diff --git "a/checkpoint-1144800/trainer_state.json" "b/checkpoint-1144800/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1144800/trainer_state.json"
@@ -0,0 +1,13555 @@
+{
+  "best_metric": 3.8442249298095703,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/binding-domain/transformer/2/checkpoints/checkpoint-992160",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 1144800,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.9471,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8172,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1814,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9731,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8113,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7151,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6002,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5309,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4452,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.3986,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3468,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.307,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2562,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2055,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1703,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1294,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1003,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0789,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0494,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0175,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0052,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 4.9664,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9506,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9281,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9143,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.8875,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8709,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8495,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8307,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.816,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.7991,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7944,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.774,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 4.7677,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7598,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7426,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7332,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.7062,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7024,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.685,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6831,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6724,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6626,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6418,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6537,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6383,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6311,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6236,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.5919,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.5914,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.5969,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.588,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.5718,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5519,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.554,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5415,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5676,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 4.5254,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5373,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5336,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5079,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5149,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5022,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4862,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4863,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5059,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.4815,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4729,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4626,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.4641,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.477,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940482703000532e-05,
+      "loss": 4.472,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396441082494794e-05,
+      "loss": 4.4601,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388055134984274e-05,
+      "loss": 4.4654,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379669187473754e-05,
+      "loss": 4.4665,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 4.4515,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 4.4397,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 4.4309,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 4.4306,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 4.4249,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 4.4169,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 4.4216,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.4222,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 4.4155,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.3922,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.3917,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.4001,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 4.4025,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926231505873767e-05,
+      "loss": 4.3837,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925394549003088e-05,
+      "loss": 4.3862,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924555954252036e-05,
+      "loss": 4.3901,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923717359500984e-05,
+      "loss": 4.374,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922878764749932e-05,
+      "loss": 4.3705,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 4.3739,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.3656,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.3711,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.3617,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918687428875045e-05,
+      "loss": 4.3454,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917848834123993e-05,
+      "loss": 4.3561,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917010239372941e-05,
+      "loss": 4.3562,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.3493,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91533468775121e-05,
+      "loss": 4.3489,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914496093000158e-05,
+      "loss": 4.3447,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913657498249106e-05,
+      "loss": 4.3402,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912818903498054e-05,
+      "loss": 4.3379,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911981946627375e-05,
+      "loss": 4.3167,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111449897566966e-05,
+      "loss": 4.3204,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103063950056446e-05,
+      "loss": 4.3271,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094678002545926e-05,
+      "loss": 4.3227,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086292055035406e-05,
+      "loss": 4.3137,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9077906107524886e-05,
+      "loss": 4.3092,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9069536538818095e-05,
+      "loss": 4.3071,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9061150591307575e-05,
+      "loss": 4.3122,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052764643797055e-05,
+      "loss": 4.3043,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9044378696286535e-05,
+      "loss": 4.3065,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9035992748776015e-05,
+      "loss": 4.3003,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027606801265495e-05,
+      "loss": 4.3057,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.3024,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.2976,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900246533753767e-05,
+      "loss": 4.3005,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899409576883088e-05,
+      "loss": 4.2983,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898570982132036e-05,
+      "loss": 4.2868,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897732387380984e-05,
+      "loss": 4.2924,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896893792629932e-05,
+      "loss": 4.2803,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89605519787888e-05,
