diff --git "a/checkpoint-1221120/trainer_state.json" "b/checkpoint-1221120/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1221120/trainer_state.json"
@@ -0,0 +1,14463 @@
+{
+  "best_metric": 3.84504771232605,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/det-adj-noun/transformer/1/checkpoints/checkpoint-992160",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 1221120,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 11.0484,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8231,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1887,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9796,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8076,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7017,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6021,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5261,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4485,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4057,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3524,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3114,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2653,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2144,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1763,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1401,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1109,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0873,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0499,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0212,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0132,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 4.97,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9591,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9332,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.924,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.8918,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8703,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8532,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8429,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8209,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.8078,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7984,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7763,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 4.7702,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7544,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7446,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969820416244367e-05,
+      "loss": 4.731,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.6995,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.707,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6792,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6872,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6706,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6715,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6541,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 4.644,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6414,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6329,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6224,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.6057,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95892359812181e-05,
+      "loss": 4.5916,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958085003370758e-05,
+      "loss": 4.6018,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957246408619706e-05,
+      "loss": 4.5889,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 4.5831,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5605,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5656,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5446,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5658,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 4.5346,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5471,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5309,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5135,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.519,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5117,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4896,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4899,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5028,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.4809,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4733,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4636,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.4683,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4828,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940482703000532e-05,
+      "loss": 4.4691,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396441082494794e-05,
+      "loss": 4.4539,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388055134984274e-05,
+      "loss": 4.4579,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379669187473754e-05,
+      "loss": 4.4594,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 4.4422,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 4.4354,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 4.438,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 4.4241,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 4.4259,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 4.424,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 4.4208,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.4236,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 4.414,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.3973,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.4026,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.3971,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 4.4014,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926231505873767e-05,
+      "loss": 4.3925,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925394549003088e-05,
+      "loss": 4.3906,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924555954252036e-05,
+      "loss": 4.3898,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923717359500984e-05,
+      "loss": 4.3824,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922878764749932e-05,
+      "loss": 4.3806,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 4.3738,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.365,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920364618377149e-05,
+      "loss": 4.3725,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919526023626097e-05,
+      "loss": 4.36,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918687428875045e-05,
+      "loss": 4.3584,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917848834123993e-05,
+      "loss": 4.3569,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917010239372941e-05,
+      "loss": 4.3583,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.3597,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.3484,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3411,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913657498249106e-05,
+      "loss": 4.3386,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912818903498054e-05,
+      "loss": 4.3364,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911980308747002e-05,
+      "loss": 4.3204,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911143351876323e-05,
+      "loss": 4.3192,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.3313,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3266,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.3135,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.3102,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3141,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3246,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3183,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90443623174828e-05,
+      "loss": 4.309,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.3139,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027606801265495e-05,
+      "loss": 4.3142,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.299,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.2944,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.2987,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994063011223415e-05,
+      "loss": 4.2982,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985677063712895e-05,
+      "loss": 4.2805,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977307495006104e-05,
+      "loss": 4.283,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968921547495584e-05,
+      "loss": 4.2819,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.2793,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2807,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894378008376775e-05,
+      "loss": 4.2873,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893539413625723e-05,
+      "loss": 4.271,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.2726,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2625,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891025267252941e-05,
+      "loss": 4.2779,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890186672501889e-05,
+      "loss": 4.2706,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889348077750837e-05,
+      "loss": 4.2629,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888509482999785e-05,
+      "loss": 4.2631,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887672526129106e-05,
+      "loss": 4.268,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886833931378054e-05,
+      "loss": 4.2652,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885995336627002e-05,
+      "loss": 4.2601,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515674187595e-05,
+      "loss": 4.2548,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8843197850052706e-05,
+      "loss": 4.2565,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834811902542186e-05,
+      "loss": 4.2567,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8826425955031666e-05,
+      "loss": 4.2552,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8818040007521146e-05,
+      "loss": 4.2382,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8809654060010626e-05,
+      "loss": 4.2466,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8801268112500106e-05,
+      "loss": 4.2348,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879289854379332e-05,
+      "loss": 4.2254,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87845125962828e-05,
+      "loss": 4.2519,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877612664877228e-05,
+      "loss": 4.236,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876774070126176e-05,
+      "loss": 4.2375,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875937113255497e-05,
+      "loss": 4.2367,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875098518504445e-05,
+      "loss": 4.2179,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.200989246368408,
+      "eval_runtime": 303.1099,
+      "eval_samples_per_second": 1258.92,
+      "eval_steps_per_second": 39.342,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874259923753393e-05,
+      "loss": 4.2206,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873421329002341e-05,
+      "loss": 4.2154,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872582734251289e-05,
+      "loss": 4.2347,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87174577738061e-05,
+      "loss": 4.2183,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870907182629558e-05,
+      "loss": 4.2267,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870068587878506e-05,
+      "loss": 4.2062,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869229993127454e-05,
+      "loss": 4.2073,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683930362567756e-05,
+      "loss": 4.2029,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675544415057236e-05,
+      "loss": 4.2093,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667158467546716e-05,
+      "loss": 4.2078,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658772520036196e-05,
+      "loss": 4.2138,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650402951329405e-05,
+      "loss": 4.2148,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642033382622614e-05,
+      "loss": 4.1937,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633647435112094e-05,
+      "loss": 4.1917,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8625261487601574e-05,
+      "loss": 4.1904,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8616875540091054e-05,
+      "loss": 4.1823,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608489592580534e-05,
+      "loss": 4.1923,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8600103645070014e-05,
+      "loss": 4.18,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591717697559493e-05,
+      "loss": 4.1807,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858333175004898e-05,
+      "loss": 4.2034,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857496218134219e-05,
+      "loss": 4.1787,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856657623383167e-05,
+      "loss": 4.1796,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855819028632115e-05,
+      "loss": 4.1809,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854980433881063e-05,
+      "loss": 4.185,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854143477010384e-05,
+      "loss": 4.1734,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853304882259332e-05,
+      "loss": 4.1691,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85246628750828e-05,
+      "loss": 4.1669,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851627692757228e-05,
+      "loss": 4.1727,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850790735886549e-05,
+      "loss": 4.1548,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849952141135497e-05,
+      "loss": 4.1653,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849113546384445e-05,
+      "loss": 4.1647,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848276589513766e-05,
+      "loss": 4.1624,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847437994762714e-05,
+      "loss": 4.1634,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846599400011662e-05,
+      "loss": 4.1538,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84576080526061e-05,
+      "loss": 4.1549,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844922210509558e-05,
+      "loss": 4.1568,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440836157585056e-05,
+      "loss": 4.1377,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432450210074536e-05,
+      "loss": 4.1478,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8424064262564016e-05,
+      "loss": 4.1345,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415694693857225e-05,
+      "loss": 4.1438,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8407308746346705e-05,
+      "loss": 4.1458,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398922798836185e-05,
+      "loss": 4.1513,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839053685132567e-05,
+      "loss": 4.1426,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838216728261888e-05,
+      "loss": 4.1381,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837378133510836e-05,
+      "loss": 4.1472,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836539538759784e-05,
+      "loss": 4.1407,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835700944008732e-05,
+      "loss": 4.1318,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834863987138053e-05,
+      "loss": 4.1281,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834025392387001e-05,
+      "loss": 4.1213,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833186797635949e-05,
+      "loss": 4.1348,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832348202884897e-05,
+      "loss": 4.1297,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831511246014218e-05,
+      "loss": 4.126,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830672651263166e-05,
+      "loss": 4.1202,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829834056512114e-05,
+      "loss": 4.1255,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289954617610625e-05,
+      "loss": 4.1096,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281585048903834e-05,
+      "loss": 4.1312,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273199101393314e-05,
+      "loss": 4.1131,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264813153882794e-05,
+      "loss": 4.1294,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256427206372274e-05,
+      "loss": 4.1176,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248057637665483e-05,
+      "loss": 4.104,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823967169015496e-05,
+      "loss": 4.1083,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823128574264444e-05,
+      "loss": 4.1133,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822289979513392e-05,
+      "loss": 4.0948,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821453022642713e-05,
+      "loss": 4.0991,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820614427891661e-05,
+      "loss": 4.1118,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819775833140609e-05,
+      "loss": 4.1016,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818937238389558e-05,
+      "loss": 4.0953,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818100281518879e-05,
+      "loss": 4.0929,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817261686767827e-05,
+      "loss": 4.0942,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816423092016775e-05,
+      "loss": 4.1115,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815584497265723e-05,
+      "loss": 4.1059,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814747540395044e-05,
+      "loss": 4.0986,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813908945643992e-05,
+      "loss": 4.1041,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81307035089294e-05,
+      "loss": 4.1011,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812231756141888e-05,
+      "loss": 4.0969,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113947992712086e-05,
+      "loss": 4.0836,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105562045201566e-05,
+      "loss": 4.0975,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8097176097691046e-05,
+      "loss": 4.0907,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808879015018053e-05,
+      "loss": 4.0873,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808042058147374e-05,
+      "loss": 4.0947,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807203463396322e-05,
+      "loss": 4.0892,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80636486864527e-05,
+      "loss": 4.0975,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805526273894218e-05,
+      "loss": 4.0911,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804689317023539e-05,
+      "loss": 4.0717,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803850722272487e-05,
+      "loss": 4.0839,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803012127521435e-05,
+      "loss": 4.0803,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802173532770383e-05,
+      "loss": 4.0875,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801336575899704e-05,
+      "loss": 4.0811,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800497981148652e-05,
+      "loss": 4.0811,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996610242779736e-05,
+      "loss": 4.0823,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988224295269215e-05,
+      "loss": 4.0778,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979838347758695e-05,
+      "loss": 4.0803,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971452400248175e-05,
+      "loss": 4.0769,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7963066452737655e-05,
+      "loss": 4.0679,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954696884030864e-05,
+      "loss": 4.0814,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946310936520344e-05,
+      "loss": 4.0761,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937924989009824e-05,
+      "loss": 4.067,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929539041499304e-05,
+      "loss": 4.075,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7921153093988784e-05,
+      "loss": 4.0735,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912767146478264e-05,
+      "loss": 4.08,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7904381198967744e-05,
+      "loss": 4.0689,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895995251457224e-05,
+      "loss": 4.0602,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788762568275043e-05,
+      "loss": 4.064,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787923973523992e-05,
+      "loss": 4.0657,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78708537877294e-05,
+      "loss": 4.043,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786248421902261e-05,
+      "loss": 4.0585,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785409827151209e-05,
+      "loss": 4.0619,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784571232400157e-05,
+      "loss": 4.0608,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783732637649105e-05,
+      "loss": 4.0496,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782894042898053e-05,
+      "loss": 4.0457,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782055448147001e-05,
+      "loss": 4.0556,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781216853395948e-05,
+      "loss": 4.0671,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780378258644896e-05,
+      "loss": 4.0595,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779541301774218e-05,
+      "loss": 4.059,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778702707023166e-05,
+      "loss": 4.0569,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777864112272114e-05,
+      "loss": 4.0696,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770271554014354e-05,
+      "loss": 4.0453,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761885606503834e-05,
+      "loss": 4.0461,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775349965899331e-05,
+      "loss": 4.0516,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774511371148279e-05,
+      "loss": 4.0536,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773672776397227e-05,
+      "loss": 4.0355,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772835819526548e-05,
+      "loss": 4.0377,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719972247754956e-05,
+      "loss": 4.0466,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7711586300244436e-05,
+      "loss": 4.0431,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703200352733916e-05,
+      "loss": 4.0457,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694814405223396e-05,
+      "loss": 4.0517,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686428457712876e-05,
+      "loss": 4.0346,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767805888900609e-05,
+      "loss": 4.0403,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766967294149557e-05,
+      "loss": 4.0327,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766128699398505e-05,
+      "loss": 4.0509,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765290104647453e-05,
+      "loss": 4.0471,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764451509896401e-05,
+      "loss": 4.0335,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763612915145349e-05,
+      "loss": 4.0371,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76277595827467e-05,
+      "loss": 4.0415,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761937363523618e-05,
+      "loss": 4.0454,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761098768772566e-05,
+      "loss": 4.0419,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760260174021514e-05,
+      "loss": 4.0368,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759421579270462e-05,
+      "loss": 4.0362,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75858298451941e-05,
+      "loss": 4.0362,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757744389768358e-05,
+      "loss": 4.0386,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756905795017306e-05,
+      "loss": 4.0246,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560688381466276e-05,
+      "loss": 4.0328,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552302433955756e-05,
+      "loss": 4.0266,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543916486445236e-05,
+      "loss": 4.0139,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535530538934716e-05,
+      "loss": 4.0427,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527144591424196e-05,
+      "loss": 4.0259,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518775022717405e-05,
+      "loss": 4.0279,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510389075206885e-05,
+      "loss": 4.0295,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7502003127696365e-05,
+      "loss": 4.0153,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.030919075012207,
+      "eval_runtime": 302.8768,
+      "eval_samples_per_second": 1259.889,
+      "eval_steps_per_second": 39.372,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493617180185845e-05,
+      "loss": 4.0194,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748523123267532e-05,
+      "loss": 4.0104,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74768452851648e-05,
+      "loss": 4.036,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746845933765428e-05,
+      "loss": 4.0148,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460073390143764e-05,
+      "loss": 4.0322,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451687442633244e-05,
+      "loss": 4.0109,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443301495122724e-05,
+      "loss": 4.0104,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434915547612204e-05,
+      "loss": 4.0034,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426545978905413e-05,
+      "loss": 4.0172,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741816003139489e-05,
+      "loss": 4.0136,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740977408388437e-05,
+      "loss": 4.0197,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740138813637385e-05,
+      "loss": 4.0226,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739301856766706e-05,
+      "loss": 4.0066,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738463262015654e-05,
+      "loss": 4.0061,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737624667264602e-05,
+      "loss": 4.0005,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73678607251355e-05,
+      "loss": 3.9984,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735947477762498e-05,
+      "loss": 4.0044,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735108883011446e-05,
+      "loss": 3.9983,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734270288260394e-05,
+      "loss": 3.9967,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733433331389716e-05,
+      "loss": 4.0162,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732594736638664e-05,
+      "loss": 3.9989,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731756141887612e-05,
+      "loss": 3.9995,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73091754713656e-05,
+      "loss": 4.0017,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730080590265881e-05,
+      "loss": 4.0059,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729241995514829e-05,
+      "loss": 3.9932,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728403400763777e-05,
+      "loss": 3.9959,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727564806012725e-05,
+      "loss": 3.9934,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267278491420456e-05,
+      "loss": 3.9928,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258892543909936e-05,
+      "loss": 3.9838,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250506596399416e-05,
+      "loss": 3.9941,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7242120648888896e-05,
+      "loss": 3.9932,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723375108018211e-05,
+      "loss": 3.9914,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722536513267159e-05,
+      "loss": 3.9973,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721697918516107e-05,
+      "loss": 3.9837,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720859323765055e-05,
+      "loss": 3.986,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720020729014003e-05,
