912855592018417e-05,
+      "loss": 3.5028,
+      "step": 1349120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.790446964450789e-05,
+      "loss": 3.4977,
+      "step": 1349632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.789608369699737e-05,
+      "loss": 3.5068,
+      "step": 1350144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7887714128290586e-05,
+      "loss": 3.5089,
+      "step": 1350656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7879328180780062e-05,
+      "loss": 3.4924,
+      "step": 1351168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7870942233269542e-05,
+      "loss": 3.5025,
+      "step": 1351680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7862556285759022e-05,
+      "loss": 3.51,
+      "step": 1352192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7854170338248502e-05,
+      "loss": 3.5023,
+      "step": 1352704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7845784390737982e-05,
+      "loss": 3.4947,
+      "step": 1353216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7837398443227462e-05,
+      "loss": 3.4979,
+      "step": 1353728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7829012495716946e-05,
+      "loss": 3.4939,
+      "step": 1354240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7820642927010155e-05,
+      "loss": 3.5171,
+      "step": 1354752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7812256979499635e-05,
+      "loss": 3.4991,
+      "step": 1355264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7803871031989115e-05,
+      "loss": 3.5045,
+      "step": 1355776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7795501463282324e-05,
+      "loss": 3.5007,
+      "step": 1356288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7787115515771807e-05,
+      "loss": 3.5147,
+      "step": 1356800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7778729568261287e-05,
+      "loss": 3.4977,
+      "step": 1357312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7770343620750767e-05,
+      "loss": 3.5071,
+      "step": 1357824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7761974052043976e-05,
+      "loss": 3.5112,
+      "step": 1358336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7753588104533456e-05,
+      "loss": 3.5013,
+      "step": 1358848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7745202157022936e-05,
+      "loss": 3.4975,
+      "step": 1359360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7736816209512416e-05,
+      "loss": 3.4994,
+      "step": 1359872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.77284302620019e-05,
+      "loss": 3.4944,
+      "step": 1360384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.772004431449138e-05,
+      "loss": 3.5005,
+      "step": 1360896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.771165836698086e-05,
+      "loss": 3.505,
+      "step": 1361408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.770327241947034e-05,
+      "loss": 3.5028,
+      "step": 1361920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7694902850763548e-05,
+      "loss": 3.4951,
+      "step": 1362432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7686516903253028e-05,
+      "loss": 3.4956,
+      "step": 1362944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7678130955742508e-05,
+      "loss": 3.4987,
+      "step": 1363456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7669745008231988e-05,
+      "loss": 3.5154,
+      "step": 1363968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.76613754395252e-05,
+      "loss": 3.5134,
+      "step": 1364480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.765298949201468e-05,
+      "loss": 3.5019,
+      "step": 1364992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.764460354450416e-05,
+      "loss": 3.4924,
+      "step": 1365504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.763621759699364e-05,
+      "loss": 3.5052,
+      "step": 1366016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7627848028286853e-05,
+      "loss": 3.5024,
+      "step": 1366528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7619462080776333e-05,
+      "loss": 3.506,
+      "step": 1367040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7611076133265813e-05,
+      "loss": 3.5003,
+      "step": 1367552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7602690185755293e-05,
+      "loss": 3.5097,
+      "step": 1368064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7594320617048502e-05,
+      "loss": 3.501,
+      "step": 1368576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7585934669537982e-05,
+      "loss": 3.5067,
+      "step": 1369088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7577548722027462e-05,
+      "loss": 3.5021,
+      "step": 1369600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7569179153320674e-05,
+      "loss": 3.5037,
+      "step": 1370112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7560793205810154e-05,
+      "loss": 3.4968,
+      "step": 1370624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7552407258299634e-05,
+      "loss": 3.492,
+      "step": 1371136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7544021310789114e-05,
+      "loss": 3.5098,
+      "step": 1371648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7535635363278594e-05,
+      "loss": 3.5081,
+      "step": 1372160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7527249415768074e-05,
+      "loss": 3.5079,
+      "step": 1372672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.751886346825755e-05,
+      "loss": 3.4953,
+      "step": 1373184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.751047752074703e-05,
+      "loss": 3.4959,
+      "step": 1373696
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.847353219985962,
+      "eval_runtime": 307.7299,
+      "eval_samples_per_second": 1240.019,
+      "eval_steps_per_second": 38.752,
+      "step": 1373760
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 9.563993345943429e+17,
+  "trial_name": null,
+  "trial_params": null
+}