+      "loss": 4.2891,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895216603127827e-05,
+      "loss": 4.2666,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894378008376775e-05,
+      "loss": 4.2785,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893539413625723e-05,
+      "loss": 4.2635,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892704094635418e-05,
+      "loss": 4.2698,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891865499884366e-05,
+      "loss": 4.2545,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891026905133314e-05,
+      "loss": 4.284,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901883103822624e-05,
+      "loss": 4.2708,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934971563121e-05,
+      "loss": 4.258,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888511120880158e-05,
+      "loss": 4.2591,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887672526129106e-05,
+      "loss": 4.2625,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886833931378054e-05,
+      "loss": 4.2661,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859969745073747e-05,
+      "loss": 4.2652,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885160017636696e-05,
+      "loss": 4.251,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884321422885644e-05,
+      "loss": 4.2577,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883482828134592e-05,
+      "loss": 4.2461,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8826442333835395e-05,
+      "loss": 4.2575,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8818056386324875e-05,
+      "loss": 4.2427,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880967043881436e-05,
+      "loss": 4.2434,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880128449130384e-05,
+      "loss": 4.2415,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879291492259705e-05,
+      "loss": 4.232,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878452897508653e-05,
+      "loss": 4.2377,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877614302757601e-05,
+      "loss": 4.2376,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876775708006549e-05,
+      "loss": 4.2349,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875937113255497e-05,
+      "loss": 4.2362,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875100156384818e-05,
+      "loss": 4.2276,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.193670272827148,
+      "eval_runtime": 303.4498,
+      "eval_samples_per_second": 1257.51,
+      "eval_steps_per_second": 39.298,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874261561633766e-05,
+      "loss": 4.2126,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873422966882714e-05,
+      "loss": 4.209,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872584372131662e-05,
+      "loss": 4.2295,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871747415260983e-05,
+      "loss": 4.2115,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709088205099316e-05,
+      "loss": 4.2307,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700702257588796e-05,
+      "loss": 4.2,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692316310078276e-05,
+      "loss": 4.2087,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683930362567756e-05,
+      "loss": 4.1919,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675560793860965e-05,
+      "loss": 4.2089,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667174846350445e-05,
+      "loss": 4.2003,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658788898839925e-05,
+      "loss": 4.2064,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650402951329405e-05,
+      "loss": 4.203,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642033382622614e-05,
+      "loss": 4.19,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633647435112094e-05,
+      "loss": 4.1892,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8625261487601574e-05,
+      "loss": 4.1858,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8616875540091054e-05,
+      "loss": 4.1791,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608489592580534e-05,
+      "loss": 4.1878,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8600103645070014e-05,
+      "loss": 4.1845,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591717697559493e-05,
+      "loss": 4.179,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858334812885271e-05,
+      "loss": 4.1972,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857496218134219e-05,
+      "loss": 4.1731,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856657623383167e-05,
+      "loss": 4.1834,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855819028632115e-05,
+      "loss": 4.1781,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854980433881063e-05,
+      "loss": 4.1866,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854143477010384e-05,
+      "loss": 4.163,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853304882259332e-05,
+      "loss": 4.1714,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85246628750828e-05,
+      "loss": 4.163,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851627692757228e-05,
+      "loss": 4.1692,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850790735886549e-05,
+      "loss": 4.1524,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849952141135497e-05,
+      "loss": 4.1573,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849113546384445e-05,
+      "loss": 4.1657,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482749516333934e-05,
+      "loss": 4.1601,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847436356882341e-05,
+      "loss": 4.1596,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846599400011662e-05,
+      "loss": 4.1638,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84576080526061e-05,
+      "loss": 4.159,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844922210509558e-05,
+      "loss": 4.1575,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440836157585056e-05,
+      "loss": 4.145,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432450210074536e-05,
+      "loss": 4.1527,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842408064136775e-05,
+      "loss": 4.138,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415694693857225e-05,
+      "loss": 4.1467,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8407308746346705e-05,
+      "loss": 4.1476,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398922798836185e-05,
+      "loss": 4.1447,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83905532301294e-05,
+      "loss": 4.1338,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838216728261888e-05,
+      "loss": 4.1425,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837378133510836e-05,
+      "loss": 4.1475,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836539538759784e-05,
+      "loss": 4.1403,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8357025818891057e-05,
+      "loss": 4.1417,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834863987138053e-05,
+      "loss": 4.113,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834025392387001e-05,
+      "loss": 4.121,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833186797635949e-05,
+      "loss": 4.138,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83234984076527e-05,
+      "loss": 4.1294,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831511246014218e-05,
+      "loss": 4.1243,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830672651263166e-05,
+      "loss": 4.1094,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829834056512114e-05,
+      "loss": 4.1135,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289970996414354e-05,
+      "loss": 4.1066,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281585048903834e-05,
+      "loss": 4.1316,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273199101393314e-05,
+      "loss": 4.1078,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264813153882794e-05,
+      "loss": 4.119,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256443585176003e-05,
+      "loss": 4.1214,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248057637665483e-05,
+      "loss": 4.1034,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823967169015496e-05,
+      "loss": 4.1098,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823128574264444e-05,
+      "loss": 4.1011,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822289979513392e-05,
+      "loss": 4.0973,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821453022642713e-05,
+      "loss": 4.098,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820614427891661e-05,
+      "loss": 4.1141,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819775833140609e-05,