+      "loss": 3.9893,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191821342629505e-05,
+      "loss": 3.9753,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183435395118985e-05,
+      "loss": 3.9795,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175049447608465e-05,
+      "loss": 3.9705,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716667987890168e-05,
+      "loss": 3.9769,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158293931391154e-05,
+      "loss": 3.985,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149907983880634e-05,
+      "loss": 3.9894,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714152203637012e-05,
+      "loss": 3.9825,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713315246766333e-05,
+      "loss": 3.9768,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712476652015281e-05,
+      "loss": 3.9906,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711638057264229e-05,
+      "loss": 3.9777,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710799462513177e-05,
+      "loss": 3.9767,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709960867762125e-05,
+      "loss": 3.975,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709122273011073e-05,
+      "loss": 3.9609,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708285316140394e-05,
+      "loss": 3.9784,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707446721389342e-05,
+      "loss": 3.9724,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70660812663829e-05,
+      "loss": 3.9737,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705769531887238e-05,
+      "loss": 3.9679,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704930937136186e-05,
+      "loss": 3.9724,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704092342385134e-05,
+      "loss": 3.9554,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032553855144554e-05,
+      "loss": 3.9807,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024167907634034e-05,
+      "loss": 3.9599,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015781960123514e-05,
+      "loss": 3.9814,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007396012612994e-05,
+      "loss": 3.9667,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6999010065102474e-05,
+      "loss": 3.962,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699064049639568e-05,
+      "loss": 3.9567,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698225454888516e-05,
+      "loss": 3.9682,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697386860137464e-05,
+      "loss": 3.9478,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696548265386412e-05,
+      "loss": 3.9556,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695711308515733e-05,
+      "loss": 3.9642,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694872713764681e-05,
+      "loss": 3.9607,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694034119013629e-05,
+      "loss": 3.9516,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693195524262577e-05,
+      "loss": 3.948,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692356929511526e-05,
+      "loss": 3.9502,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691519972640847e-05,
+      "loss": 3.9713,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690681377889795e-05,
+      "loss": 3.9599,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689842783138743e-05,
+      "loss": 3.9555,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689004188387691e-05,
+      "loss": 3.9649,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688165593636639e-05,
+      "loss": 3.961,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687326998885587e-05,
+      "loss": 3.958,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686488404134534e-05,
+      "loss": 3.9424,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856514472638556e-05,
+      "loss": 3.9616,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848128525128036e-05,
+      "loss": 3.9557,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683974257761751e-05,
+      "loss": 3.9454,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831356630106996e-05,
+      "loss": 3.961,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822987061400205e-05,
+      "loss": 3.9528,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681460111388969e-05,
+      "loss": 3.9591,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806215166379165e-05,
+      "loss": 3.9558,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6797829218868645e-05,
+      "loss": 3.94,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678945965016186e-05,
+      "loss": 3.9494,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678107370265134e-05,
+      "loss": 3.946,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772687755140814e-05,
+      "loss": 3.9523,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764301807630294e-05,
+      "loss": 3.9508,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755915860119774e-05,
+      "loss": 3.947,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674754629141298e-05,
+      "loss": 3.9516,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673916034390246e-05,
+      "loss": 3.9466,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673077439639194e-05,
+      "loss": 3.9471,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672238844888143e-05,
+      "loss": 3.9442,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671400250137091e-05,
+      "loss": 3.9456,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670563293266412e-05,
+      "loss": 3.954,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66972469851536e-05,
+      "loss": 3.9446,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668886103764308e-05,
+      "loss": 3.9408,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668047509013256e-05,
+      "loss": 3.946,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667210552142577e-05,
+      "loss": 3.9466,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666371957391525e-05,
+      "loss": 3.955,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665533362640473e-05,
+      "loss": 3.9415,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664694767889421e-05,
+      "loss": 3.9363,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663857811018742e-05,
+      "loss": 3.9403,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66301921626769e-05,
+      "loss": 3.9404,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621806215166383e-05,
+      "loss": 3.9182,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613420267655863e-05,
+      "loss": 3.9351,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660505069894907e-05,
+      "loss": 3.9398,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659666475143855e-05,
+      "loss": 3.9325,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658827880392803e-05,
+      "loss": 3.9314,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657989285641751e-05,
+      "loss": 3.927,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657150690890699e-05,
+      "loss": 3.9292,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65631373402002e-05,
+      "loss": 3.9496,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655475139268968e-05,
+      "loss": 3.9352,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654636544517916e-05,
+      "loss": 3.9424,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653797949766864e-05,
+      "loss": 3.9358,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652960992896185e-05,
+      "loss": 3.9475,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652122398145134e-05,
+      "loss": 3.9289,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651283803394082e-05,
+      "loss": 3.9288,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6504468465234026e-05,
+      "loss": 3.9297,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6496082517723506e-05,
+      "loss": 3.9367,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487696570212986e-05,
+      "loss": 3.9193,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479310622702466e-05,
+      "loss": 3.9214,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470924675191946e-05,
+      "loss": 3.9251,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462555106485155e-05,
+      "loss": 3.9295,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454169158974635e-05,
+      "loss": 3.9307,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445783211464115e-05,
+      "loss": 3.9375,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437397263953595e-05,
+      "loss": 3.9183,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429027695246804e-05,
+      "loss": 3.9237,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642064174773629e-05,
+      "loss": 3.9183,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641225580022577e-05,
+      "loss": 3.9372,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640386985271525e-05,
+      "loss": 3.9361,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639550028400846e-05,
+      "loss": 3.923,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638711433649794e-05,
+      "loss": 3.9195,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637872838898742e-05,
+      "loss": 3.9295,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63703424414769e-05,
+      "loss": 3.9306,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636197287277011e-05,
+      "loss": 3.9315,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635358692525959e-05,
+      "loss": 3.9229,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634520097774907e-05,
+      "loss": 3.9254,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633681503023855e-05,
+      "loss": 3.9259,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632844546153176e-05,
+      "loss": 3.9314,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6320059514021244e-05,
+      "loss": 3.9166,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311673566510724e-05,
+      "loss": 3.9202,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303287619000204e-05,
+      "loss": 3.9183,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294918050293413e-05,
+      "loss": 3.9001,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286532102782893e-05,
+      "loss": 3.9326,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6278146155272373e-05,
+      "loss": 3.9187,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626976020776185e-05,
+      "loss": 3.9185,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626139063905506e-05,
+      "loss": 3.9236,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625300469154454e-05,
+      "loss": 3.9071,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.955702543258667,
+      "eval_runtime": 332.6666,
+      "eval_samples_per_second": 1147.067,
+      "eval_steps_per_second": 35.847,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624461874403402e-05,
+      "loss": 3.9132,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.62362327965235e-05,
+      "loss": 3.9036,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622786322781671e-05,
+      "loss": 3.9266,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.62194772803062e-05,
+      "loss": 3.9089,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621109133279568e-05,
+      "loss": 3.9336,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620270538528516e-05,
+      "loss": 3.8994,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619433581657837e-05,
+      "loss": 3.9112,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618594986906785e-05,
+      "loss": 3.8926,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617756392155733e-05,
+      "loss": 3.9158,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616917797404681e-05,
+      "loss": 3.9105,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616079202653629e-05,
+      "loss": 3.9118,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152422457829496e-05,
+      "loss": 3.9244,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6144052889122705e-05,
+      "loss": 3.8999,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135666941612185e-05,
+      "loss": 3.904,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127280994101665e-05,
+      "loss": 3.8977,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611889504659115e-05,
+      "loss": 3.898,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611050909908063e-05,
+      "loss": 3.9013,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610212315157011e-05,
+      "loss": 3.8985,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609373720405959e-05,
+      "loss": 3.8976,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085351256549065e-05,
+      "loss": 3.9142,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607698168784228e-05,
+      "loss": 3.9004,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606859574033176e-05,
+      "loss": 3.8995,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606020979282124e-05,
+      "loss": 3.9042,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605184022411445e-05,
+      "loss": 3.9072,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604345427660393e-05,
+      "loss": 3.8939,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603506832909341e-05,
+      "loss": 3.8984,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602668238158289e-05,
+      "loss": 3.8964,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60183128128761e-05,
+      "loss": 3.8911,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009926865365585e-05,
+      "loss": 3.8893,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001540917855065e-05,
+      "loss": 3.8919,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599315497034454e-05,
+      "loss": 3.9017,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5984785401637754e-05,
+      "loss": 3.8929,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976399454127234e-05,
+      "loss": 3.905,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5968013506616714e-05,
+      "loss": 3.8904,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595962755910619e-05,
+      "loss": 3.8908,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595124161159567e-05,
+      "loss": 3.8915,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594285566408515e-05,
+      "loss": 3.8846,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593446971657463e-05,
+      "loss": 3.8822,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592608376906411e-05,
+      "loss": 3.8787,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591771420035732e-05,
+      "loss": 3.882,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59093282528468e-05,
+      "loss": 3.8901,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590094230533628e-05,
+      "loss": 3.8923,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589257273662949e-05,
+      "loss": 3.8899,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588418678911897e-05,
+      "loss": 3.8854,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587580084160845e-05,
+      "loss": 3.8984,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586741489409793e-05,
+      "loss": 3.8848,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585904532539114e-05,
+      "loss": 3.8845,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585065937788062e-05,
+      "loss": 3.8807,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58422734303701e-05,
+      "loss": 3.8698,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583388748285958e-05,
+      "loss": 3.891,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582551791415279e-05,
+      "loss": 3.8788,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581713196664228e-05,
+      "loss": 3.8822,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580874601913176e-05,
+      "loss": 3.8755,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580036007162124e-05,
+      "loss": 3.8864,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791990502914446e-05,
+      "loss": 3.862,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783604555403926e-05,
+      "loss": 3.8892,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5775218607893406e-05,
+      "loss": 3.8738,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766832660382886e-05,
+      "loss": 3.8861,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758463091676095e-05,
+      "loss": 3.8822,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5750077144165575e-05,
+      "loss": 3.8716,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741691196655055e-05,
+      "loss": 3.8653,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733305249144535e-05,
+      "loss": 3.8857,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724935680437744e-05,
+      "loss": 3.8626,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571654973292723e-05,
+      "loss": 3.8617,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570816378541671e-05,
+      "loss": 3.8799,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569977783790619e-05,
+      "loss": 3.8746,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56914082691994e-05,
+      "loss": 3.8611,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568302232168888e-05,
+      "loss": 3.8636,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567463637417836e-05,
+      "loss": 3.8636,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566625042666784e-05,
+      "loss": 3.8826,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565788085796105e-05,
+      "loss": 3.8732,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564949491045053e-05,
+      "loss": 3.8679,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564110896294001e-05,
+      "loss": 3.879,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563272301542949e-05,
+      "loss": 3.8778,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56243534467227e-05,
+      "loss": 3.8755,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615967499212184e-05,
+      "loss": 3.8567,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5607581551701664e-05,
+      "loss": 3.8742,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5599195604191144e-05,
+      "loss": 3.869,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559082603548435e-05,
+      "loss": 3.8618,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558244008797383e-05,
+      "loss": 3.881,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557405414046331e-05,
+      "loss": 3.8672,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556566819295279e-05,
+      "loss": 3.8755,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557298624246e-05,
+      "loss": 3.8708,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554891267673548e-05,
+      "loss": 3.8586,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554052672922496e-05,
+      "loss": 3.866,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553214078171444e-05,
+      "loss": 3.8629,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552377121300765e-05,
+      "loss": 3.8672,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551538526549714e-05,
+      "loss": 3.8692,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550699931798662e-05,
+      "loss": 3.8605,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54986133704761e-05,
+      "loss": 3.8701,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549024380176931e-05,
+      "loss": 3.8701,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548185785425879e-05,
+      "loss": 3.8626,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547347190674827e-05,
+      "loss": 3.8635,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546508595923775e-05,
+      "loss": 3.8626,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5456716390530956e-05,
+      "loss": 3.8761,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5448330443020436e-05,
+      "loss": 3.8614,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5439944495509916e-05,
+      "loss": 3.8608,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431558547999396e-05,
+      "loss": 3.8647,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5423188979292605e-05,
+      "loss": 3.8627,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541480303178209e-05,
+      "loss": 3.8748,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540641708427157e-05,
+      "loss": 3.8612,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539803113676105e-05,
+      "loss": 3.86,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538966156805426e-05,
+      "loss": 3.8575,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538127562054374e-05,
+      "loss": 3.8641,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537288967303322e-05,
+      "loss": 3.8395,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53645037255227e-05,
+      "loss": 3.852,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535613415681591e-05,
+      "loss": 3.8634,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534774820930539e-05,
+      "loss": 3.8548,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533936226179487e-05,
+      "loss": 3.8488,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533097631428435e-05,
+      "loss": 3.8493,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532260674557756e-05,
+      "loss": 3.8526,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314220798067045e-05,
+      "loss": 3.8746,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305834850556525e-05,
+      "loss": 3.8535,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297448903046005e-05,
+      "loss": 3.8626,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5289079334339214e-05,
+      "loss": 3.8577,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5280693386828694e-05,
+      "loss": 3.8705,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5272307439318174e-05,
+      "loss": 3.8551,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5263921491807654e-05,
+      "loss": 3.848,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525555192310086e-05,
+      "loss": 3.8573,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524716597559034e-05,
+      "loss": 3.8591,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523878002807982e-05,
+      "loss": 3.8424,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52303940805693e-05,
+      "loss": 3.8437,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522202451186251e-05,
+      "loss": 3.8514,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213638564352e-05,
+      "loss": 3.8528,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520525261684148e-05,
+      "loss": 3.8496,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519686666933096e-05,
+      "loss": 3.8649,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518849710062417e-05,
+      "loss": 3.8406,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518011115311365e-05,
+      "loss": 3.8472,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517172520560313e-05,
+      "loss": 3.8414,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516333925809261e-05,
+      "loss": 3.8633,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515496968938582e-05,
+      "loss": 3.8621,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51465837418753e-05,
+      "loss": 3.8497,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513819779436478e-05,
+      "loss": 3.8421,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512981184685426e-05,
+      "loss": 3.855,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121442278147466e-05,
+      "loss": 3.8557,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5113056330636946e-05,
+      "loss": 3.8544,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510467038312643e-05,
+      "loss": 3.8516,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509628443561591e-05,
+      "loss": 3.8483,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508791486690912e-05,
+      "loss": 3.8531,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50795289193986e-05,
+      "loss": 3.861,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507114297188808e-05,
+      "loss": 3.8418,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506275702437756e-05,
+      "loss": 3.844,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505438745567077e-05,
+      "loss": 3.8443,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504600150816025e-05,
+      "loss": 3.8338,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503761556064973e-05,
+      "loss": 3.8571,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502922961313921e-05,
+      "loss": 3.8462,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502086004443242e-05,
+      "loss": 3.8479,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50124740969219e-05,
+      "loss": 3.8463,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5004088149411386e-05,
+      "loss": 3.8411,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9141323566436768,
+      "eval_runtime": 322.8123,
+      "eval_samples_per_second": 1182.083,
+      "eval_steps_per_second": 36.941,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499570220190086e-05,
+      "loss": 3.8416,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498731625439034e-05,
+      "loss": 3.8319,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497893030687982e-05,
+      "loss": 3.8525,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49705443593693e-05,
+      "loss": 3.8405,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496215841185878e-05,
+      "loss": 3.8604,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495377246434826e-05,
+      "loss": 3.8362,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494538651683774e-05,
+      "loss": 3.8373,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493700056932722e-05,
+      "loss": 3.8235,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492863100062043e-05,
+      "loss": 3.8399,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492024505310991e-05,
+      "loss": 3.844,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491185910559939e-05,
+      "loss": 3.8403,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490347315808887e-05,
+      "loss": 3.8492,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489511996818581e-05,
+      "loss": 3.8348,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488673402067529e-05,
+      "loss": 3.8342,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487834807316477e-05,
+      "loss": 3.8279,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486996212565425e-05,
+      "loss": 3.825,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486157617814373e-05,
+      "loss": 3.8331,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485319023063321e-05,
+      "loss": 3.8316,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484480428312269e-05,
+      "loss": 3.8267,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483641833561217e-05,
+      "loss": 3.8432,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482804876690538e-05,
+      "loss": 3.8336,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481966281939486e-05,
+      "loss": 3.8315,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481127687188434e-05,
+      "loss": 3.8343,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480289092437382e-05,
+      "loss": 3.8394,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479452135566704e-05,
+      "loss": 3.8223,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478613540815652e-05,
+      "loss": 3.837,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4777749460646e-05,
+      "loss": 3.8226,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4769379891939207e-05,
+      "loss": 3.8259,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4760993944428687e-05,
+      "loss": 3.8221,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752607996918166e-05,
+      "loss": 3.8248,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4744222049407646e-05,
+      "loss": 3.8333,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4735852480700856e-05,
+      "loss": 3.83,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4727466533190335e-05,
+      "loss": 3.8343,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4719080585679815e-05,
+      "loss": 3.8254,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710694638169295e-05,
+      "loss": 3.8196,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4702308690658775e-05,
+      "loss": 3.8266,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469392274314826e-05,
+      "loss": 3.8206,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468553679563774e-05,
+      "loss": 3.8137,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467715084812722e-05,
+      "loss": 3.8144,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466878127942043e-05,
+      "loss": 3.8156,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466039533190991e-05,
+      "loss": 3.8241,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465200938439939e-05,
+      "loss": 3.8296,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46436398156926e-05,
+      "loss": 3.824,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463525386818208e-05,
+      "loss": 3.8169,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462686792067156e-05,
+      "loss": 3.8319,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461848197316104e-05,
+      "loss": 3.824,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461011240445425e-05,
+      "loss": 3.8192,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460172645694373e-05,
+      "loss": 3.8153,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4593340509433216e-05,
+      "loss": 3.8031,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584954561922696e-05,
+      "loss": 3.8277,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576584993215905e-05,
+      "loss": 3.8143,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4568199045705385e-05,
+      "loss": 3.819,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559813098194865e-05,
+      "loss": 3.8164,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551427150684345e-05,
+      "loss": 3.8118,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4543057581977554e-05,
+      "loss": 3.8038,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534671634467034e-05,
+      "loss": 3.8188,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4526285686956514e-05,
+      "loss": 3.8141,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517899739445994e-05,
+      "loss": 3.8256,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45095301707392e-05,
+      "loss": 3.8174,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450114422322868e-05,
+      "loss": 3.8099,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449275827571816e-05,
+      "loss": 3.802,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448437232820764e-05,
+      "loss": 3.82,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447600275950086e-05,
+      "loss": 3.8024,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446761681199034e-05,
+      "loss": 3.7986,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445923086447982e-05,
+      "loss": 3.8142,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445084491696929e-05,
+      "loss": 3.8154,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444247534826251e-05,
+      "loss": 3.798,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443408940075199e-05,
+      "loss": 3.8004,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442570345324147e-05,
+      "loss": 3.7985,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441731750573094e-05,
+      "loss": 3.8177,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4408947937024156e-05,
+      "loss": 3.813,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400561989513636e-05,
+      "loss": 3.8089,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4392176042003116e-05,
+      "loss": 3.8165,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4383790094492596e-05,
+      "loss": 3.8129,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437542052578581e-05,
+      "loss": 3.8182,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436703457827529e-05,
+      "loss": 3.7994,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4358648630764765e-05,
+      "loss": 3.8094,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4350262683254245e-05,
+      "loss": 3.8088,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434189311454746e-05,
+      "loss": 3.7979,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333507167036934e-05,
+      "loss": 3.8221,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4325121219526414e-05,
+      "loss": 3.807,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316735272015894e-05,
+      "loss": 3.8136,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430836570330911e-05,
+      "loss": 3.8163,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429997975579859e-05,
+      "loss": 3.7993,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429159380828807e-05,
+      "loss": 3.8028,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428320786077755e-05,
+      "loss": 3.8023,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274838292070766e-05,
+      "loss": 3.8047,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426645234456024e-05,
+      "loss": 3.8073,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425806639704972e-05,
+      "loss": 3.8042,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42496804495392e-05,
+      "loss": 3.8089,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424131088083241e-05,
+      "loss": 3.8093,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423292493332189e-05,
+      "loss": 3.8004,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224555364615104e-05,
+      "loss": 3.804,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216169417104584e-05,
+      "loss": 3.8028,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207783469594064e-05,
+      "loss": 3.8162,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4199397522083544e-05,
+      "loss": 3.8058,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4191011574573024e-05,
+      "loss": 3.7998,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182625627062504e-05,
+      "loss": 3.8058,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4174239679551984e-05,
+      "loss": 3.8037,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165853732041464e-05,
+      "loss": 3.8171,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415748416333467e-05,
+      "loss": 3.8041,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414909821582415e-05,
+      "loss": 3.8002,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414071226831363e-05,
+      "loss": 3.7982,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413232632080311e-05,
+      "loss": 3.8083,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412395675209632e-05,
+      "loss": 3.7827,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41155708045858e-05,
+      "loss": 3.7942,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410718485707528e-05,
+      "loss": 3.8034,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409879890956476e-05,
+      "loss": 3.8017,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409042934085798e-05,
+      "loss": 3.7904,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4082059772151186e-05,
+      "loss": 3.7902,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4073673824640666e-05,
+      "loss": 3.7966,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4065287877130146e-05,
+      "loss": 3.813,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056901929619626e-05,
+      "loss": 3.8012,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4048515982109106e-05,
+      "loss": 3.8017,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4040130034598586e-05,
+      "loss": 3.8018,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031744087088066e-05,
+      "loss": 3.8087,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4023358139577546e-05,
+      "loss": 3.8021,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4014988570870755e-05,
+      "loss": 3.7891,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4006602623360235e-05,
+      "loss": 3.799,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3998216675849715e-05,
+      "loss": 3.8028,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39898307283392e-05,
+      "loss": 3.7849,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398146115963241e-05,
+      "loss": 3.7897,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397307521212189e-05,
+      "loss": 3.7914,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396468926461137e-05,
+      "loss": 3.7976,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395630331710085e-05,
+      "loss": 3.7941,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394793374839406e-05,
+      "loss": 3.8068,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393954780088354e-05,
+      "loss": 3.7833,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393116185337302e-05,
+      "loss": 3.7931,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39227759058625e-05,
+      "loss": 3.7862,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391440633715571e-05,
+      "loss": 3.8095,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390602038964519e-05,
+      "loss": 3.8,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389763444213467e-05,
+      "loss": 3.8004,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3889248494624155e-05,
+      "loss": 3.7832,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880878925917365e-05,
+      "loss": 3.7999,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872492978406845e-05,
+      "loss": 3.8044,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3864107030896325e-05,
+      "loss": 3.7964,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855721083385804e-05,
+      "loss": 3.7922,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3847351514679014e-05,
+      "loss": 3.7994,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838965567168494e-05,
+      "loss": 3.7953,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830579619657973e-05,
+      "loss": 3.8065,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3822193672147453e-05,
+      "loss": 3.7827,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381382410344066e-05,
+      "loss": 3.7917,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380543815593014e-05,
+      "loss": 3.788,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379705220841962e-05,
+      "loss": 3.781,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378866626090911e-05,
+      "loss": 3.7993,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378028031339859e-05,
+      "loss": 3.7927,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377189436588807e-05,
+      "loss": 3.7947,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376350841837754e-05,
+      "loss": 3.7903,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375512247086702e-05,
+      "loss": 3.7899,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8890888690948486,
+      "eval_runtime": 303.1724,
+      "eval_samples_per_second": 1258.66,
+      "eval_steps_per_second": 39.334,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374675290216024e-05,
+      "loss": 3.7866,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373836695464972e-05,
+      "loss": 3.7761,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372998100713919e-05,
+      "loss": 3.7992,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372159505962867e-05,
+      "loss": 3.7879,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371322549092189e-05,
+      "loss": 3.8029,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370483954341137e-05,
+      "loss": 3.7848,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369645359590085e-05,
+      "loss": 3.7779,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368808402719406e-05,
+      "loss": 3.773,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367969807968354e-05,
+      "loss": 3.7845,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3671312132173016e-05,
+      "loss": 3.7931,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3662926184662496e-05,
+      "loss": 3.7869,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365455661595571e-05,
+      "loss": 3.7966,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364618704724892e-05,
+      "loss": 3.7816,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36378010997384e-05,
+      "loss": 3.7852,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362941515222788e-05,
+      "loss": 3.776,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362102920471736e-05,
+      "loss": 3.7702,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361264325720684e-05,
+      "loss": 3.7804,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3604257309696314e-05,
+      "loss": 3.7799,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35958713621858e-05,
+      "loss": 3.7745,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358748541467528e-05,
+      "loss": 3.7898,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357911584596849e-05,
+      "loss": 3.7855,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357072989845797e-05,
+      "loss": 3.7806,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356234395094745e-05,
+      "loss": 3.7796,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355395800343693e-05,
+      "loss": 3.7852,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354558843473014e-05,
+      "loss": 3.7715,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353720248721962e-05,
+      "loss": 3.7856,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35288165397091e-05,
+      "loss": 3.7727,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3520446971002314e-05,
+      "loss": 3.7747,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351206102349179e-05,
+      "loss": 3.7747,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350367507598127e-05,
+      "loss": 3.7715,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349528912847075e-05,
+      "loss": 3.7824,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486919559763963e-05,
+      "loss": 3.7814,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347853361225344e-05,
+      "loss": 3.7814,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347014766474292e-05,
+      "loss": 3.7748,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34617617172324e-05,
+      "loss": 3.7683,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345337576972188e-05,
+      "loss": 3.7815,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344498982221136e-05,
+      "loss": 3.7711,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343660387470084e-05,
+      "loss": 3.7599,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342821792719032e-05,
+      "loss": 3.7668,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341984835848353e-05,
+      "loss": 3.764,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341146241097301e-05,
+      "loss": 3.7714,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340307646346249e-05,
+      "loss": 3.7809,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339469051595197e-05,
+      "loss": 3.7719,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338632094724519e-05,
+      "loss": 3.7718,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337793499973467e-05,
+      "loss": 3.7793,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336954905222415e-05,
+      "loss": 3.7716,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336117948351736e-05,
+      "loss": 3.7726,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335279353600684e-05,
+      "loss": 3.7661,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334440758849632e-05,
+      "loss": 3.7497,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33360216409858e-05,
+      "loss": 3.7797,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3327652072279006e-05,
+      "loss": 3.7643,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3319266124768486e-05,
+      "loss": 3.7718,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3310880177257966e-05,
+      "loss": 3.7652,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3302494229747446e-05,
+      "loss": 3.7622,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3294124661040655e-05,
+      "loss": 3.7588,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328573871353014e-05,
+      "loss": 3.7662,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327735276601962e-05,
+      "loss": 3.7677,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32689668185091e-05,
+      "loss": 3.7754,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326059724980231e-05,
+      "loss": 3.7666,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325221130229179e-05,
+      "loss": 3.7657,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324382535478127e-05,
+      "loss": 3.75,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323543940727075e-05,
+      "loss": 3.7754,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322706983856396e-05,
+      "loss": 3.754,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321868389105344e-05,
+      "loss": 3.751,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321029794354292e-05,
+      "loss": 3.7642,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32019119960324e-05,
+      "loss": 3.7679,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319354242732561e-05,
+      "loss": 3.7474,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3185156479815095e-05,
+      "loss": 3.7572,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176770532304575e-05,
+      "loss": 3.7473,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3168384584794055e-05,
+      "loss": 3.7724,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3160015016087264e-05,
+      "loss": 3.7629,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151629068576744e-05,
+      "loss": 3.7628,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314325949986995e-05,
+      "loss": 3.768,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313487355235943e-05,
+      "loss": 3.7613,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312648760484891e-05,
+      "loss": 3.7751,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311810165733839e-05,
+      "loss": 3.7493,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310971570982787e-05,
+      "loss": 3.7659,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310132976231735e-05,
+      "loss": 3.7597,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309294381480683e-05,
+      "loss": 3.7512,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308455786729631e-05,
+      "loss": 3.7686,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307618829858953e-05,
+      "loss": 3.7658,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306780235107901e-05,
+      "loss": 3.7642,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305941640356849e-05,
+      "loss": 3.7702,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305103045605797e-05,
+      "loss": 3.7516,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304266088735118e-05,
+      "loss": 3.7556,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303427493984066e-05,
+      "loss": 3.7528,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302588899233014e-05,
+      "loss": 3.7582,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301750304481962e-05,
+      "loss": 3.7593,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300913347611283e-05,
+      "loss": 3.759,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300074752860231e-05,
+      "loss": 3.7581,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299236158109179e-05,
+      "loss": 3.7653,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.298397563358127e-05,
+      "loss": 3.7533,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297562244367821e-05,
+      "loss": 3.757,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296723649616769e-05,
+      "loss": 3.7594,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295885054865717e-05,
+      "loss": 3.7688,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295046460114665e-05,
+      "loss": 3.763,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294207865363613e-05,
+      "loss": 3.7499,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293369270612561e-05,
+      "loss": 3.7605,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292530675861509e-05,
+      "loss": 3.7593,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2916920811104565e-05,
+      "loss": 3.7652,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290855124239778e-05,
+      "loss": 3.7597,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290018167369099e-05,
+      "loss": 3.7547,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289179572618047e-05,
+      "loss": 3.7472,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2883409778669956e-05,
+      "loss": 3.7668,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2875023831159436e-05,
+      "loss": 3.7362,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2866637883648916e-05,
+      "loss": 3.7531,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2858251936138396e-05,
+      "loss": 3.7554,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284986598862787e-05,
+      "loss": 3.7519,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2841496419921085e-05,
+      "loss": 3.7469,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2833110472410565e-05,
+      "loss": 3.7462,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282472452490004e-05,
+      "loss": 3.7515,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281633857738952e-05,
+      "loss": 3.7661,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807952629879e-05,
+      "loss": 3.7573,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279956668236848e-05,
+      "loss": 3.7531,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2791197113661694e-05,
+      "loss": 3.7595,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782811166151174e-05,
+      "loss": 3.763,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2774425218640654e-05,
+      "loss": 3.7578,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2766039271130134e-05,
+      "loss": 3.7439,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757653323619614e-05,
+      "loss": 3.7554,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2749267376109094e-05,
+      "loss": 3.7551,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2740881428598574e-05,
+      "loss": 3.7399,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2732495481088054e-05,
+      "loss": 3.7441,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272412591238126e-05,
+      "loss": 3.7505,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271573996487074e-05,
+      "loss": 3.7498,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270735401736022e-05,
+      "loss": 3.7507,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26989680698497e-05,
+      "loss": 3.76,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269059850114291e-05,
+      "loss": 3.7416,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268221255363239e-05,
+      "loss": 3.746,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267382660612187e-05,
+      "loss": 3.7462,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266544065861136e-05,
+      "loss": 3.7626,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265707108990457e-05,
+      "loss": 3.7538,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264868514239405e-05,
+      "loss": 3.7562,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264029919488353e-05,
+      "loss": 3.7426,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263191324737301e-05,
+      "loss": 3.7558,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2623543678666216e-05,
+      "loss": 3.759,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2615157731155696e-05,
+      "loss": 3.7516,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2606771783645176e-05,
+      "loss": 3.749,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2598385836134656e-05,
+      "loss": 3.7546,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2590016267427865e-05,
+      "loss": 3.75,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2581630319917345e-05,
+      "loss": 3.7651,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2573244372406825e-05,
+      "loss": 3.7386,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256487480370004e-05,
+      "loss": 3.7528,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255648885618952e-05,
+      "loss": 3.7392,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2548102908679e-05,
+      "loss": 3.7408,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253971696116848e-05,
+      "loss": 3.7571,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253133101365796e-05,
+      "loss": 3.7436,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252294506614744e-05,
+      "loss": 3.7594,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251455911863692e-05,
+      "loss": 3.7398,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25061731711264e-05,
+      "loss": 3.7523,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8726401329040527,
+      "eval_runtime": 302.2559,
+      "eval_samples_per_second": 1262.477,
+      "eval_steps_per_second": 39.453,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249780360241961e-05,
+      "loss": 3.7567,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248941765490909e-05,
+      "loss": 3.7283,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248103170739857e-05,
+      "loss": 3.7523,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247264575988805e-05,
+      "loss": 3.7484,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246425981237753e-05,
+      "loss": 3.7593,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245587386486701e-05,
+      "loss": 3.7397,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244748791735649e-05,
+      "loss": 3.7372,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243910196984597e-05,
+      "loss": 3.7347,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243073240113918e-05,
+      "loss": 3.7408,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242234645362866e-05,
+      "loss": 3.7481,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241396050611814e-05,
+      "loss": 3.7466,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240557455860762e-05,
+      "loss": 3.7526,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2397221368704564e-05,
+      "loss": 3.7434,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2388835421194044e-05,
+      "loss": 3.7404,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2380449473683524e-05,
+      "loss": 3.7335,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2372063526173004e-05,
+      "loss": 3.7258,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363677578662484e-05,
+      "loss": 3.7406,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2355291631151963e-05,
+      "loss": 3.7341,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346905683641443e-05,
+      "loss": 3.7352,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233851973613092e-05,
+      "loss": 3.7448,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233015016742413e-05,
+      "loss": 3.7468,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232176421991361e-05,
+      "loss": 3.7382,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231337827240309e-05,
+      "loss": 3.7374,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230499232489257e-05,
+      "loss": 3.742,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229662275618578e-05,
+      "loss": 3.7341,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228823680867526e-05,
+      "loss": 3.7397,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227985086116474e-05,
+      "loss": 3.7316,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227146491365422e-05,
+      "loss": 3.7356,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226309534494744e-05,
+      "loss": 3.7336,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225470939743692e-05,
+      "loss": 3.7323,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22463234499264e-05,
+      "loss": 3.7353,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237953881219606e-05,
+      "loss": 3.7409,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2229567933709086e-05,
+      "loss": 3.7386,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2221181986198566e-05,
+      "loss": 3.7367,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2212796038688046e-05,
+      "loss": 3.7273,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2204426469981255e-05,
+      "loss": 3.739,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2196040522470735e-05,
+      "loss": 3.7338,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2187654574960215e-05,