+      "loss": 4.1041,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818937238389558e-05,
+      "loss": 4.094,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8181019193992524e-05,
+      "loss": 4.0922,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172633246482004e-05,
+      "loss": 4.0914,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816424729897148e-05,
+      "loss": 4.1089,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815586135146096e-05,
+      "loss": 4.1106,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814747540395044e-05,
+      "loss": 4.1028,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813908945643992e-05,
+      "loss": 4.1123,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81307035089294e-05,
+      "loss": 4.1091,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812231756141888e-05,
+      "loss": 4.1061,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113947992712086e-05,
+      "loss": 4.0954,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105562045201566e-05,
+      "loss": 4.0924,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8097176097691046e-05,
+      "loss": 4.0914,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808879015018053e-05,
+      "loss": 4.0896,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808042058147374e-05,
+      "loss": 4.0852,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807203463396322e-05,
+      "loss": 4.0928,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80636486864527e-05,
+      "loss": 4.0994,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805526273894218e-05,
+      "loss": 4.0884,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804689317023539e-05,
+      "loss": 4.0756,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803850722272487e-05,
+      "loss": 4.0743,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803012127521435e-05,
+      "loss": 4.0816,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802175170650756e-05,
+      "loss": 4.0913,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801336575899704e-05,
+      "loss": 4.0769,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800497981148652e-05,
+      "loss": 4.077,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996593863976e-05,
+      "loss": 4.0834,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988207916465486e-05,
+      "loss": 4.0737,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979821968954966e-05,
+      "loss": 4.0719,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971436021444446e-05,
+      "loss": 4.0733,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7963050073933926e-05,
+      "loss": 4.0756,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954680505227135e-05,
+      "loss": 4.077,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946294557716615e-05,
+      "loss": 4.0763,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937908610206095e-05,
+      "loss": 4.0604,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929522662695575e-05,
+      "loss": 4.0734,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7921153093988784e-05,
+      "loss": 4.0684,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912767146478264e-05,
+      "loss": 4.0771,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7904381198967744e-05,
+      "loss": 4.0689,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895995251457224e-05,
+      "loss": 4.0686,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788762568275043e-05,
+      "loss": 4.0664,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787923973523992e-05,
+      "loss": 4.0696,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78708537877294e-05,
+      "loss": 4.0469,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786246784021888e-05,
+      "loss": 4.0523,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785409827151209e-05,
+      "loss": 4.0611,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784571232400157e-05,
+      "loss": 4.0573,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783732637649105e-05,
+      "loss": 4.0517,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782894042898053e-05,
+      "loss": 4.0499,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782055448147001e-05,
+      "loss": 4.0495,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781216853395948e-05,
+      "loss": 4.0547,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780378258644896e-05,
+      "loss": 4.0499,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779541301774218e-05,
+      "loss": 4.0521,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778702707023166e-05,
+      "loss": 4.0511,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777864112272114e-05,
+      "loss": 4.0573,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777025517521062e-05,
+      "loss": 4.0546,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761885606503834e-05,
+      "loss": 4.0487,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775349965899331e-05,
+      "loss": 4.0574,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774513009028652e-05,
+      "loss": 4.0564,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7736744142776e-05,
+      "loss": 4.0412,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772835819526548e-05,
+      "loss": 4.0521,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719972247754956e-05,
+      "loss": 4.0408,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7711586300244436e-05,
+      "loss": 4.0533,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703200352733916e-05,
+      "loss": 4.0278,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694814405223396e-05,
+      "loss": 4.0462,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686428457712876e-05,
+      "loss": 4.0303,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767805888900609e-05,
+      "loss": 4.0347,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766967294149557e-05,
+      "loss": 4.0245,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766128699398505e-05,
+      "loss": 4.0552,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765290104647453e-05,
+      "loss": 4.0443,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7644547856571476e-05,
+      "loss": 4.0357,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7636161909060956e-05,
+      "loss": 4.031,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762777596155043e-05,
+      "loss": 4.0399,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761939001403991e-05,
+      "loss": 4.0441,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761100406652939e-05,
+      "loss": 4.0413,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760261811901887e-05,
+      "loss": 4.0345,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759423217150835e-05,
+      "loss": 4.0368,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758584622399783e-05,
+      "loss": 4.0313,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577476655291045e-05,
+      "loss": 4.0453,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569090707780525e-05,
+      "loss": 4.0263,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560704760270005e-05,
+      "loss": 4.0304,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552318812759485e-05,
+      "loss": 4.0278,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543949244052694e-05,
+      "loss": 4.0224,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75355796753459e-05,
+      "loss": 4.0309,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752719372783538e-05,
+      "loss": 4.0258,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751880778032486e-05,
+      "loss": 4.0306,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751042183281434e-05,
+      "loss": 4.0269,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750205226410755e-05,
+      "loss": 4.0221,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.025816917419434,
+      "eval_runtime": 303.3157,
+      "eval_samples_per_second": 1258.065,
+      "eval_steps_per_second": 39.315,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749366631659703e-05,
+      "loss": 4.0104,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748528036908652e-05,
+      "loss": 4.008,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476894421576e-05,
+      "loss": 4.0255,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746850847406548e-05,
+      "loss": 4.012,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746012252655496e-05,
+      "loss": 4.0361,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745173657904444e-05,
+      "loss": 4.0032,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744335063153392e-05,
+      "loss": 4.0142,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74349646840234e-05,
+      "loss": 3.9996,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742657873651288e-05,