+      "loss": 3.7225,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2179268627449695e-05,
+      "loss": 3.7227,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2170882679939175e-05,
+      "loss": 3.7227,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2162496732428655e-05,
+      "loss": 3.7286,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2154110784918135e-05,
+      "loss": 3.7381,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214574121621135e-05,
+      "loss": 3.7315,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213735526870083e-05,
+      "loss": 3.7352,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212896932119031e-05,
+      "loss": 3.7401,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212058337367979e-05,
+      "loss": 3.7288,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211219742616927e-05,
+      "loss": 3.7319,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210382785746248e-05,
+      "loss": 3.7286,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209544190995196e-05,
+      "loss": 3.7038,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208705596244144e-05,
+      "loss": 3.7417,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207867001493092e-05,
+      "loss": 3.7235,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20702840674204e-05,
+      "loss": 3.737,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206191449871361e-05,
+      "loss": 3.7232,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205352855120309e-05,
+      "loss": 3.7276,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2045142603692575e-05,
+      "loss": 3.7168,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036756656182055e-05,
+      "loss": 3.7258,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2028370708671535e-05,
+      "loss": 3.7275,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201998476116101e-05,
+      "loss": 3.7372,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201159881365049e-05,
+      "loss": 3.7345,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200321286613997e-05,
+      "loss": 3.7206,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1994843297433184e-05,
+      "loss": 3.7099,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198645734992266e-05,
+      "loss": 3.74,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197807140241214e-05,
+      "loss": 3.712,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196968545490162e-05,
+      "loss": 3.708,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196131588619483e-05,
+      "loss": 3.7271,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195292993868431e-05,
+      "loss": 3.7316,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194454399117379e-05,
+      "loss": 3.706,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193615804366327e-05,
+      "loss": 3.7171,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192778847495648e-05,
+      "loss": 3.7122,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191940252744596e-05,
+      "loss": 3.7298,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191101657993544e-05,
+      "loss": 3.7256,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190263063242492e-05,
+      "loss": 3.7251,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189427744252187e-05,
+      "loss": 3.7254,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188589149501135e-05,
+      "loss": 3.7255,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187750554750083e-05,
+      "loss": 3.736,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18691195999903e-05,
+      "loss": 3.7088,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186073365247978e-05,
+      "loss": 3.729,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185234770496927e-05,
+      "loss": 3.7238,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184397813626248e-05,
+      "loss": 3.7125,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1835592188751956e-05,
+      "loss": 3.7279,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1827206241241436e-05,
+      "loss": 3.7289,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818820293730916e-05,
+      "loss": 3.7262,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1810434346220396e-05,
+      "loss": 3.7306,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1802048398709876e-05,
+      "loss": 3.7101,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1793662451199356e-05,
+      "loss": 3.7152,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1785276503688836e-05,
+      "loss": 3.7163,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776906934982045e-05,
+      "loss": 3.7216,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1768520987471525e-05,
+      "loss": 3.7216,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1760135039961005e-05,
+      "loss": 3.7259,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175176547125422e-05,
+      "loss": 3.7114,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17433795237437e-05,
+      "loss": 3.7293,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173499357623318e-05,
+      "loss": 3.7158,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172660762872266e-05,
+      "loss": 3.719,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171822168121214e-05,
+      "loss": 3.722,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170985211250535e-05,
+      "loss": 3.7281,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170146616499483e-05,
+      "loss": 3.7278,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169308021748431e-05,
+      "loss": 3.7114,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168469426997379e-05,
+      "loss": 3.7218,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167630832246327e-05,
+      "loss": 3.7225,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166793875375648e-05,
+      "loss": 3.7248,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165955280624596e-05,
+      "loss": 3.7238,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165116685873544e-05,
+      "loss": 3.7169,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164278091122492e-05,
+      "loss": 3.7149,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1634394963714405e-05,
+      "loss": 3.7258,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1626009016203885e-05,
+      "loss": 3.7004,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1617623068693365e-05,
+      "loss": 3.713,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1609237121182845e-05,
+      "loss": 3.7163,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1600867552476054e-05,
+      "loss": 3.7169,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159249798376926e-05,
+      "loss": 3.7081,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158411203625874e-05,
+      "loss": 3.7121,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157572608874822e-05,
+      "loss": 3.7157,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15673401412377e-05,
+      "loss": 3.7273,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155895419372718e-05,
+      "loss": 3.7216,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155056824621666e-05,
+      "loss": 3.7144,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154219867750987e-05,
+      "loss": 3.7241,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153381272999936e-05,
+      "loss": 3.7272,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152542678248884e-05,
+      "loss": 3.7231,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151704083497832e-05,
+      "loss": 3.706,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150865488746779e-05,
+      "loss": 3.718,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150026893995727e-05,
+      "loss": 3.7178,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149189937125049e-05,
+      "loss": 3.7025,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148351342373997e-05,
+      "loss": 3.7078,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147512747622944e-05,
+      "loss": 3.7133,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146674152871892e-05,
+      "loss": 3.7097,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14583555812084e-05,
+      "loss": 3.7162,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144996963369788e-05,
+      "loss": 3.7234,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144158368618736e-05,
+      "loss": 3.7081,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143319773867684e-05,
+      "loss": 3.7079,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424828169970056e-05,
+      "loss": 3.7135,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1416442222459536e-05,
+      "loss": 3.7234,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1408056274949016e-05,
+      "loss": 3.7168,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1399670327438496e-05,
+      "loss": 3.7235,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1391300758731705e-05,
+      "loss": 3.7049,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1382931190024914e-05,
+      "loss": 3.7203,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1374545242514394e-05,
+      "loss": 3.7225,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1366159295003874e-05,
+      "loss": 3.7143,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357773347493354e-05,
+      "loss": 3.7174,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1349387399982834e-05,
+      "loss": 3.7153,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1341001452472314e-05,
+      "loss": 3.7152,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332615504961794e-05,
+      "loss": 3.7294,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1324229557451274e-05,
+      "loss": 3.7058,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131585998874449e-05,
+      "loss": 3.7144,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130747404123397e-05,
+      "loss": 3.7048,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129908809372345e-05,
+      "loss": 3.7064,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129070214621293e-05,
+      "loss": 3.7161,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128231619870241e-05,
+      "loss": 3.711,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127393025119189e-05,
+      "loss": 3.7203,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126554430368137e-05,
+      "loss": 3.705,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125717473497458e-05,
+      "loss": 3.7189,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8624958992004395,
+      "eval_runtime": 303.3265,
+      "eval_samples_per_second": 1258.021,
+      "eval_steps_per_second": 39.314,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124878878746406e-05,
+      "loss": 3.7061,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124040283995354e-05,
+      "loss": 3.6947,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123201689244302e-05,
+      "loss": 3.7139,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12236309449325e-05,
+      "loss": 3.7115,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1215261376225714e-05,
+      "loss": 3.722,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1206875428715194e-05,
+      "loss": 3.7079,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1198489481204674e-05,
+      "loss": 3.7033,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1190103533694154e-05,
+      "loss": 3.7028,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118171758618363e-05,
+      "loss": 3.7067,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117333163867311e-05,
+      "loss": 3.7103,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116494569116259e-05,
+      "loss": 3.7106,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115655974365207e-05,
+      "loss": 3.7161,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1148190174945276e-05,
+      "loss": 3.7067,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1139804227434756e-05,
+      "loss": 3.7066,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1131418279924236e-05,
+      "loss": 3.7019,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1123032332413716e-05,
+      "loss": 3.6916,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1114646384903196e-05,
+      "loss": 3.7036,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110626043739268e-05,
+      "loss": 3.704,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109789086868589e-05,
+      "loss": 3.7004,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108950492117537e-05,
+      "loss": 3.7021,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108111897366485e-05,
+      "loss": 3.719,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107273302615433e-05,
+      "loss": 3.7041,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106434707864381e-05,
+      "loss": 3.7032,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105596113113329e-05,
+      "loss": 3.704,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10475915624265e-05,
+      "loss": 3.7019,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103920561491598e-05,
+      "loss": 3.7059,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103081966740546e-05,
+      "loss": 3.6997,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102243371989494e-05,
+      "loss": 3.7041,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101404777238442e-05,
+      "loss": 3.696,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10056618248739e-05,
+      "loss": 3.6943,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.099727587736338e-05,
+      "loss": 3.7003,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988906308656597e-05,
+      "loss": 3.7085,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0980520361146076e-05,
+      "loss": 3.7073,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0972134413635556e-05,
+      "loss": 3.7047,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0963748466125036e-05,
+      "loss": 3.6928,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0955378897418245e-05,
+      "loss": 3.6985,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0946992949907725e-05,
+      "loss": 3.6989,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0938607002397205e-05,
+      "loss": 3.6922,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0930221054886685e-05,
+      "loss": 3.6891,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092183510737616e-05,
+      "loss": 3.6879,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091344915986564e-05,
+      "loss": 3.6963,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090506321235512e-05,
+      "loss": 3.701,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0896693643648334e-05,
+      "loss": 3.6973,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0888307696137814e-05,
+      "loss": 3.7035,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0879921748627294e-05,
+      "loss": 3.7015,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0871535801116774e-05,
+      "loss": 3.6973,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0863149853606254e-05,
+      "loss": 3.6994,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0854763906095734e-05,
+      "loss": 3.6942,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0846377958585214e-05,
+      "loss": 3.6726,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0837992011074694e-05,
+      "loss": 3.7067,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08296224423679e-05,
+      "loss": 3.6885,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082125287366111e-05,
+      "loss": 3.7041,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081286692615059e-05,
+      "loss": 3.6939,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080448097864007e-05,
+      "loss": 3.6918,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079609503112955e-05,
+      "loss": 3.6824,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078770908361904e-05,
+      "loss": 3.6891,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077932313610852e-05,
+      "loss": 3.6988,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0770937188598e-05,
+      "loss": 3.6991,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076256761989121e-05,
+      "loss": 3.7007,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075418167238069e-05,
+      "loss": 3.69,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074579572487017e-05,
+      "loss": 3.6773,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073740977735965e-05,
+      "loss": 3.7061,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072902382984913e-05,
+      "loss": 3.6824,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072063788233861e-05,
+      "loss": 3.6724,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071225193482809e-05,
+      "loss": 3.6938,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070386598731757e-05,
+      "loss": 3.6996,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0695496418610777e-05,
+      "loss": 3.6765,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0687110471100257e-05,
+      "loss": 3.6863,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0678724523589736e-05,
+      "loss": 3.6775,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067033857607922e-05,
+      "loss": 3.6957,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066196900737243e-05,
+      "loss": 3.6948,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065358305986191e-05,
+      "loss": 3.6883,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064521349115512e-05,
+      "loss": 3.692,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06368275436446e-05,
+      "loss": 3.6927,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062844159613408e-05,
+      "loss": 3.7028,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062005564862356e-05,
+      "loss": 3.6782,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061166970111304e-05,
+      "loss": 3.6978,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060330013240625e-05,
+      "loss": 3.6859,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059491418489573e-05,
+      "loss": 3.6842,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058652823738521e-05,
+      "loss": 3.695,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057814228987469e-05,
+      "loss": 3.6985,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056975634236418e-05,
+      "loss": 3.6916,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0561386773657386e-05,
+      "loss": 3.701,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0553000826146866e-05,
+      "loss": 3.6774,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0544614878636346e-05,
+      "loss": 3.6822,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053622893112582e-05,
+      "loss": 3.6852,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05278429836153e-05,
+      "loss": 3.6924,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051945703610478e-05,
+      "loss": 3.685,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051107108859426e-05,
+      "loss": 3.6934,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050268514108374e-05,
+      "loss": 3.6812,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049431557237695e-05,
+      "loss": 3.6985,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048592962486643e-05,
+      "loss": 3.688,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0477543677355915e-05,
+      "loss": 3.6849,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0469157729845395e-05,
+      "loss": 3.6896,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0460788161138604e-05,
+      "loss": 3.6966,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0452402213628084e-05,
+      "loss": 3.6954,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0444016266117564e-05,
+      "loss": 3.6826,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0435630318607044e-05,
+      "loss": 3.6872,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0427244371096524e-05,
+      "loss": 3.6906,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041887480238973e-05,
+      "loss": 3.6937,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041048885487921e-05,
+      "loss": 3.6924,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040210290736869e-05,
+      "loss": 3.6832,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039371695985817e-05,
+      "loss": 3.6824,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038534739115138e-05,
+      "loss": 3.6935,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037696144364087e-05,
+      "loss": 3.6675,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036857549613035e-05,
+      "loss": 3.6841,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036018954861983e-05,
+      "loss": 3.6827,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035181997991304e-05,
+      "loss": 3.6883,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034343403240252e-05,
+      "loss": 3.6747,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0335048084892e-05,
+      "loss": 3.683,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032666213738148e-05,
+      "loss": 3.6865,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031827618987096e-05,
+      "loss": 3.6909,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030989024236044e-05,
+      "loss": 3.6886,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030150429484992e-05,
+      "loss": 3.6842,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0293134726143126e-05,
+      "loss": 3.6916,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284748778632606e-05,
+      "loss": 3.6985,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0276362831122086e-05,
+      "loss": 3.6901,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0267976883611566e-05,
+      "loss": 3.6738,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0259590936101046e-05,
+      "loss": 3.6867,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025120498859053e-05,
+      "loss": 3.6885,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024281904108001e-05,
+      "loss": 3.669,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023444947237322e-05,
+      "loss": 3.6775,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02260635248627e-05,
+      "loss": 3.6812,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021767757735218e-05,
+      "loss": 3.6814,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0209291629841655e-05,
+      "loss": 3.6803,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0200905682331135e-05,
+      "loss": 3.6933,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0192519734820615e-05,
+      "loss": 3.6804,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0184133787310095e-05,
+      "loss": 3.6754,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0175764218603304e-05,
+      "loss": 3.6795,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0167378271092784e-05,
+      "loss": 3.6961,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015899232358227e-05,
+      "loss": 3.6833,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0150622754875486e-05,
+      "loss": 3.6939,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014223680736496e-05,
+      "loss": 3.6725,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013385085985444e-05,
+      "loss": 3.6889,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012546491234392e-05,
+      "loss": 3.6921,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011709534363713e-05,
+      "loss": 3.687,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010870939612661e-05,
+      "loss": 3.6842,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010032344861609e-05,
+      "loss": 3.6851,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009193750110557e-05,
+      "loss": 3.6844,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008355155359505e-05,
+      "loss": 3.6997,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007516560608453e-05,
+      "loss": 3.6755,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006679603737774e-05,
+      "loss": 3.6843,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0058410089867224e-05,
+      "loss": 3.6774,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0050024142356704e-05,
+      "loss": 3.6766,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0041638194846184e-05,
+      "loss": 3.6779,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0033252247335664e-05,
+      "loss": 3.6817,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0024866299825144e-05,
+      "loss": 3.6924,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001649673111835e-05,
+      "loss": 3.6749,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000811078360783e-05,
+      "loss": 3.6871,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.855154275894165,
+      "eval_runtime": 302.1496,
+      "eval_samples_per_second": 1262.921,
+      "eval_steps_per_second": 39.467,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999972483609731e-05,
+      "loss": 3.6872,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999133888858679e-05,
+      "loss": 3.665,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998295294107627e-05,
+      "loss": 3.686,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997456699356575e-05,
+      "loss": 3.6832,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996619742485896e-05,
+      "loss": 3.6897,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995781147734844e-05,
+      "loss": 3.6751,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994942552983792e-05,
+      "loss": 3.6764,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994103958232741e-05,
+      "loss": 3.6695,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993265363481689e-05,
+      "loss": 3.675,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99242840661101e-05,
+      "loss": 3.6837,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991589811859958e-05,
+      "loss": 3.6776,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990751217108906e-05,
+      "loss": 3.6858,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989914260238227e-05,
+      "loss": 3.6857,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989075665487175e-05,
+      "loss": 3.6669,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9882370707361227e-05,
+      "loss": 3.669,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9873984759850707e-05,
+      "loss": 3.6677,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865598812340187e-05,
+      "loss": 3.6744,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9857212864829666e-05,
+      "loss": 3.6719,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848826917319146e-05,
+      "loss": 3.669,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9840440969808626e-05,
+      "loss": 3.6763,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983207140110184e-05,
+      "loss": 3.6865,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823685453591315e-05,
+      "loss": 3.6755,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9815299506080795e-05,
+      "loss": 3.6734,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9806913558570275e-05,
+      "loss": 3.6716,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979854398986349e-05,
+      "loss": 3.6737,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9790158042352964e-05,
+      "loss": 3.6767,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781772094842444e-05,
+      "loss": 3.6681,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9773386147331924e-05,
+      "loss": 3.6809,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976501657862514e-05,
+      "loss": 3.6658,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975663063111461e-05,
+      "loss": 3.6629,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974824468360409e-05,
+      "loss": 3.674,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973985873609358e-05,
+      "loss": 3.6734,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973147278858306e-05,
+      "loss": 3.6756,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972310321987627e-05,
+      "loss": 3.6749,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971471727236575e-05,
+      "loss": 3.6637,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970633132485523e-05,
+      "loss": 3.6671,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969794537734471e-05,
+      "loss": 3.6764,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968955942983419e-05,
+      "loss": 3.6592,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96811898611274e-05,
+      "loss": 3.662,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967280391361688e-05,
+      "loss": 3.6601,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966441796610636e-05,
+      "loss": 3.664,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965603201859584e-05,
+      "loss": 3.6703,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964766244988905e-05,
+      "loss": 3.6708,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9639276502378534e-05,
+      "loss": 3.6785,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9630890554868014e-05,
+      "loss": 3.6667,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9622504607357494e-05,
+      "loss": 3.6726,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96141350386507e-05,
+      "loss": 3.6668,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960574909114018e-05,
+      "loss": 3.6679,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959736314362966e-05,
+      "loss": 3.6422,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958897719611914e-05,
+      "loss": 3.677,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958059124860862e-05,
+      "loss": 3.6614,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957222167990183e-05,
+      "loss": 3.6742,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956383573239131e-05,
+      "loss": 3.6653,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955544978488079e-05,
+      "loss": 3.6645,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954706383737027e-05,
+      "loss": 3.6497,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953869426866349e-05,
+      "loss": 3.6614,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953030832115297e-05,
+      "loss": 3.6725,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952192237364245e-05,
+      "loss": 3.6639,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951353642613193e-05,
+      "loss": 3.6726,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9505166857425136e-05,
+      "loss": 3.6668,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9496780909914616e-05,
+      "loss": 3.644,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9488394962404096e-05,
+      "loss": 3.6752,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9480009014893576e-05,
+      "loss": 3.6568,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9471639446186785e-05,
+      "loss": 3.6462,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9463253498676265e-05,
+      "loss": 3.6682,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454867551165745e-05,
+      "loss": 3.6635,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9446481603655225e-05,
+      "loss": 3.648,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9438095656144705e-05,
+      "loss": 3.6569,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942972608743792e-05,
+      "loss": 3.653,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94213401399274e-05,
+      "loss": 3.6643,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941295419241688e-05,
+      "loss": 3.6657,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940456824490636e-05,
+      "loss": 3.6614,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939619867619957e-05,
+      "loss": 3.6665,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938781272868905e-05,
+      "loss": 3.6625,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937942678117853e-05,
+      "loss": 3.676,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937104083366801e-05,
+      "loss": 3.6484,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936267126496122e-05,
+      "loss": 3.6681,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93542853174507e-05,
+      "loss": 3.657,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934589936994018e-05,
+      "loss": 3.6541,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933751342242966e-05,
+      "loss": 3.6671,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9329143853722875e-05,
+      "loss": 3.6687,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9320757906212355e-05,
+      "loss": 3.6641,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9312371958701835e-05,
+      "loss": 3.673,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9303986011191315e-05,
+      "loss": 3.6478,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9295616442484524e-05,
+      "loss": 3.6556,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9287230494974004e-05,
+      "loss": 3.6559,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9278844547463484e-05,
+      "loss": 3.6631,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9270458599952964e-05,
+      "loss": 3.6559,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9262072652442443e-05,
+      "loss": 3.6645,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925370308373565e-05,
+      "loss": 3.6571,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924531713622513e-05,
+      "loss": 3.6667,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923693118871461e-05,
+      "loss": 3.6607,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922854524120409e-05,
+      "loss": 3.6581,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922017567249731e-05,
+      "loss": 3.6634,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921178972498679e-05,
+      "loss": 3.6676,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920340377747627e-05,
+      "loss": 3.6713,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919501782996575e-05,
+      "loss": 3.6556,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918664826125896e-05,
+      "loss": 3.6553,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917826231374844e-05,
+      "loss": 3.6633,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916987636623792e-05,
+      "loss": 3.6694,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916149041872739e-05,
+      "loss": 3.6632,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915310447121687e-05,
+      "loss": 3.6608,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9144734902510086e-05,
+      "loss": 3.658,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9136348954999566e-05,
+      "loss": 3.6642,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9127963007489046e-05,
+      "loss": 3.6415,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9119577059978526e-05,
+      "loss": 3.6591,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911120749127174e-05,
+      "loss": 3.6458,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910282154376122e-05,
+      "loss": 3.6598,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9094435596250695e-05,
+      "loss": 3.6494,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9086049648740175e-05,
+      "loss": 3.6596,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907768008003339e-05,
+      "loss": 3.6567,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9069294132522864e-05,
+      "loss": 3.6642,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9060908185012344e-05,
+      "loss": 3.6602,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9052522237501824e-05,
+      "loss": 3.6594,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9044136289991304e-05,
+      "loss": 3.6643,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903576672128452e-05,
+      "loss": 3.6671,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9027380773774e-05,
+      "loss": 3.6623,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901899482626348e-05,
+      "loss": 3.6517,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901060887875296e-05,
+      "loss": 3.6552,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900223931004617e-05,
+      "loss": 3.6616,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899385336253565e-05,
+      "loss": 3.6484,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898546741502513e-05,
+      "loss": 3.6479,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897708146751461e-05,
+      "loss": 3.6524,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896869552000409e-05,
+      "loss": 3.6547,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89603259512973e-05,
+      "loss": 3.65,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895194000378678e-05,
+      "loss": 3.6685,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894355405627626e-05,
+      "loss": 3.6542,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893516810876574e-05,
+      "loss": 3.6505,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892678216125522e-05,
+      "loss": 3.6502,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891841259254843e-05,
+      "loss": 3.6664,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891002664503791e-05,
+      "loss": 3.659,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890164069752739e-05,
+      "loss": 3.6623,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889325475001687e-05,
+      "loss": 3.6463,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888486880250635e-05,
+      "loss": 3.6633,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887649923379956e-05,
+      "loss": 3.665,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886811328628904e-05,
+      "loss": 3.6599,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885972733877852e-05,
+      "loss": 3.6575,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8851341391268e-05,
+      "loss": 3.6525,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884297182256121e-05,
+      "loss": 3.6584,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883458587505069e-05,
+      "loss": 3.6726,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882619992754017e-05,
+      "loss": 3.6494,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881781398002966e-05,
+      "loss": 3.6569,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880944441132287e-05,
+      "loss": 3.6517,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880105846381235e-05,
+      "loss": 3.6512,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879267251630183e-05,
+      "loss": 3.6507,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878428656879131e-05,
+      "loss": 3.6555,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8775917000084516e-05,
+      "loss": 3.6638,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8767531052573996e-05,
+      "loss": 3.6538,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8759145105063476e-05,
+      "loss": 3.6572,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.850785255432129,
+      "eval_runtime": 304.0309,
+      "eval_samples_per_second": 1255.106,
+      "eval_steps_per_second": 39.223,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8750759157552956e-05,
+      "loss": 3.6457,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8742373210042436e-05,
+      "loss": 3.6406,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8734003641335645e-05,
+      "loss": 3.6571,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8725617693825125e-05,
+      "loss": 3.655,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871723174631461e-05,
+      "loss": 3.6666,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870884579880409e-05,
+      "loss": 3.6481,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.87004762300973e-05,
+      "loss": 3.6515,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869209028258678e-05,
+      "loss": 3.6404,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.868370433507626e-05,
+      "loss": 3.6494,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.867531838756574e-05,
+      "loss": 3.6569,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.866693244005522e-05,
+      "loss": 3.6507,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865856287134843e-05,
+      "loss": 3.6573,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865017692383791e-05,
+      "loss": 3.664,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864179097632739e-05,
+      "loss": 3.6399,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.863340502881687e-05,
+      "loss": 3.6437,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862503546011008e-05,
+      "loss": 3.64,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8616649512599565e-05,
+      "loss": 3.6494,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8608263565089045e-05,
+      "loss": 3.6421,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8599877617578525e-05,
+      "loss": 3.6457,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8591491670068e-05,
+      "loss": 3.6484,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8583122101361214e-05,
+      "loss": 3.6568,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8574736153850694e-05,
+      "loss": 3.6514,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8566350206340174e-05,
+      "loss": 3.6457,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855796425882965e-05,
+      "loss": 3.6468,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854959469012286e-05,
+      "loss": 3.6492,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854120874261234e-05,
+      "loss": 3.6524,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853282279510182e-05,
+      "loss": 3.6367,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85244368475913e-05,
+      "loss": 3.6565,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851606727888452e-05,
+      "loss": 3.6409,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8507681331374e-05,
+      "loss": 3.6377,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849929538386347e-05,
+      "loss": 3.6502,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849090943635295e-05,
+      "loss": 3.6488,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848253986764617e-05,
+      "loss": 3.6468,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847415392013565e-05,
+      "loss": 3.6552,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846576797262512e-05,
+      "loss": 3.6397,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84573820251146e-05,
+      "loss": 3.6429,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.844899607760408e-05,
+      "loss": 3.6474,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.844061013009356e-05,
+      "loss": 3.632,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843224056138677e-05,
+      "loss": 3.6397,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.842385461387626e-05,
+      "loss": 3.6326,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.841546866636574e-05,
+      "loss": 3.6383,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840708271885522e-05,
+      "loss": 3.6457,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8398713150148426e-05,
+      "loss": 3.6429,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8390327202637906e-05,
+      "loss": 3.6513,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8381941255127386e-05,
+      "loss": 3.6423,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8373555307616866e-05,
+      "loss": 3.6483,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8365185738910075e-05,
+      "loss": 3.6437,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8356799791399555e-05,
+      "loss": 3.6418,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8348413843889035e-05,
+      "loss": 3.6221,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8340027896378515e-05,
+      "loss": 3.6475,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8331658327671724e-05,
+      "loss": 3.6358,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832327238016121e-05,
+      "loss": 3.6488,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831488643265069e-05,
+      "loss": 3.637,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830650048514017e-05,
+      "loss": 3.638,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.829813091643338e-05,
+      "loss": 3.6278,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828974496892286e-05,
+      "loss": 3.6364,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828135902141234e-05,
+      "loss": 3.6477,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.827297307390182e-05,
+      "loss": 3.6371,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826460350519503e-05,
+      "loss": 3.6517,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.825621755768451e-05,
+      "loss": 3.642,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824783161017399e-05,
+      "loss": 3.6131,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823944566266347e-05,
+      "loss": 3.6511,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823105971515295e-05,
+      "loss": 3.6312,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822269014644616e-05,
+      "loss": 3.6232,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8214304198935644e-05,
+      "loss": 3.6387,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8205918251425124e-05,
+      "loss": 3.6416,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8197532303914604e-05,
+      "loss": 3.6242,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818916273520781e-05,
+      "loss": 3.6314,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818077678769729e-05,
+      "loss": 3.6238,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817239084018677e-05,
+      "loss": 3.643,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816400489267625e-05,
+      "loss": 3.6397,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815563532396946e-05,
+      "loss": 3.634,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814724937645894e-05,
+      "loss": 3.6449,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813886342894842e-05,
+      "loss": 3.6372,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81304774814379e-05,
+      "loss": 3.65,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812210791273111e-05,
+      "loss": 3.6275,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81137219652206e-05,
+      "loss": 3.6407,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810533601771008e-05,
+      "loss": 3.633,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.809695007019956e-05,
+      "loss": 3.6335,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808856412268904e-05,
+      "loss": 3.6396,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808019455398225e-05,
+      "loss": 3.6401,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8071808606471727e-05,
+      "loss": 3.6437,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8063422658961207e-05,
+      "loss": 3.6459,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8055036711450686e-05,
+      "loss": 3.6294,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8046667142743896e-05,
+      "loss": 3.6292,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8038281195233376e-05,
+      "loss": 3.6301,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8029895247722855e-05,
+      "loss": 3.6413,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8021509300212335e-05,
+      "loss": 3.6304,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.801313973150555e-05,
+      "loss": 3.6358,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.800475378399503e-05,
+      "loss": 3.6391,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.799636783648451e-05,
+      "loss": 3.6393,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798798188897399e-05,
+      "loss": 3.6355,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79796123202672e-05,
+      "loss": 3.6341,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797122637275668e-05,
+      "loss": 3.6389,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796284042524616e-05,
+      "loss": 3.6425,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.795445447773564e-05,
+      "loss": 3.6462,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.794608490902885e-05,
+      "loss": 3.6275,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793769896151833e-05,
+      "loss": 3.6368,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792931301400781e-05,
+      "loss": 3.6368,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792092706649729e-05,
+      "loss": 3.6429,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7912557497790505e-05,
+      "loss": 3.6392,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7904171550279985e-05,
+      "loss": 3.6391,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7895785602769465e-05,
+      "loss": 3.6323,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7887399655258945e-05,
+      "loss": 3.6352,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7879030086552154e-05,
+      "loss": 3.6199,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7870644139041634e-05,
+      "loss": 3.6319,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7862258191531114e-05,
+      "loss": 3.6225,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7853872244020594e-05,
+      "loss": 3.6394,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78455026753138e-05,
+      "loss": 3.6273,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.783711672780328e-05,
+      "loss": 3.631,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782873078029276e-05,
+      "loss": 3.6353,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782034483278224e-05,
+      "loss": 3.6373,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781197526407546e-05,
+      "loss": 3.6384,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780358931656494e-05,
+      "loss": 3.6335,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.779520336905442e-05,
+      "loss": 3.6391,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77868174215439e-05,
+      "loss": 3.6409,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777844785283711e-05,
+      "loss": 3.6363,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777006190532659e-05,
+      "loss": 3.6328,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776167595781607e-05,
+      "loss": 3.6289,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.775329001030555e-05,
+      "loss": 3.6352,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7744920441598757e-05,
+      "loss": 3.6306,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7736534494088237e-05,
+      "loss": 3.6178,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7728148546577716e-05,
+      "loss": 3.6279,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7719762599067196e-05,
+      "loss": 3.6279,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771139303036041e-05,
+      "loss": 3.6304,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770300708284989e-05,
+      "loss": 3.6411,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769462113533937e-05,
+      "loss": 3.6313,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768623518782885e-05,
+      "loss": 3.6238,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7677849240318325e-05,
+      "loss": 3.6292,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766947967161154e-05,
+      "loss": 3.6442,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766109372410102e-05,
+      "loss": 3.6321,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7652707776590494e-05,
+      "loss": 3.6404,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7644321829079974e-05,
+      "loss": 3.6196,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763595226037319e-05,
+      "loss": 3.6452,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.762756631286267e-05,
+      "loss": 3.6371,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761918036535215e-05,
+      "loss": 3.6353,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761079441784163e-05,
+      "loss": 3.6332,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760240847033111e-05,
+      "loss": 3.6295,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759402252282059e-05,
+      "loss": 3.6327,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75856529541138e-05,
+      "loss": 3.6557,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.757726700660328e-05,
+      "loss": 3.6229,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756888105909276e-05,
+      "loss": 3.6357,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756049511158224e-05,
+      "loss": 3.6301,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755212554287545e-05,
+      "loss": 3.621,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754373959536493e-05,
+      "loss": 3.6293,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753535364785441e-05,
+      "loss": 3.6331,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752698407914762e-05,
+      "loss": 3.6386,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7518598131637104e-05,
+      "loss": 3.6308,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7510212184126584e-05,
+      "loss": 3.632,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.847637891769409,
+      "eval_runtime": 304.0055,
+      "eval_samples_per_second": 1255.211,
+      "eval_steps_per_second": 39.226,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7501826236616064e-05,
+      "loss": 3.6339,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7493440289105544e-05,
+      "loss": 3.6172,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7485054341595024e-05,
+      "loss": 3.6334,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7476668394084504e-05,
+      "loss": 3.6348,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746829882537771e-05,
+      "loss": 3.643,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745991287786719e-05,
+      "loss": 3.6295,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745152693035667e-05,
+      "loss": 3.6263,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744314098284615e-05,
+      "loss": 3.6201,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.743475503533563e-05,
+      "loss": 3.6229,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.742638546662884e-05,
+      "loss": 3.634,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.741799951911832e-05,
+      "loss": 3.6293,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.74096135716078e-05,
+      "loss": 3.6368,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740124400290102e-05,
+      "loss": 3.6362,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73928580553905e-05,
+      "loss": 3.6179,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738447210787998e-05,
+      "loss": 3.6186,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737608616036946e-05,
+      "loss": 3.6154,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736770021285894e-05,
+      "loss": 3.6314,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735931426534842e-05,
+      "loss": 3.6206,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73509283178379e-05,
+      "loss": 3.6205,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734254237032738e-05,
+      "loss": 3.6231,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7334172801620586e-05,
+      "loss": 3.6362,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7325786854110066e-05,
+      "loss": 3.6244,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7317400906599546e-05,
+      "loss": 3.6279,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7309014959089026e-05,
+      "loss": 3.6235,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7300645390382235e-05,
+      "loss": 3.629,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729225944287172e-05,
+      "loss": 3.626,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.72838734953612e-05,
+      "loss": 3.6141,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727548754785068e-05,
+      "loss": 3.6331,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726711797914389e-05,
+      "loss": 3.6149,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725873203163337e-05,
+      "loss": 3.6129,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.725034608412285e-05,
+      "loss": 3.6324,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724196013661233e-05,
+      "loss": 3.6234,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723359056790554e-05,
+      "loss": 3.6239,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722520462039502e-05,
+      "loss": 3.6315,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.72168186728845e-05,
+      "loss": 3.6174,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720843272537398e-05,
+      "loss": 3.6223,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720006315666719e-05,
+      "loss": 3.6254,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7191677209156676e-05,
+      "loss": 3.6082,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7183291261646155e-05,
+      "loss": 3.6176,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717490531413563e-05,