+      "loss": 4.0168,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741819278900236e-05,
+      "loss": 4.0104,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740980684149184e-05,
+      "loss": 4.0148,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740143727278505e-05,
+      "loss": 4.0112,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7393084082881986e-05,
+      "loss": 4.0006,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738469813537147e-05,
+      "loss": 4.0058,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737631218786095e-05,
+      "loss": 4.0009,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736792624035043e-05,
+      "loss": 3.9962,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735954029283991e-05,
+      "loss": 3.9989,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735115434532939e-05,
+      "loss": 4.0031,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734276839781887e-05,
+      "loss": 3.9956,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733438245030835e-05,
+      "loss": 4.0186,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732601288160156e-05,
+      "loss": 3.9946,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731762693409104e-05,
+      "loss": 4.0023,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730924098658052e-05,
+      "loss": 3.9999,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730085503907e-05,
+      "loss": 4.0071,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729246909155948e-05,
+      "loss": 3.9882,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728409952285269e-05,
+      "loss": 3.9995,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727571357534217e-05,
+      "loss": 3.9898,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726732762783165e-05,
+      "loss": 3.9936,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725894168032114e-05,
+      "loss": 3.9816,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725055573281062e-05,
+      "loss": 3.9846,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724216978530009e-05,
+      "loss": 3.9933,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233800216593306e-05,
+      "loss": 3.9943,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225414269082786e-05,
+      "loss": 3.9929,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7217028321572266e-05,
+      "loss": 3.9954,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720864237406174e-05,
+      "loss": 3.9893,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720025642655122e-05,
+      "loss": 3.993,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191886857844435e-05,
+      "loss": 3.98,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718350091033391e-05,
+      "loss": 3.9924,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717511496282339e-05,
+      "loss": 3.9728,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7166729015312875e-05,
+      "loss": 3.9838,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715835944660609e-05,
+      "loss": 3.9814,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149973499095564e-05,
+      "loss": 3.9867,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141587551585044e-05,
+      "loss": 3.981,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133201604074524e-05,
+      "loss": 3.9785,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124815656564004e-05,
+      "loss": 3.988,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711644608785721e-05,
+      "loss": 3.985,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710806014034669e-05,
+      "loss": 3.9824,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709967419283617e-05,
+      "loss": 3.9655,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709128824532565e-05,
+      "loss": 3.9633,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708290229781513e-05,
+      "loss": 3.9786,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707451635030461e-05,
+      "loss": 3.9756,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706614678159783e-05,
+      "loss": 3.9727,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705776083408731e-05,
+      "loss": 3.956,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704937488657679e-05,
+      "loss": 3.9609,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704098893906627e-05,
+      "loss": 3.9608,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703260299155575e-05,
+      "loss": 3.9742,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702421704404523e-05,
+      "loss": 3.9586,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701583109653471e-05,
+      "loss": 3.9708,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700744514902419e-05,
+      "loss": 3.9689,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69990755803174e-05,
+      "loss": 3.9576,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990706011610606e-05,
+      "loss": 3.9633,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982320064100086e-05,
+      "loss": 3.9571,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973934116589566e-05,
+      "loss": 3.9494,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965548169079046e-05,
+      "loss": 3.9584,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695717860037226e-05,
+      "loss": 3.9627,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694879265286174e-05,
+      "loss": 3.9677,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694040670535122e-05,
+      "loss": 3.9445,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69320207578407e-05,
+      "loss": 3.9506,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692363481033018e-05,
+      "loss": 3.9477,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691524886281966e-05,
+      "loss": 3.9745,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690686291530914e-05,
+      "loss": 3.9606,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689849334660235e-05,
+      "loss": 3.9622,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689010739909183e-05,
+      "loss": 3.9749,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688172145158131e-05,
+      "loss": 3.9698,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687333550407079e-05,
+      "loss": 3.9667,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686494955656027e-05,
+      "loss": 3.9577,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685656360904975e-05,
+      "loss": 3.9588,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684817766153923e-05,
+      "loss": 3.9553,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683979171402871e-05,
+      "loss": 3.9529,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831438524125655e-05,
+      "loss": 3.9519,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6823052576615135e-05,
+      "loss": 3.9563,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6814666629104615e-05,
+      "loss": 3.9651,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806280681594095e-05,
+      "loss": 3.9529,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679789473408357e-05,
+      "loss": 3.9457,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789525165376784e-05,
+      "loss": 3.939,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6781139217866264e-05,
+      "loss": 3.9517,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772753270355744e-05,
+      "loss": 3.9571,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676436732284522e-05,
+      "loss": 3.9479,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755981375334704e-05,
+      "loss": 3.9445,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747595427824184e-05,
+      "loss": 3.9572,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739209480313664e-05,
+      "loss": 3.9357,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673083991160687e-05,
+      "loss": 3.9472,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672245396409635e-05,
+      "loss": 3.9445,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671406801658583e-05,
+      "loss": 3.9524,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670568206907531e-05,
+      "loss": 3.9501,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669731250036852e-05,
+      "loss": 3.9455,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688926552858e-05,
+      "loss": 3.9346,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668054060534748e-05,
+      "loss": 3.9475,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667215465783696e-05,
+      "loss": 3.9398,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666376871032644e-05,
+      "loss": 3.9546,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665539914161966e-05,
+      "loss": 3.9432,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664701319410914e-05,
+      "loss": 3.9447,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663862724659862e-05,
+      "loss": 3.9369,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66302412990881e-05,
+      "loss": 3.9507,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662187173038131e-05,
+      "loss": 3.9221,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661348578287079e-05,
+      "loss": 3.9308,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660509983536027e-05,
+      "loss": 3.938,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659671388784975e-05,
+      "loss": 3.9372,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658832794033923e-05,
+      "loss": 3.9259,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579958371632436e-05,
+      "loss": 3.9309,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571572424121916e-05,
+      "loss": 3.9257,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563186476611396e-05,
+      "loss": 3.935,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554800529100876e-05,
+      "loss": 3.9288,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654643096039409e-05,
+      "loss": 3.9275,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653804501288357e-05,
+      "loss": 3.9368,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652965906537305e-05,
+      "loss": 3.9393,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652127311786253e-05,
+      "loss": 3.9387,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651290354915574e-05,
+      "loss": 3.9294,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650451760164522e-05,
+      "loss": 3.9401,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64961316541347e-05,
+      "loss": 3.9332,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648774570662418e-05,
+      "loss": 3.9254,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647937613791739e-05,
+      "loss": 3.9401,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647099019040687e-05,
+      "loss": 3.9206,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646260424289635e-05,
+      "loss": 3.9384,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645421829538583e-05,
+      "loss": 3.9168,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445848726679045e-05,
+      "loss": 3.9318,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437462779168525e-05,
+      "loss": 3.9145,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429076831658005e-05,
+      "loss": 3.9194,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420690884147485e-05,
+      "loss": 3.9148,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412321315440694e-05,
+      "loss": 3.9392,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403935367930174e-05,
+      "loss": 3.9302,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395549420419654e-05,
+      "loss": 3.9266,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387163472909134e-05,
+      "loss": 3.9201,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637879390420234e-05,
+      "loss": 3.9247,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637040795669182e-05,
+      "loss": 3.9369,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63620220091813e-05,
+      "loss": 3.9258,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635363606167078e-05,
+      "loss": 3.9232,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345266492964e-05,
+      "loss": 3.9316,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633688054545348e-05,
+      "loss": 3.9227,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632849459794296e-05,
+      "loss": 3.9296,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632010865043244e-05,
+      "loss": 3.917,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631173908172565e-05,
+      "loss": 3.9222,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630335313421513e-05,
+      "loss": 3.9178,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629496718670461e-05,
+      "loss": 3.9182,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628658123919409e-05,
+      "loss": 3.9178,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62782116704873e-05,
+      "loss": 3.9237,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626982572297678e-05,
+      "loss": 3.9168,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626143977546626e-05,
+      "loss": 3.9198,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6253053827955737e-05,
+      "loss": 3.917,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9517478942871094,
+      "eval_runtime": 317.1615,
+      "eval_samples_per_second": 1203.144,
+      "eval_steps_per_second": 37.599,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244667880445217e-05,
+      "loss": 3.9094,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623629831173843e-05,
+      "loss": 3.9033,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622791236422791e-05,
+      "loss": 3.9194,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621952641671739e-05,
+      "loss": 3.9056,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621114046920687e-05,
+      "loss": 3.9333,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620277090050008e-05,
+      "loss": 3.899,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619438495298956e-05,
+      "loss": 3.9128,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618599900547904e-05,
+      "loss": 3.8954,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617761305796852e-05,
+      "loss": 3.9121,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6169227110458e-05,
+      "loss": 3.9105,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616085754175121e-05,
+      "loss": 3.9065,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615247159424069e-05,
+      "loss": 3.9097,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6144102025533906e-05,
+      "loss": 3.897,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135716078023386e-05,
+      "loss": 3.9065,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127330130512866e-05,
+      "loss": 3.8984,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6118944183002346e-05,
+      "loss": 3.8973,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110558235491826e-05,
+      "loss": 3.8978,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61021722879813e-05,
+      "loss": 3.9016,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609378634047078e-05,
+      "loss": 3.8979,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608540039296026e-05,
+      "loss": 3.9134,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6077030824253475e-05,
+      "loss": 3.897,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606864487674295e-05,
+      "loss": 3.9022,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606025892923243e-05,
+      "loss": 3.9007,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605187298172191e-05,
+      "loss": 3.9105,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043503413015124e-05,
+      "loss": 3.8895,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6035117465504604e-05,
+      "loss": 3.9047,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026731517994084e-05,
+      "loss": 3.892,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018345570483564e-05,
+      "loss": 3.8969,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600997600177677e-05,
+      "loss": 3.882,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600159005426625e-05,
+      "loss": 3.8938,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599320410675573e-05,
+      "loss": 3.8946,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598481815924521e-05,
+      "loss": 3.899,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597644859053842e-05,
+      "loss": 3.8971,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59680626430279e-05,
+      "loss": 3.9012,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595967669551738e-05,
+      "loss": 3.8906,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595129074800686e-05,
+      "loss": 3.8989,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594292117930008e-05,
+      "loss": 3.8885,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593453523178956e-05,
+      "loss": 3.8972,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592614928427904e-05,
+      "loss": 3.8797,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591776333676852e-05,
+      "loss": 3.891,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5909393768061727e-05,
+      "loss": 3.8904,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5901007820551206e-05,
+      "loss": 3.8963,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5892621873040686e-05,
+      "loss": 3.8877,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5884235925530166e-05,