+      "loss": 3.6115,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7166535745428845e-05,
+      "loss": 3.6149,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7158149797918324e-05,
+      "loss": 3.6256,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7149763850407804e-05,
+      "loss": 3.6203,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714137790289728e-05,
+      "loss": 3.6287,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7133008334190493e-05,
+      "loss": 3.6185,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7124622386679973e-05,
+      "loss": 3.6248,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7116236439169453e-05,
+      "loss": 3.6183,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7107850491658927e-05,
+      "loss": 3.6178,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709948092295214e-05,
+      "loss": 3.5993,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709109497544163e-05,
+      "loss": 3.6261,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70827090279311e-05,
+      "loss": 3.6144,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.707432308042058e-05,
+      "loss": 3.6263,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70659535117138e-05,
+      "loss": 3.6155,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.705756756420328e-05,
+      "loss": 3.6106,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704918161669275e-05,
+      "loss": 3.6077,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704079566918223e-05,
+      "loss": 3.6128,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703240972167171e-05,
+      "loss": 3.6256,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702404015296493e-05,
+      "loss": 3.6137,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70156542054544e-05,
+      "loss": 3.6317,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.700726825794388e-05,
+      "loss": 3.6205,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699888231043337e-05,
+      "loss": 3.5897,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6990512741726576e-05,
+      "loss": 3.6292,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6982126794216056e-05,
+      "loss": 3.61,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6973740846705536e-05,
+      "loss": 3.6019,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6965354899195016e-05,
+      "loss": 3.6136,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6956985330488225e-05,
+      "loss": 3.6219,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6948599382977705e-05,
+      "loss": 3.6025,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6940213435467185e-05,
+      "loss": 3.6112,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6931827487956665e-05,
+      "loss": 3.6029,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6923457919249874e-05,
+      "loss": 3.6144,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6915071971739354e-05,
+      "loss": 3.6218,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6906686024228834e-05,
+      "loss": 3.61,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689830007671832e-05,
+      "loss": 3.6221,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688993050801153e-05,
+      "loss": 3.6117,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688154456050101e-05,
+      "loss": 3.6272,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.687315861299049e-05,
+      "loss": 3.606,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686477266547997e-05,
+      "loss": 3.6222,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685638671796945e-05,
+      "loss": 3.6099,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684801714926266e-05,
+      "loss": 3.6095,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683963120175214e-05,
+      "loss": 3.6182,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683124525424162e-05,
+      "loss": 3.6154,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68228593067311e-05,
+      "loss": 3.6191,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681448973802431e-05,
+      "loss": 3.6283,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680610379051379e-05,
+      "loss": 3.6107,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6797717843003274e-05,
+      "loss": 3.6047,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6789331895492754e-05,
+      "loss": 3.6064,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678096232678596e-05,
+      "loss": 3.6191,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677257637927544e-05,
+      "loss": 3.607,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676419043176492e-05,
+      "loss": 3.6167,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.67558044842544e-05,
+      "loss": 3.6173,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.674743491554761e-05,
+      "loss": 3.616,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673904896803709e-05,
+      "loss": 3.6118,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673066302052657e-05,
+      "loss": 3.6107,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672227707301605e-05,
+      "loss": 3.6201,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.671390750430926e-05,
+      "loss": 3.6183,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.670552155679874e-05,
+      "loss": 3.6219,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669713560928822e-05,
+      "loss": 3.6099,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668874966177771e-05,
+      "loss": 3.611,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668038009307092e-05,
+      "loss": 3.6186,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66719941455604e-05,
+      "loss": 3.6199,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.666360819804988e-05,
+      "loss": 3.6168,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.665522225053936e-05,
+      "loss": 3.6145,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6646852681832566e-05,
+      "loss": 3.6139,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6638466734322046e-05,
+      "loss": 3.6103,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6630080786811526e-05,
+      "loss": 3.6028,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6621694839301006e-05,
+      "loss": 3.6079,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6613325270594215e-05,
+      "loss": 3.6017,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6604939323083695e-05,
+      "loss": 3.6191,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6596553375573175e-05,
+      "loss": 3.6065,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.658816742806266e-05,
+      "loss": 3.6099,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657979785935587e-05,
+      "loss": 3.6097,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657141191184535e-05,
+      "loss": 3.6206,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656302596433483e-05,
+      "loss": 3.6135,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655464001682431e-05,
+      "loss": 3.6125,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654627044811752e-05,
+      "loss": 3.6203,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6537884500607e-05,
+      "loss": 3.6201,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652949855309648e-05,
+      "loss": 3.6144,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652111260558596e-05,
+      "loss": 3.6092,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651274303687917e-05,
+      "loss": 3.607,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650435708936865e-05,
+      "loss": 3.6143,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649597114185813e-05,
+      "loss": 3.6117,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6487585194347615e-05,
+      "loss": 3.5915,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6479215625640824e-05,
+      "loss": 3.6106,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6470829678130304e-05,
+      "loss": 3.6077,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6462443730619784e-05,
+      "loss": 3.6081,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6454057783109264e-05,
+      "loss": 3.6189,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644568821440247e-05,
+      "loss": 3.6094,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643730226689195e-05,
+      "loss": 3.6048,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642891631938143e-05,
+      "loss": 3.6045,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642053037187091e-05,
+      "loss": 3.6252,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641214442436039e-05,
+      "loss": 3.6133,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64037748556536e-05,
+      "loss": 3.6199,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.639538890814308e-05,
+      "loss": 3.5993,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.638700296063257e-05,
+      "loss": 3.6202,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637861701312205e-05,
+      "loss": 3.6148,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637024744441526e-05,
+      "loss": 3.6158,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636186149690474e-05,
+      "loss": 3.6128,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635347554939422e-05,
+      "loss": 3.6033,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634510598068743e-05,
+      "loss": 3.6131,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633672003317691e-05,
+      "loss": 3.6329,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632833408566639e-05,
+      "loss": 3.6027,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631994813815587e-05,
+      "loss": 3.6119,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631156219064535e-05,
+      "loss": 3.6069,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6303192621938556e-05,
+      "loss": 3.6065,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6294806674428036e-05,
+      "loss": 3.6046,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628642072691752e-05,
+      "loss": 3.6115,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6278034779407e-05,
+      "loss": 3.6146,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626966521070021e-05,
+      "loss": 3.616,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626127926318969e-05,
+      "loss": 3.6086,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.845874071121216,
+      "eval_runtime": 310.4534,
+      "eval_samples_per_second": 1229.141,
+      "eval_steps_per_second": 38.412,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.625289331567917e-05,
+      "loss": 3.6006,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.624450736816865e-05,
+      "loss": 3.5962,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.623613779946186e-05,
+      "loss": 3.6132,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622775185195134e-05,
+      "loss": 3.6135,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621936590444082e-05,
+      "loss": 3.6189,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.62109799569303e-05,
+      "loss": 3.6112,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620261038822351e-05,
+      "loss": 3.6059,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.619422444071299e-05,
+      "loss": 3.5989,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6185838493202476e-05,
+      "loss": 3.6016,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617745254569195e-05,
+      "loss": 3.6048,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6169082976985165e-05,
+      "loss": 3.6157,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6160697029474645e-05,
+      "loss": 3.6177,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6152327460767854e-05,
+      "loss": 3.6219,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6143941513257334e-05,
+      "loss": 3.5904,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6135555565746814e-05,
+      "loss": 3.6021,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6127169618236294e-05,
+      "loss": 3.5898,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6118783670725774e-05,
+      "loss": 3.6109,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611039772321525e-05,
+      "loss": 3.5997,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610201177570473e-05,
+      "loss": 3.5968,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6093625828194214e-05,
+      "loss": 3.6041,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608525625948742e-05,
+      "loss": 3.6184,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60768703119769e-05,
+      "loss": 3.5997,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606848436446638e-05,
+      "loss": 3.6071,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606009841695586e-05,
+      "loss": 3.6022,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605172884824907e-05,
+      "loss": 3.6098,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.604334290073855e-05,
+      "loss": 3.6023,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603495695322803e-05,
+      "loss": 3.5946,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602657100571751e-05,
+      "loss": 3.6101,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601820143701072e-05,
+      "loss": 3.5972,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60098154895002e-05,
+      "loss": 3.5937,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600142954198968e-05,
+      "loss": 3.6092,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599304359447917e-05,
+      "loss": 3.6033,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598467402577238e-05,
+      "loss": 3.6061,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597628807826186e-05,
+      "loss": 3.61,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596790213075134e-05,
+      "loss": 3.5969,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595951618324082e-05,
+      "loss": 3.6032,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5951146614534026e-05,
+      "loss": 3.6037,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5942760667023506e-05,
+      "loss": 3.5858,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5934374719512986e-05,
+      "loss": 3.5982,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5925988772002466e-05,
+      "loss": 3.5886,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5917619203295675e-05,
+      "loss": 3.5932,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5909233255785155e-05,
+      "loss": 3.6041,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5900847308274635e-05,
+      "loss": 3.6056,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589246136076412e-05,
+      "loss": 3.6018,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588409179205733e-05,
+      "loss": 3.6026,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587570584454681e-05,
+      "loss": 3.6002,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586731989703629e-05,
+      "loss": 3.6011,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585893394952577e-05,
+      "loss": 3.5946,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585056438081898e-05,
+      "loss": 3.5832,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584217843330846e-05,
+      "loss": 3.6014,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583379248579794e-05,
+      "loss": 3.596,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582540653828742e-05,
+      "loss": 3.6042,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581703696958063e-05,
+      "loss": 3.5994,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580865102207011e-05,
+      "loss": 3.5915,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580026507455959e-05,
+      "loss": 3.5879,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579187912704907e-05,
+      "loss": 3.5902,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5783493179538555e-05,
+      "loss": 3.6062,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5775123610831764e-05,
+      "loss": 3.5883,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5766737663321244e-05,
+      "loss": 3.6139,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5758351715810724e-05,
+      "loss": 3.6001,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5749965768300204e-05,
+      "loss": 3.5742,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574159619959341e-05,
+      "loss": 3.6046,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573321025208289e-05,
+      "loss": 3.5891,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572482430457237e-05,
+      "loss": 3.5829,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571643835706185e-05,
+      "loss": 3.5928,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570806878835506e-05,
+      "loss": 3.5997,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569968284084454e-05,
+      "loss": 3.5818,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569129689333402e-05,
+      "loss": 3.5926,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568291094582351e-05,
+      "loss": 3.5811,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567454137711672e-05,
+      "loss": 3.5957,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56661554296062e-05,
+      "loss": 3.6014,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565776948209568e-05,
+      "loss": 3.5923,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564938353458516e-05,
+      "loss": 3.6014,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564101396587837e-05,
+      "loss": 3.5916,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563262801836785e-05,
+      "loss": 3.6069,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562424207085733e-05,
+      "loss": 3.5894,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5615856123346807e-05,
+      "loss": 3.5994,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5607486554640016e-05,
+      "loss": 3.5859,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5599100607129496e-05,
+      "loss": 3.5908,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5590714659618976e-05,
+      "loss": 3.6019,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558232871210846e-05,
+      "loss": 3.5933,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557395914340167e-05,
+      "loss": 3.5996,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556557319589115e-05,
+      "loss": 3.6084,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555718724838063e-05,
+      "loss": 3.5915,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554880130087011e-05,
+      "loss": 3.5772,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554043173216332e-05,
+      "loss": 3.59,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55320457846528e-05,
+      "loss": 3.5988,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.552365983714228e-05,
+      "loss": 3.5924,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551527388963176e-05,
+      "loss": 3.5932,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550690432092497e-05,
+      "loss": 3.6002,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549851837341445e-05,
+      "loss": 3.5939,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549013242590393e-05,
+      "loss": 3.5929,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5481746478393416e-05,
+      "loss": 3.5915,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5473376909686625e-05,
+      "loss": 3.5993,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5464990962176105e-05,
+      "loss": 3.601,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5456605014665585e-05,
+      "loss": 3.6063,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5448219067155065e-05,
+      "loss": 3.5852,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5439849498448274e-05,
+      "loss": 3.5966,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5431463550937754e-05,
+      "loss": 3.5967,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5423077603427234e-05,
+      "loss": 3.5973,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5414691655916714e-05,
+      "loss": 3.5991,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.540632208720992e-05,
+      "loss": 3.5939,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53979361396994e-05,
+      "loss": 3.5956,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538955019218888e-05,
+      "loss": 3.5885,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538116424467837e-05,
+      "loss": 3.5908,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537279467597158e-05,
+      "loss": 3.5843,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536440872846106e-05,
+      "loss": 3.5848,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535602278095054e-05,
+      "loss": 3.5971,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534763683344002e-05,
+      "loss": 3.5898,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533926726473323e-05,
+      "loss": 3.5836,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533088131722271e-05,
+      "loss": 3.587,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532249536971219e-05,
+      "loss": 3.6021,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531410942220167e-05,
+      "loss": 3.598,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.530573985349488e-05,
+      "loss": 3.5932,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.529735390598436e-05,
+      "loss": 3.6009,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528896795847384e-05,
+      "loss": 3.6009,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528058201096332e-05,
+      "loss": 3.5929,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527221244225653e-05,
+      "loss": 3.5881,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526382649474601e-05,
+      "loss": 3.5901,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.525544054723549e-05,
+      "loss": 3.5911,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524705459972497e-05,
+      "loss": 3.5973,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523868503101818e-05,
+      "loss": 3.5685,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523029908350766e-05,
+      "loss": 3.5908,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522191313599714e-05,
+      "loss": 3.5883,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521352718848662e-05,
+      "loss": 3.5924,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520515761977983e-05,
+      "loss": 3.5972,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519677167226931e-05,
+      "loss": 3.5911,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518838572475879e-05,
+      "loss": 3.5861,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517999977724828e-05,
+      "loss": 3.5817,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517161382973775e-05,
+      "loss": 3.606,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5163244261030966e-05,
+      "loss": 3.5898,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5154858313520446e-05,
+      "loss": 3.6034,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.514647236600992e-05,
+      "loss": 3.5797,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51380864184994e-05,
+      "loss": 3.5963,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5129716849792615e-05,
+      "loss": 3.5998,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5121330902282095e-05,
+      "loss": 3.5922,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.511294495477157e-05,
+      "loss": 3.5963,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.510455900726105e-05,
+      "loss": 3.5848,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5096189438554264e-05,
+      "loss": 3.5948,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5087803491043744e-05,
+      "loss": 3.609,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5079417543533224e-05,
+      "loss": 3.5882,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5071031596022704e-05,
+      "loss": 3.5912,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506266202731592e-05,
+      "loss": 3.5908,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505427607980539e-05,
+      "loss": 3.5848,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504589013229487e-05,
+      "loss": 3.586,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.503750418478435e-05,
+      "loss": 3.5972,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502913461607757e-05,
+      "loss": 3.5887,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502074866856704e-05,
+      "loss": 3.5967,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501236272105652e-05,
+      "loss": 3.5901,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8455843925476074,
+      "eval_runtime": 311.3215,
+      "eval_samples_per_second": 1225.714,
+      "eval_steps_per_second": 38.304,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5003976773546e-05,
+      "loss": 3.6109,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499559082603548e-05,
+      "loss": 3.5736,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498720487852497e-05,
+      "loss": 3.5903,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497881893101445e-05,
+      "loss": 3.5977,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497044936230766e-05,
+      "loss": 3.5965,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496206341479714e-05,
+      "loss": 3.5931,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495367746728662e-05,
+      "loss": 3.5881,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.49452915197761e-05,
+      "loss": 3.5801,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4936921951069306e-05,
+      "loss": 3.5816,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4928536003558786e-05,
+      "loss": 3.5865,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4920150056048266e-05,
+      "loss": 3.5947,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4911764108537746e-05,
+      "loss": 3.596,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.490341091863469e-05,
+      "loss": 3.6065,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489502497112417e-05,
+      "loss": 3.5695,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.488663902361365e-05,
+      "loss": 3.5806,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487825307610313e-05,
+      "loss": 3.5696,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486986712859261e-05,
+      "loss": 3.5925,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486148118108209e-05,
+      "loss": 3.5812,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.485309523357157e-05,
+      "loss": 3.5785,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484470928606105e-05,
+      "loss": 3.5807,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483633971735426e-05,
+      "loss": 3.6,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482795376984374e-05,
+      "loss": 3.5815,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481956782233322e-05,
+      "loss": 3.5878,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48111818748227e-05,
+      "loss": 3.5844,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480281230611591e-05,
+      "loss": 3.5975,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479442635860539e-05,
+      "loss": 3.583,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478604041109487e-05,
+      "loss": 3.5755,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4777654463584356e-05,
+      "loss": 3.5929,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4769284894877565e-05,
+      "loss": 3.5725,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4760898947367045e-05,
+      "loss": 3.5741,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4752512999856525e-05,
+      "loss": 3.591,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4744127052346005e-05,
+      "loss": 3.5844,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4735757483639214e-05,
+      "loss": 3.5848,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4727371536128694e-05,
+      "loss": 3.5928,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4718985588618174e-05,
+      "loss": 3.5775,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4710599641107654e-05,
+      "loss": 3.5831,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470223007240086e-05,
+      "loss": 3.5855,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469384412489034e-05,
+      "loss": 3.5695,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468545817737982e-05,
+      "loss": 3.5806,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467707222986931e-05,