+      "loss": 3.888,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5875866356823375e-05,
+      "loss": 3.8949,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867480409312855e-05,
+      "loss": 3.8921,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5859094461802335e-05,
+      "loss": 3.8911,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5850708514291815e-05,
+      "loss": 3.8807,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584233894558503e-05,
+      "loss": 3.8652,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583395299807451e-05,
+      "loss": 3.8882,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582556705056399e-05,
+      "loss": 3.8842,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581718110305347e-05,
+      "loss": 3.88,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580881153434668e-05,
+      "loss": 3.874,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580042558683616e-05,
+      "loss": 3.8708,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579203963932564e-05,
+      "loss": 3.8675,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578365369181512e-05,
+      "loss": 3.8844,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577528412310833e-05,
+      "loss": 3.8708,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576689817559781e-05,
+      "loss": 3.8777,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575851222808729e-05,
+      "loss": 3.8851,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575012628057677e-05,
+      "loss": 3.8708,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741756711869985e-05,
+      "loss": 3.8715,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733370764359465e-05,
+      "loss": 3.8756,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724984816848945e-05,
+      "loss": 3.8572,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716598869338425e-05,
+      "loss": 3.8713,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5708229300631634e-05,
+      "loss": 3.8795,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699843353121114e-05,
+      "loss": 3.8773,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691457405610594e-05,
+      "loss": 3.8592,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5683071458100074e-05,
+      "loss": 3.8659,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567470188939328e-05,
+      "loss": 3.8622,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566631594188276e-05,
+      "loss": 3.8824,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565792999437224e-05,
+      "loss": 3.8751,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564954404686172e-05,
+      "loss": 3.8773,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564117447815494e-05,
+      "loss": 3.8875,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563278853064442e-05,
+      "loss": 3.886,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56244025831339e-05,
+      "loss": 3.8868,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561601663562338e-05,
+      "loss": 3.8662,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560764706691659e-05,
+      "loss": 3.8762,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559926111940607e-05,
+      "loss": 3.8741,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559087517189555e-05,
+      "loss": 3.8681,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558248922438503e-05,
+      "loss": 3.8727,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5574119655678236e-05,
+      "loss": 3.8703,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5565733708167716e-05,
+      "loss": 3.8788,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557347760657196e-05,
+      "loss": 3.8718,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548961813146676e-05,
+      "loss": 3.8557,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554059224443989e-05,
+      "loss": 3.8592,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553220629692937e-05,
+      "loss": 3.8697,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552382034941885e-05,
+      "loss": 3.8725,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551543440190833e-05,
+      "loss": 3.8639,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550704845439781e-05,
+      "loss": 3.8609,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549867888569102e-05,
+      "loss": 3.8727,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54902929381805e-05,
+      "loss": 3.8577,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548190699066998e-05,
+      "loss": 3.8614,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547352104315946e-05,
+      "loss": 3.866,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546515147445267e-05,
+      "loss": 3.8717,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545676552694215e-05,
+      "loss": 3.8715,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544837957943163e-05,
+      "loss": 3.8628,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543999363192111e-05,
+      "loss": 3.8499,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431624063214326e-05,
+      "loss": 3.87,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5423238115703806e-05,
+      "loss": 3.8563,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414852168193286e-05,
+      "loss": 3.8767,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406466220682766e-05,
+      "loss": 3.8661,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5398096651975975e-05,
+      "loss": 3.865,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389710704465455e-05,
+      "loss": 3.8568,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5381324756954935e-05,
+      "loss": 3.8704,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372938809444415e-05,
+      "loss": 3.8449,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364569240737624e-05,
+      "loss": 3.8539,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5356183293227104e-05,
+      "loss": 3.8533,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347797345716584e-05,
+      "loss": 3.8592,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53394277770098e-05,
+      "loss": 3.8466,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533104182949928e-05,
+      "loss": 3.8549,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532265588198876e-05,
+      "loss": 3.8519,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531426993447824e-05,
+      "loss": 3.8506,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530590036577145e-05,
+      "loss": 3.8554,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529751441826093e-05,
+      "loss": 3.848,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528912847075041e-05,
+      "loss": 3.8629,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528074252323989e-05,
+      "loss": 3.8611,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52723729545331e-05,
+      "loss": 3.8609,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526398700702258e-05,
+      "loss": 3.8542,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525560105951206e-05,
+      "loss": 3.8619,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524721511200154e-05,
+      "loss": 3.8614,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523884554329475e-05,
+      "loss": 3.8475,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523045959578423e-05,
+      "loss": 3.8613,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522207364827371e-05,
+      "loss": 3.847,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521368770076319e-05,
+      "loss": 3.8609,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52053181320564e-05,
+      "loss": 3.8441,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519693218454588e-05,
+      "loss": 3.8547,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518854623703536e-05,
+      "loss": 3.8376,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518016028952484e-05,
+      "loss": 3.8445,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517179072081805e-05,
+      "loss": 3.8419,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516340477330753e-05,
+      "loss": 3.8597,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515501882579701e-05,
+      "loss": 3.8531,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514663287828649e-05,
+      "loss": 3.8566,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513826330957971e-05,
+      "loss": 3.8464,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512987736206919e-05,
+      "loss": 3.8524,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512149141455867e-05,
+      "loss": 3.862,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511310546704815e-05,
+      "loss": 3.8517,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104735898341356e-05,