+      "loss": 3.5686,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466870266116252e-05,
+      "loss": 3.576,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4660316713652e-05,
+      "loss": 3.5824,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465193076614148e-05,
+      "loss": 3.5897,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464354481863096e-05,
+      "loss": 3.5842,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463517524992417e-05,
+      "loss": 3.583,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462678930241365e-05,
+      "loss": 3.5843,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461840335490313e-05,
+      "loss": 3.5805,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461001740739261e-05,
+      "loss": 3.5774,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4601647838685816e-05,
+      "loss": 3.5634,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4593261891175296e-05,
+      "loss": 3.5811,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4584875943664776e-05,
+      "loss": 3.5806,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.457648999615426e-05,
+      "loss": 3.5829,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.456810404864374e-05,
+      "loss": 3.5818,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455973447993695e-05,
+      "loss": 3.5727,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455134853242643e-05,
+      "loss": 3.5681,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.454296258491591e-05,
+      "loss": 3.5711,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.453457663740539e-05,
+      "loss": 3.588,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45262070686986e-05,
+      "loss": 3.5734,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.451782112118808e-05,
+      "loss": 3.5929,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450943517367756e-05,
+      "loss": 3.5796,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450104922616704e-05,
+      "loss": 3.5611,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449267965746025e-05,
+      "loss": 3.5832,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448429370994973e-05,
+      "loss": 3.5715,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447590776243922e-05,
+      "loss": 3.5643,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44675218149287e-05,
+      "loss": 3.5705,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4459152246221906e-05,
+      "loss": 3.582,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4450766298711386e-05,
+      "loss": 3.5638,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4442380351200866e-05,
+      "loss": 3.5795,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4433994403690346e-05,
+      "loss": 3.5585,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4425624834983555e-05,
+      "loss": 3.5778,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4417238887473035e-05,
+      "loss": 3.5819,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4408852939962515e-05,
+      "loss": 3.574,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4400466992451995e-05,
+      "loss": 3.5812,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4392097423745204e-05,
+      "loss": 3.5731,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4383711476234684e-05,
+      "loss": 3.5879,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.437532552872417e-05,
+      "loss": 3.5717,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436693958121365e-05,
+      "loss": 3.5832,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435857001250686e-05,
+      "loss": 3.5719,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435018406499634e-05,
+      "loss": 3.5712,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434179811748582e-05,
+      "loss": 3.5796,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433341216997529e-05,
+      "loss": 3.5734,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432504260126851e-05,
+      "loss": 3.5825,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.431665665375799e-05,
+      "loss": 3.5868,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.430827070624747e-05,
+      "loss": 3.5778,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429988475873694e-05,
+      "loss": 3.5589,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429151519003016e-05,
+      "loss": 3.571,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.428312924251964e-05,
+      "loss": 3.5772,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4274743295009124e-05,
+      "loss": 3.576,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.42663573474986e-05,
+      "loss": 3.5731,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425798777879181e-05,
+      "loss": 3.5807,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424960183128129e-05,
+      "loss": 3.5764,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4241215883770766e-05,
+      "loss": 3.5732,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4232829936260246e-05,
+      "loss": 3.5742,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.422446036755346e-05,
+      "loss": 3.5814,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421607442004294e-05,
+      "loss": 3.5808,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4207688472532415e-05,
+      "loss": 3.5872,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4199302525021895e-05,
+      "loss": 3.5659,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419093295631511e-05,
+      "loss": 3.5773,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418254700880459e-05,
+      "loss": 3.5792,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417416106129407e-05,
+      "loss": 3.5796,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.416577511378355e-05,
+      "loss": 3.5825,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.415740554507677e-05,
+      "loss": 3.5752,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414901959756624e-05,
+      "loss": 3.5762,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414063365005572e-05,
+      "loss": 3.5715,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41322477025452e-05,
+      "loss": 3.5733,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4123878133838416e-05,
+      "loss": 3.5641,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.411549218632789e-05,
+      "loss": 3.5668,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.410710623881737e-05,
+      "loss": 3.5779,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409872029130685e-05,
+      "loss": 3.5752,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4090350722600065e-05,
+      "loss": 3.5673,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4081964775089545e-05,
+      "loss": 3.5657,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4073578827579025e-05,
+      "loss": 3.5842,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4065192880068505e-05,
+      "loss": 3.5796,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4056823311361714e-05,
+      "loss": 3.5757,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4048437363851194e-05,
+      "loss": 3.5821,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4040051416340674e-05,
+      "loss": 3.5847,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4031665468830154e-05,
+      "loss": 3.5753,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402329590012336e-05,
+      "loss": 3.5709,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401490995261284e-05,
+      "loss": 3.5743,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400652400510232e-05,
+      "loss": 3.5749,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39981380575918e-05,
+      "loss": 3.5793,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398976848888502e-05,
+      "loss": 3.552,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39813825413745e-05,
+      "loss": 3.5722,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397299659386398e-05,
+      "loss": 3.5697,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.396462702515719e-05,
+      "loss": 3.5713,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.395624107764667e-05,
+      "loss": 3.5801,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.394785513013615e-05,
+      "loss": 3.5733,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393946918262563e-05,
+      "loss": 3.5672,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393108323511511e-05,
+      "loss": 3.5647,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3922713666408316e-05,
+      "loss": 3.5878,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3914327718897796e-05,
+      "loss": 3.5717,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3905941771387276e-05,
+      "loss": 3.5839,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3897555823876756e-05,
+      "loss": 3.5644,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.388918625516997e-05,
+      "loss": 3.578,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.388080030765945e-05,
+      "loss": 3.5817,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387241436014893e-05,
+      "loss": 3.5753,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.386402841263841e-05,
+      "loss": 3.5748,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.385565884393162e-05,
+      "loss": 3.5688,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38472728964211e-05,
+      "loss": 3.5773,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383888694891058e-05,
+      "loss": 3.593,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383050100140006e-05,
+      "loss": 3.5713,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382213143269327e-05,
+      "loss": 3.5693,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.381374548518275e-05,
+      "loss": 3.5747,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380535953767223e-05,
+      "loss": 3.5679,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379697359016171e-05,
+      "loss": 3.565,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3788604021454926e-05,
+      "loss": 3.5837,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3780218073944406e-05,
+      "loss": 3.5679,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3771832126433886e-05,
+      "loss": 3.5759,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3763446178923366e-05,
+      "loss": 3.5753,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.84504771232605,
+      "eval_runtime": 307.6838,
+      "eval_samples_per_second": 1240.205,
+      "eval_steps_per_second": 38.757,
+      "step": 992160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3755076610216575e-05,
+      "loss": 3.5918,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3746690662706055e-05,
+      "loss": 3.5613,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3738304715195535e-05,
+      "loss": 3.5699,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3729918767685015e-05,
+      "loss": 3.5818,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3721549198978224e-05,
+      "loss": 3.5774,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3713163251467704e-05,
+      "loss": 3.5763,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3704777303957184e-05,
+      "loss": 3.5662,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3696391356446664e-05,
+      "loss": 3.5633,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.368802178773988e-05,
+      "loss": 3.5643,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367963584022936e-05,
+      "loss": 3.5729,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367124989271884e-05,
+      "loss": 3.57,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.366286394520832e-05,
+      "loss": 3.585,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.365449437650153e-05,
+      "loss": 3.583,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.364610842899101e-05,
+      "loss": 3.5531,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.363772248148049e-05,
+      "loss": 3.5649,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.362933653396997e-05,
+      "loss": 3.5541,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.362096696526318e-05,
+      "loss": 3.5761,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.361258101775266e-05,
+      "loss": 3.5602,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.360419507024214e-05,
+      "loss": 3.5623,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.359580912273162e-05,
+      "loss": 3.5658,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.358743955402483e-05,
+      "loss": 3.5772,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357905360651431e-05,
+      "loss": 3.5674,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357066765900379e-05,
+      "loss": 3.5681,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.356228171149327e-05,
+      "loss": 3.5693,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.355391214278648e-05,
+      "loss": 3.5765,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.354552619527596e-05,
+      "loss": 3.5687,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.353714024776544e-05,
+      "loss": 3.5588,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.352875430025492e-05,
+      "loss": 3.5734,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.352038473154813e-05,
+      "loss": 3.5549,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351199878403761e-05,
+      "loss": 3.5611,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.350361283652709e-05,
+      "loss": 3.5701,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.349522688901657e-05,
+      "loss": 3.5653,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.348685732030979e-05,
+      "loss": 3.5688,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347847137279927e-05,
+      "loss": 3.5738,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347008542528875e-05,
+      "loss": 3.5632,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346169947777823e-05,
+      "loss": 3.5663,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3453329909071436e-05,
+      "loss": 3.5678,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3444943961560916e-05,
+      "loss": 3.5523,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3436558014050396e-05,
+      "loss": 3.5645,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3428172066539876e-05,
+      "loss": 3.5521,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3419802497833085e-05,
+      "loss": 3.5625,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3411416550322565e-05,
+      "loss": 3.5584,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3403030602812045e-05,
+      "loss": 3.5733,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3394644655301525e-05,
+      "loss": 3.5657,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.338627508659474e-05,
+      "loss": 3.5666,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.337788913908422e-05,
+      "loss": 3.5668,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.33695031915737e-05,
+      "loss": 3.5631,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336111724406318e-05,
+      "loss": 3.5604,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.335274767535639e-05,
+      "loss": 3.5461,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.334436172784587e-05,
+      "loss": 3.5625,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.333597578033535e-05,
+      "loss": 3.5668,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.332758983282483e-05,
+      "loss": 3.5667,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.331922026411804e-05,
+      "loss": 3.5603,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.331083431660752e-05,
+      "loss": 3.5545,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3302448369097e-05,
+      "loss": 3.5548,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.329406242158648e-05,
+      "loss": 3.55,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3285692852879694e-05,
+      "loss": 3.578,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3277306905369174e-05,
+      "loss": 3.5549,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3268920957858654e-05,
+      "loss": 3.5738,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3260535010348134e-05,
+      "loss": 3.563,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325216544164134e-05,
+      "loss": 3.5441,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.324377949413082e-05,
+      "loss": 3.5653,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.32353935466203e-05,
+      "loss": 3.5519,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.322700759910978e-05,
+      "loss": 3.5514,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321863803040299e-05,
+      "loss": 3.5517,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321025208289247e-05,
+      "loss": 3.5661,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320186613538195e-05,
+      "loss": 3.5475,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.319348018787143e-05,
+      "loss": 3.5597,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.318511061916465e-05,
+      "loss": 3.5434,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.317672467165413e-05,
+      "loss": 3.5586,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.316833872414361e-05,
+      "loss": 3.5651,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315995277663309e-05,
+      "loss": 3.5602,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.31515832079263e-05,
+      "loss": 3.5628,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.314319726041578e-05,
+      "loss": 3.5575,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.313481131290526e-05,
+      "loss": 3.5697,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3126425365394737e-05,
+      "loss": 3.5523,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3118055796687946e-05,
+      "loss": 3.564,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3109669849177426e-05,
+      "loss": 3.558,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3101283901666906e-05,
+      "loss": 3.5526,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3092897954156386e-05,
+      "loss": 3.5616,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3084528385449595e-05,
+      "loss": 3.5596,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.307614243793908e-05,
+      "loss": 3.5674,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.306775649042856e-05,
+      "loss": 3.5713,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.305937054291804e-05,
+      "loss": 3.5566,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.305100097421125e-05,
+      "loss": 3.5471,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304261502670073e-05,
+      "loss": 3.5534,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.303422907919021e-05,
+      "loss": 3.5606,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3025843131679683e-05,
+      "loss": 3.5596,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30174735629729e-05,
+      "loss": 3.5575,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300908761546238e-05,
+      "loss": 3.5651,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.300070166795186e-05,
+      "loss": 3.5588,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299231572044133e-05,
+      "loss": 3.5543,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.298394615173455e-05,
+      "loss": 3.5598,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2975560204224035e-05,
+      "loss": 3.5661,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.296717425671351e-05,
+      "loss": 3.5611,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.295878830920299e-05,
+      "loss": 3.5719,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2950418740496204e-05,
+      "loss": 3.5453,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2942032792985684e-05,
+      "loss": 3.5668,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.293364684547516e-05,
+      "loss": 3.5645,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.292526089796464e-05,
+      "loss": 3.5591,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.291689132925785e-05,
+      "loss": 3.5648,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.290850538174733e-05,
+      "loss": 3.5592,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2900119434236806e-05,
+      "loss": 3.5591,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2891733486726286e-05,
+      "loss": 3.5552,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.28833639180195e-05,
+      "loss": 3.5573,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.287497797050898e-05,
+      "loss": 3.5473,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.286659202299846e-05,
+      "loss": 3.5541,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.285820607548794e-05,
+      "loss": 3.5584,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284983650678116e-05,
+      "loss": 3.5577,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284145055927063e-05,
+      "loss": 3.5484,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.283306461176011e-05,
+      "loss": 3.5516,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.282467866424959e-05,
+      "loss": 3.5616,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.281630909554281e-05,
+      "loss": 3.5647,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.280792314803228e-05,
+      "loss": 3.5574,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279953720052176e-05,
+      "loss": 3.5681,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279115125301124e-05,
+      "loss": 3.566,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2782781684304456e-05,
+      "loss": 3.5628,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2774395736793936e-05,
+      "loss": 3.5519,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2766009789283416e-05,
+      "loss": 3.557,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2757623841772896e-05,
+      "loss": 3.5558,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2749254273066105e-05,
+      "loss": 3.5641,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2740868325555585e-05,
+      "loss": 3.5374,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2732482378045065e-05,
+      "loss": 3.5543,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2724096430534544e-05,
+      "loss": 3.5491,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2715726861827754e-05,
+      "loss": 3.5558,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2707340914317234e-05,
+      "loss": 3.5654,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2698954966806713e-05,
+      "loss": 3.5558,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2690569019296193e-05,
+      "loss": 3.5477,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268219945058941e-05,
+      "loss": 3.55,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.267381350307889e-05,
+      "loss": 3.5713,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.266542755556837e-05,
+      "loss": 3.556,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.265704160805785e-05,
+      "loss": 3.5701,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264867203935106e-05,
+      "loss": 3.5467,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264028609184054e-05,
+      "loss": 3.5623,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263190014433002e-05,
+      "loss": 3.5643,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.26235141968195e-05,
+      "loss": 3.5586,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.261514462811271e-05,
+      "loss": 3.5533,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.260675868060219e-05,
+      "loss": 3.5551,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.259837273309167e-05,
+      "loss": 3.5644,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258998678558115e-05,
+      "loss": 3.5707,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258161721687436e-05,
+      "loss": 3.5589,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.257323126936384e-05,
+      "loss": 3.5507,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.256484532185332e-05,
+      "loss": 3.5617,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.25564593743428e-05,
+      "loss": 3.55,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.254808980563601e-05,
+      "loss": 3.5466,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253970385812549e-05,
+      "loss": 3.5679,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253131791061497e-05,
+      "loss": 3.5507,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.252293196310445e-05,
+      "loss": 3.5637,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.251456239439766e-05,
+      "loss": 3.5527,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.84675669670105,
+      "eval_runtime": 304.0271,
+      "eval_samples_per_second": 1255.122,
+      "eval_steps_per_second": 39.223,
+      "step": 1068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.250617644688714e-05,
+      "loss": 3.5502,
+      "step": 1068544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.249779049937662e-05,
+      "loss": 3.5469,
+      "step": 1069056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.24894045518661e-05,
+      "loss": 3.5524,
+      "step": 1069568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.248103498315932e-05,
+      "loss": 3.5649,
+      "step": 1070080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2472649035648797e-05,
+      "loss": 3.5649,
+      "step": 1070592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2464263088138277e-05,
+      "loss": 3.5594,
+      "step": 1071104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2455877140627757e-05,
+      "loss": 3.5495,
+      "step": 1071616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2447507571920966e-05,
+      "loss": 3.5468,
+      "step": 1072128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2439121624410446e-05,
+      "loss": 3.5474,
+      "step": 1072640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2430735676899926e-05,
+      "loss": 3.5553,
+      "step": 1073152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2422349729389405e-05,
+      "loss": 3.5579,
+      "step": 1073664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2413980160682615e-05,
+      "loss": 3.5658,
+      "step": 1074176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2405610591975824e-05,
+      "loss": 3.5694,
+      "step": 1074688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.239722464446531e-05,
+      "loss": 3.539,
+      "step": 1075200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.238883869695479e-05,
+      "loss": 3.5467,
+      "step": 1075712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.238045274944427e-05,
+      "loss": 3.5398,
+      "step": 1076224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.237206680193375e-05,
+      "loss": 3.5584,
+      "step": 1076736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.236368085442323e-05,
+      "loss": 3.5451,
+      "step": 1077248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.235529490691271e-05,
+      "loss": 3.5476,
+      "step": 1077760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.234690895940219e-05,
+      "loss": 3.5464,
+      "step": 1078272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.23385393906954e-05,
+      "loss": 3.5599,
+      "step": 1078784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.233015344318488e-05,
+      "loss": 3.55,
+      "step": 1079296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.232176749567436e-05,
+      "loss": 3.5524,
+      "step": 1079808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.231338154816384e-05,
+      "loss": 3.553,
+      "step": 1080320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.230501197945705e-05,
+      "loss": 3.5621,
+      "step": 1080832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.229662603194653e-05,
+      "loss": 3.5487,
+      "step": 1081344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.228824008443601e-05,
+      "loss": 3.5453,
+      "step": 1081856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.227985413692549e-05,
+      "loss": 3.5566,
+      "step": 1082368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2271468189414975e-05,
+      "loss": 3.5392,
+      "step": 1082880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2263098620708184e-05,
+      "loss": 3.5446,
+      "step": 1083392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2254712673197664e-05,
+      "loss": 3.5529,
+      "step": 1083904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2246326725687144e-05,
+      "loss": 3.5522,
+      "step": 1084416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2237940778176624e-05,
+      "loss": 3.5504,
+      "step": 1084928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.222957120946983e-05,
+      "loss": 3.5569,
+      "step": 1085440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.222118526195931e-05,
+      "loss": 3.5493,
+      "step": 1085952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.221279931444879e-05,
+      "loss": 3.549,
+      "step": 1086464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2204429745742e-05,
+      "loss": 3.5493,
+      "step": 1086976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.219604379823148e-05,
+      "loss": 3.5368,
+      "step": 1087488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.218765785072096e-05,
+      "loss": 3.5479,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.217927190321044e-05,
+      "loss": 3.5386,
+      "step": 1088512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.217088595569993e-05,
+      "loss": 3.5447,
+      "step": 1089024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.216251638699314e-05,
+      "loss": 3.5444,
+      "step": 1089536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.215413043948262e-05,
+      "loss": 3.5587,
+      "step": 1090048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.21457444919721e-05,
+      "loss": 3.5474,
+      "step": 1090560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2137374923265307e-05,
+      "loss": 3.5514,
+      "step": 1091072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2128988975754787e-05,
+      "loss": 3.55,
+      "step": 1091584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2120603028244266e-05,
+      "loss": 3.5495,
+      "step": 1092096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2112217080733746e-05,
+      "loss": 3.5428,
+      "step": 1092608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2103847512026956e-05,
+      "loss": 3.5332,
+      "step": 1093120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2095461564516435e-05,
+      "loss": 3.5443,
+      "step": 1093632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2087075617005915e-05,
+      "loss": 3.5496,
+      "step": 1094144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2078689669495395e-05,
+      "loss": 3.5503,
+      "step": 1094656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.207032010078861e-05,
+      "loss": 3.5469,
+      "step": 1095168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.206193415327809e-05,
+      "loss": 3.534,
+      "step": 1095680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.205354820576757e-05,
+      "loss": 3.5434,
+      "step": 1096192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.204516225825705e-05,
+      "loss": 3.5327,
+      "step": 1096704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.203679268955026e-05,
+      "loss": 3.5625,
+      "step": 1097216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.202840674203974e-05,
+      "loss": 3.5367,
+      "step": 1097728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.202002079452922e-05,
+      "loss": 3.5574,
+      "step": 1098240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20116348470187e-05,
+      "loss": 3.5428,
+      "step": 1098752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.200326527831191e-05,
+      "loss": 3.5309,
+      "step": 1099264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.199487933080139e-05,
+      "loss": 3.5438,
+      "step": 1099776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.198649338329087e-05,
+      "loss": 3.5419,
+      "step": 1100288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.197810743578035e-05,
+      "loss": 3.5376,
+      "step": 1100800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1969737867073565e-05,
+      "loss": 3.5309,
+      "step": 1101312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1961351919563045e-05,
+      "loss": 3.5529,
+      "step": 1101824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1952965972052525e-05,
+      "loss": 3.5348,
+      "step": 1102336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1944580024542005e-05,
+      "loss": 3.5373,
+      "step": 1102848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1936210455835214e-05,
+      "loss": 3.528,
+      "step": 1103360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1927824508324694e-05,
+      "loss": 3.5441,
+      "step": 1103872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1919438560814174e-05,
+      "loss": 3.5475,
+      "step": 1104384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1911052613303654e-05,
+      "loss": 3.5421,
+      "step": 1104896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.190268304459686e-05,
+      "loss": 3.544,
+      "step": 1105408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.189429709708634e-05,
+      "loss": 3.546,
+      "step": 1105920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.188591114957582e-05,
+      "loss": 3.5513,
+      "step": 1106432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.18775252020653e-05,
+      "loss": 3.5407,
+      "step": 1106944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.186915563335852e-05,
+      "loss": 3.5443,
+      "step": 1107456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1860769685848e-05,
+      "loss": 3.5419,
+      "step": 1107968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185238373833748e-05,
+      "loss": 3.542,
+      "step": 1108480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.184399779082695e-05,
+      "loss": 3.5446,
+      "step": 1108992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.183562822212017e-05,
+      "loss": 3.546,
+      "step": 1109504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.182724227460965e-05,
+      "loss": 3.5471,
+      "step": 1110016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.181885632709913e-05,
+      "loss": 3.5535,
+      "step": 1110528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.18104703795886e-05,
+      "loss": 3.5433,
+      "step": 1111040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1802100810881817e-05,
+      "loss": 3.5304,
+      "step": 1111552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1793714863371296e-05,
+      "loss": 3.5375,
+      "step": 1112064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1785328915860776e-05,
+      "loss": 3.5481,
+      "step": 1112576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1776942968350256e-05,
+      "loss": 3.541,
+      "step": 1113088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.176857339964347e-05,
+      "loss": 3.5411,
+      "step": 1113600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.176018745213295e-05,
+      "loss": 3.5502,
+      "step": 1114112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1751801504622425e-05,
+      "loss": 3.5438,
+      "step": 1114624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1743415557111905e-05,
+      "loss": 3.539,
+      "step": 1115136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.173504598840512e-05,
+      "loss": 3.5412,
+      "step": 1115648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.17266600408946e-05,
+      "loss": 3.5514,
+      "step": 1116160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1718274093384074e-05,
+      "loss": 3.5474,
+      "step": 1116672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1709888145873554e-05,
+      "loss": 3.5566,
+      "step": 1117184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170151857716677e-05,
+      "loss": 3.5297,
+      "step": 1117696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.169313262965625e-05,
+      "loss": 3.5483,
+      "step": 1118208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.168474668214573e-05,
+      "loss": 3.5464,
+      "step": 1118720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.167636073463521e-05,
+      "loss": 3.5458,
+      "step": 1119232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1667991165928426e-05,
+      "loss": 3.5486,
+      "step": 1119744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.16596052184179e-05,
+      "loss": 3.5455,
+      "step": 1120256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.165121927090738e-05,
+      "loss": 3.5408,
+      "step": 1120768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.164283332339686e-05,
+      "loss": 3.5443,
+      "step": 1121280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1634463754690075e-05,
+      "loss": 3.5375,
+      "step": 1121792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.162607780717955e-05,
+      "loss": 3.5372,
+      "step": 1122304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.161769185966903e-05,
+      "loss": 3.5351,
+      "step": 1122816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.160930591215851e-05,
+      "loss": 3.5421,
+      "step": 1123328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.160093634345172e-05,
+      "loss": 3.5429,
+      "step": 1123840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1592550395941204e-05,
+      "loss": 3.5294,
+      "step": 1124352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1584164448430684e-05,
+      "loss": 3.532,
+      "step": 1124864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1575778500920164e-05,
+      "loss": 3.5456,
+      "step": 1125376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.156740893221337e-05,
+      "loss": 3.553,
+      "step": 1125888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.155902298470285e-05,
+      "loss": 3.5455,
+      "step": 1126400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.155063703719233e-05,
+      "loss": 3.5448,
+      "step": 1126912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154225108968181e-05,
+      "loss": 3.5497,
+      "step": 1127424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.153388152097502e-05,
+      "loss": 3.5461,
+      "step": 1127936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.15254955734645e-05,
+      "loss": 3.5388,
+      "step": 1128448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.151710962595398e-05,
+      "loss": 3.5397,
+      "step": 1128960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.150872367844346e-05,
+      "loss": 3.5426,
+      "step": 1129472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.150035410973667e-05,
+      "loss": 3.5485,
+      "step": 1129984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149196816222616e-05,
+      "loss": 3.5205,
+      "step": 1130496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.148358221471564e-05,
+      "loss": 3.543,
+      "step": 1131008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.147519626720512e-05,
+      "loss": 3.5312,
+      "step": 1131520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1466826698498326e-05,
+      "loss": 3.5418,
+      "step": 1132032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1458440750987806e-05,
+      "loss": 3.5457,
+      "step": 1132544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1450054803477286e-05,
+      "loss": 3.5434,
+      "step": 1133056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1441668855966766e-05,
+      "loss": 3.5304,
+      "step": 1133568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1433299287259975e-05,
+      "loss": 3.5368,
+      "step": 1134080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1424913339749455e-05,
+      "loss": 3.5513,
+      "step": 1134592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1416527392238935e-05,
+      "loss": 3.5415,
+      "step": 1135104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1408141444728415e-05,
+      "loss": 3.5514,
+      "step": 1135616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1399755497217895e-05,
+      "loss": 3.5371,
+      "step": 1136128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.139138592851111e-05,
+      "loss": 3.5431,
+      "step": 1136640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.138299998100059e-05,
+      "loss": 3.5492,
+      "step": 1137152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.137461403349007e-05,
+      "loss": 3.5452,
+      "step": 1137664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.136624446478328e-05,
+      "loss": 3.5391,
+      "step": 1138176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.135785851727276e-05,
+      "loss": 3.5393,
+      "step": 1138688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.134947256976224e-05,
+      "loss": 3.5513,
+      "step": 1139200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.134108662225172e-05,
+      "loss": 3.5538,
+      "step": 1139712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133271705354493e-05,
+      "loss": 3.5429,
+      "step": 1140224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.132433110603441e-05,
+      "loss": 3.5338,
+      "step": 1140736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.131594515852389e-05,
+      "loss": 3.543,
+      "step": 1141248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.130755921101337e-05,
+      "loss": 3.5347,
+      "step": 1141760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.129918964230658e-05,
+      "loss": 3.5294,
+      "step": 1142272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1290803694796065e-05,
+      "loss": 3.5535,
+      "step": 1142784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1282417747285545e-05,
+      "loss": 3.5366,
+      "step": 1143296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1274031799775025e-05,
+      "loss": 3.5469,
+      "step": 1143808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1265662231068234e-05,
+      "loss": 3.5373,
+      "step": 1144320
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.847018003463745,
+      "eval_runtime": 304.3354,
+      "eval_samples_per_second": 1253.85,
+      "eval_steps_per_second": 39.184,
+      "step": 1144800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1257276283557714e-05,
+      "loss": 3.526,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1248890336047194e-05,
+      "loss": 3.5351,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1240504388536674e-05,
+      "loss": 3.5386,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1232118441026154e-05,
+      "loss": 3.5471,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.122374887231936e-05,
+      "loss": 3.5499,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.121536292480884e-05,
+      "loss": 3.5431,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.120697697729832e-05,
+      "loss": 3.5367,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11985910297878e-05,
+      "loss": 3.5304,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.119022146108101e-05,
+      "loss": 3.5322,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11818355135705e-05,
+      "loss": 3.5425,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.117344956605998e-05,
+      "loss": 3.537,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.116506361854946e-05,
+      "loss": 3.555,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1156710428646397e-05,
+      "loss": 3.5532,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1148324481135877e-05,
+      "loss": 3.5258,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1139938533625357e-05,
+      "loss": 3.5274,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1131552586114836e-05,
+      "loss": 3.5254,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1123166638604316e-05,
+      "loss": 3.542,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1114780691093796e-05,
+      "loss": 3.5332,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1106394743583276e-05,
+      "loss": 3.5331,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1098008796072756e-05,
+      "loss": 3.5263,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1089639227365965e-05,
+      "loss": 3.5523,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.108125327985545e-05,
+      "loss": 3.5314,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107286733234493e-05,
+      "loss": 3.5382,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.106448138483441e-05,
+      "loss": 3.5391,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.105611181612762e-05,
+      "loss": 3.5491,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.10477258686171e-05,
+      "loss": 3.529,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.103933992110658e-05,
+      "loss": 3.5318,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.103095397359606e-05,
+      "loss": 3.5353,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.102256802608554e-05,
+      "loss": 3.5343,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.101419845737875e-05,
+      "loss": 3.5266,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.100581250986823e-05,
+      "loss": 3.5342,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.099742656235771e-05,
+      "loss": 3.5383,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.098905699365092e-05,
+      "loss": 3.5349,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0980671046140406e-05,
+      "loss": 3.5399,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0972285098629886e-05,
+      "loss": 3.533,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0963899151119366e-05,
+      "loss": 3.536,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0955529582412575e-05,
+      "loss": 3.533,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0947143634902055e-05,
+      "loss": 3.5259,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0938757687391535e-05,
+      "loss": 3.5325,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0930371739881015e-05,
+      "loss": 3.5195,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0922002171174224e-05,
+      "loss": 3.5324,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0913616223663704e-05,
+      "loss": 3.5286,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0905230276153184e-05,
+      "loss": 3.546,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0896844328642664e-05,
+      "loss": 3.5334,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0888458381132144e-05,
+      "loss": 3.5329,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.088008881242536e-05,
+      "loss": 3.5404,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087170286491484e-05,
+      "loss": 3.5294,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086331691740432e-05,
+      "loss": 3.5267,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.085494734869753e-05,
+      "loss": 3.519,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.084656140118701e-05,
+      "loss": 3.5334,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.083817545367649e-05,
+      "loss": 3.533,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.082978950616597e-05,
+      "loss": 3.5323,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.082140355865545e-05,
+      "loss": 3.5356,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.081303398994866e-05,
+      "loss": 3.5179,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.080464804243814e-05,
+      "loss": 3.5275,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.079626209492762e-05,
+      "loss": 3.5164,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.07878761474171e-05,
+      "loss": 3.5452,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.077950657871031e-05,
+      "loss": 3.521,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.077112063119979e-05,
+      "loss": 3.5454,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076273468368927e-05,
+      "loss": 3.531,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0754348736178746e-05,
+      "loss": 3.5131,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.074597916747196e-05,
+      "loss": 3.5287,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.073759321996144e-05,
+      "loss": 3.5283,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072920727245092e-05,
+      "loss": 3.5195,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0720821324940395e-05,
+      "loss": 3.516,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071245175623361e-05,
+      "loss": 3.5377,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.070406580872309e-05,
+      "loss": 3.5228,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.069567986121257e-05,
+      "loss": 3.5204,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.068729391370205e-05,
+      "loss": 3.5142,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.067892434499527e-05,
+      "loss": 3.5274,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.067053839748475e-05,
+      "loss": 3.5344,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066215244997422e-05,
+      "loss": 3.5276,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.06537665024637e-05,
+      "loss": 3.5265,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0645396933756916e-05,
+      "loss": 3.5311,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0637010986246396e-05,
+      "loss": 3.5355,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.062862503873587e-05,
+      "loss": 3.5243,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.062023909122535e-05,
+      "loss": 3.5304,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0611869522518565e-05,
+      "loss": 3.5302,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0603483575008045e-05,
+      "loss": 3.5246,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.059509762749752e-05,
+      "loss": 3.5299,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0586711679987005e-05,
+      "loss": 3.5299,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.057834211128022e-05,
+      "loss": 3.5323,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0569956163769694e-05,
+      "loss": 3.5401,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0561570216259174e-05,
+      "loss": 3.5252,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0553184268748654e-05,
+      "loss": 3.5175,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.054481470004187e-05,
+      "loss": 3.5227,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.053642875253134e-05,
+      "loss": 3.5354,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.052804280502082e-05,
+      "loss": 3.532,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.05196568575103e-05,
+      "loss": 3.5237,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051128728880352e-05,
+      "loss": 3.5338,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0502901341292995e-05,
+      "loss": 3.5327,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0494515393782475e-05,
+      "loss": 3.5221,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.048614582507569e-05,
+      "loss": 3.5254,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0477759877565164e-05,
+      "loss": 3.5402,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0469373930054644e-05,
+      "loss": 3.5266,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0460987982544127e-05,
+      "loss": 3.5393,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0452618413837343e-05,
+      "loss": 3.5201,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0444232466326816e-05,
+      "loss": 3.5333,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0435846518816296e-05,
+      "loss": 3.5335,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0427460571305776e-05,
+      "loss": 3.5276,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0419091002598992e-05,
+      "loss": 3.5369,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.041070505508847e-05,
+      "loss": 3.5297,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.040231910757795e-05,
+      "loss": 3.5253,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.039393316006743e-05,
+      "loss": 3.5235,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0385563591360638e-05,
+      "loss": 3.527,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0377177643850118e-05,
+      "loss": 3.5198,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0368791696339598e-05,
+      "loss": 3.5213,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.036040574882908e-05,
+      "loss": 3.5336,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.035203618012229e-05,
+      "loss": 3.5273,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.034365023261177e-05,
+      "loss": 3.5139,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.033526428510125e-05,
+      "loss": 3.5185,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.032687833759073e-05,
+      "loss": 3.5286,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0318508768883942e-05,
+      "loss": 3.5383,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0310122821373422e-05,
+      "loss": 3.5287,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0301736873862902e-05,
+      "loss": 3.5307,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0293350926352382e-05,
+      "loss": 3.5332,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.028498135764559e-05,
+      "loss": 3.5396,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.027659541013507e-05,
+      "loss": 3.5213,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.026820946262455e-05,
+      "loss": 3.5261,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0259823515114035e-05,
+      "loss": 3.5283,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0251453946407244e-05,
+      "loss": 3.531,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0243067998896724e-05,
+      "loss": 3.5107,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0234682051386204e-05,
+      "loss": 3.5243,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0226296103875684e-05,
+      "loss": 3.5195,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0217926535168896e-05,
+      "loss": 3.5238,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0209540587658376e-05,
+      "loss": 3.5263,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0201154640147856e-05,
+      "loss": 3.5375,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0192768692637336e-05,
+      "loss": 3.5175,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0184399123930545e-05,
+      "loss": 3.5215,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0176013176420025e-05,
+      "loss": 3.5308,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0167627228909505e-05,
+      "loss": 3.532,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0159241281398988e-05,
+      "loss": 3.5345,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0150871712692197e-05,
+      "loss": 3.5227,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0142485765181677e-05,
+      "loss": 3.5276,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0134099817671157e-05,
+      "loss": 3.5344,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0125713870160637e-05,
+      "loss": 3.5331,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.011734430145385e-05,
+      "loss": 3.5233,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.010895835394333e-05,
+      "loss": 3.5267,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.010057240643281e-05,
+      "loss": 3.5364,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.009218645892229e-05,
+      "loss": 3.5339,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.00838168902155e-05,
+      "loss": 3.5332,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.007543094270498e-05,
+      "loss": 3.5207,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.006704499519446e-05,
+      "loss": 3.5268,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0058659047683942e-05,
+      "loss": 3.5159,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.005028947897715e-05,
+      "loss": 3.5163,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.004190353146663e-05,
+      "loss": 3.5393,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.003351758395611e-05,
+      "loss": 3.5237,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.002513163644559e-05,
+      "loss": 3.5295,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0016762067738803e-05,
+      "loss": 3.5247,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.0008376120228283e-05,
+      "loss": 3.5199,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8478143215179443,
+      "eval_runtime": 305.5063,
+      "eval_samples_per_second": 1249.045,
+      "eval_steps_per_second": 39.034,
+      "step": 1221120
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 8.453713616814735e+17,
+  "trial_name": null,
+  "trial_params": null
+}