+      "loss": 3.8503,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096349950830836e-05,
+      "loss": 3.8581,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087964003320316e-05,
+      "loss": 3.8488,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079578055809796e-05,
+      "loss": 3.8587,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5071208487103005e-05,
+      "loss": 3.8409,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062822539592485e-05,
+      "loss": 3.8531,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5054436592081965e-05,
+      "loss": 3.8455,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5046050644571445e-05,
+      "loss": 3.8428,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503768107586466e-05,
+      "loss": 3.8487,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502929512835414e-05,
+      "loss": 3.8463,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502090918084362e-05,
+      "loss": 3.8466,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012523233333094e-05,
+      "loss": 3.8443,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500415366462631e-05,
+      "loss": 3.8491,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9105277061462402,
+      "eval_runtime": 316.9798,
+      "eval_samples_per_second": 1203.834,
+      "eval_steps_per_second": 37.621,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499576771711579e-05,
+      "loss": 3.8344,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498738176960527e-05,
+      "loss": 3.832,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497899582209474e-05,
+      "loss": 3.8473,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497060987458422e-05,
+      "loss": 3.8368,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49622239270737e-05,
+      "loss": 3.8602,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495383797956318e-05,
+      "loss": 3.8354,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494545203205266e-05,
+      "loss": 3.8371,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493706608454214e-05,
+      "loss": 3.824,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492869651583536e-05,
+      "loss": 3.8407,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492031056832484e-05,
+      "loss": 3.8445,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491192462081432e-05,
+      "loss": 3.8347,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49035386733038e-05,
+      "loss": 3.8368,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489518548340074e-05,
+      "loss": 3.8299,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886799535890216e-05,
+      "loss": 3.8344,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878413588379696e-05,
+      "loss": 3.8323,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4870027640869176e-05,
+      "loss": 3.8266,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861641693358656e-05,
+      "loss": 3.8295,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4853255745848136e-05,
+      "loss": 3.836,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844869798337616e-05,
+      "loss": 3.8281,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4836483850827096e-05,
+      "loss": 3.8422,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482811428212031e-05,
+      "loss": 3.8319,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481972833460979e-05,
+      "loss": 3.8365,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481134238709927e-05,
+      "loss": 3.831,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480295643958875e-05,
+      "loss": 3.8409,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479458687088196e-05,
+      "loss": 3.8211,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478620092337144e-05,
+      "loss": 3.8342,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477781497586092e-05,
+      "loss": 3.8284,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47694290283504e-05,
+      "loss": 3.8247,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476105945964361e-05,
+      "loss": 3.8168,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475267351213309e-05,
+      "loss": 3.8242,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474428756462257e-05,
+      "loss": 3.8332,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473590161711205e-05,
+      "loss": 3.8315,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4727532048405266e-05,
+      "loss": 3.8328,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4719146100894746e-05,
+      "loss": 3.8325,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710760153384226e-05,
+      "loss": 3.8232,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4702374205873705e-05,
+      "loss": 3.8316,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4694004637166915e-05,
+      "loss": 3.8263,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685618689656395e-05,
+      "loss": 3.8289,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4677232742145874e-05,
+      "loss": 3.8159,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668846794635354e-05,
+      "loss": 3.8254,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660477225928564e-05,
+      "loss": 3.8238,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4652091278418043e-05,
+      "loss": 3.831,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643705330907523e-05,
+      "loss": 3.8245,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4635319383397003e-05,
+      "loss": 3.8231,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462694981469022e-05,
+      "loss": 3.834,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46185638671797e-05,
+      "loss": 3.8246,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461017791966918e-05,
+      "loss": 3.8248,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460179197215866e-05,
+      "loss": 3.8158,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45934387822556e-05,
+      "loss": 3.801,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458505283474508e-05,
+      "loss": 3.826,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457666688723456e-05,
+      "loss": 3.8209,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456828093972404e-05,
+      "loss": 3.8183,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455989499221352e-05,
+      "loss": 3.8042,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551525423506726e-05,
+      "loss": 3.8101,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4543139475996206e-05,
+      "loss": 3.7992,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453475352848569e-05,
+      "loss": 3.8209,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452636758097517e-05,
+      "loss": 3.809,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451798163346465e-05,
+      "loss": 3.812,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450959568595413e-05,
+      "loss": 3.8241,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450120973844361e-05,
+      "loss": 3.8075,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449284016973682e-05,
+      "loss": 3.8116,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44844542222263e-05,
+      "loss": 3.8155,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447606827471578e-05,
+      "loss": 3.7964,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446768232720526e-05,
+      "loss": 3.8103,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445929637969474e-05,
+      "loss": 3.8135,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445091043218422e-05,
+      "loss": 3.8168,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442524484673695e-05,
+      "loss": 3.7988,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443415491596691e-05,
+      "loss": 3.8048,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442576896845639e-05,
+      "loss": 3.7983,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441738302094588e-05,
+      "loss": 3.8225,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440899707343535e-05,
+      "loss": 3.8147,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440061112592483e-05,
+      "loss": 3.8117,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439222517841431e-05,
+      "loss": 3.8279,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438383923090379e-05,
+      "loss": 3.8198,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437545328339327e-05,
+      "loss": 3.833,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436708371468648e-05,
+      "loss": 3.807,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435869776717596e-05,
+      "loss": 3.8125,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435031181966544e-05,
+      "loss": 3.814,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434192587215492e-05,
+      "loss": 3.8043,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433355630344813e-05,
+      "loss": 3.8145,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4325170355937615e-05,
+      "loss": 3.8091,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316784408427095e-05,
+      "loss": 3.8183,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308398460916575e-05,
+      "loss": 3.809,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4300028892209784e-05,
+      "loss": 3.7983,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291642944699264e-05,
+      "loss": 3.807,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4283256997188744e-05,
+      "loss": 3.8015,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427488742848195e-05,
+      "loss": 3.8146,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426650148097143e-05,
+      "loss": 3.8088,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425811553346091e-05,
+      "loss": 3.8003,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424972958595039e-05,
+      "loss": 3.8135,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424134363843987e-05,
+      "loss": 3.8041,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423295769092935e-05,
+      "loss": 3.7963,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422457174341883e-05,
+      "loss": 3.8053,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421618579590831e-05,
+      "loss": 3.8145,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420781622720153e-05,
+      "loss": 3.8136,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419943027969101e-05,
+      "loss": 3.8075,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419104433218049e-05,
+      "loss": 3.7901,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418265838466997e-05,
+      "loss": 3.8095,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417427243715945e-05,
+      "loss": 3.7998,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416588648964893e-05,
+      "loss": 3.8144,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415750054213841e-05,
+      "loss": 3.811,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414911459462788e-05,
+      "loss": 3.8064,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41407450259211e-05,
+      "loss": 3.7983,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413235907841058e-05,
+      "loss": 3.8163,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412397313090006e-05,
+      "loss": 3.7818,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411558718338953e-05,
+      "loss": 3.7993,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4107217614682747e-05,
+      "loss": 3.794,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409883166717223e-05,
+      "loss": 3.8062,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409044571966171e-05,
+      "loss": 3.7865,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4082059772151186e-05,
+      "loss": 3.795,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40736902034444e-05,
+      "loss": 3.7961,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406530425593388e-05,
+      "loss": 3.7934,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056918308423355e-05,
+      "loss": 3.7966,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404854873971657e-05,
+      "loss": 3.7955,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404016279220605e-05,
+      "loss": 3.8043,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403177684469553e-05,
+      "loss": 3.8052,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4023390897185004e-05,
+      "loss": 3.8034,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401502132847822e-05,
+      "loss": 3.796,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40066353809677e-05,
+      "loss": 3.8067,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399824943345719e-05,
+      "loss": 3.8007,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398986348594666e-05,
+      "loss": 3.7919,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981493917239876e-05,
+      "loss": 3.8061,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3973107969729356e-05,
+      "loss": 3.7887,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396472202221883e-05,
+      "loss": 3.8023,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395633607470831e-05,
+      "loss": 3.789,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947966506001525e-05,
+      "loss": 3.8003,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3939580558491005e-05,
+      "loss": 3.7867,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393119461098048e-05,
+      "loss": 3.7876,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392280866346996e-05,
+      "loss": 3.7902,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3914439094763174e-05,
+      "loss": 3.8029,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3906053147252654e-05,
+      "loss": 3.7965,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389768357854587e-05,
+      "loss": 3.8001,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388929763103535e-05,
+      "loss": 3.7894,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388091168352483e-05,
+      "loss": 3.8008,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38725257360143e-05,
+      "loss": 3.8067,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386413978850378e-05,
+      "loss": 3.7978,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385575384099326e-05,
+      "loss": 3.7949,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384736789348274e-05,
+      "loss": 3.8019,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383898194597222e-05,
+      "loss": 3.7946,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383061237726543e-05,
+      "loss": 3.8024,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382222642975491e-05,
+      "loss": 3.7881,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381384048224439e-05,
+      "loss": 3.7974,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380545453473388e-05,
+      "loss": 3.7903,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379708496602709e-05,
+      "loss": 3.7916,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378869901851657e-05,
+      "loss": 3.7939,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378031307100605e-05,
+      "loss": 3.7922,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377192712349553e-05,
+      "loss": 3.7945,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763557554788736e-05,
+      "loss": 3.7843,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3755171607278216e-05,
+      "loss": 3.7965,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.886021614074707,
+      "eval_runtime": 346.7835,
+      "eval_samples_per_second": 1100.372,
+      "eval_steps_per_second": 34.387,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3746785659767696e-05,
+      "loss": 3.7855,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3738399712257176e-05,
+      "loss": 3.7774,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3730030143550385e-05,
+      "loss": 3.7944,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3721644196039865e-05,
+      "loss": 3.7849,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3713258248529345e-05,
+      "loss": 3.8034,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370487230101883e-05,
+      "loss": 3.7858,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369650273231204e-05,
+      "loss": 3.7781,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368811678480152e-05,
+      "loss": 3.7783,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3679730837291e-05,
+      "loss": 3.783,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367134488978048e-05,
+      "loss": 3.7903,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366297532107369e-05,
+      "loss": 3.7828,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365458937356317e-05,
+      "loss": 3.7847,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364621980485638e-05,
+      "loss": 3.7782,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363783385734586e-05,
+      "loss": 3.7791,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362944790983534e-05,
+      "loss": 3.7835,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362106196232482e-05,
+      "loss": 3.7724,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36126760148143e-05,
+      "loss": 3.776,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3604290067303786e-05,
+      "loss": 3.7818,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3595904119793266e-05,
+      "loss": 3.7787,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3587518172282746e-05,
+      "loss": 3.7886,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3579164982379684e-05,
+      "loss": 3.7839,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3570779034869164e-05,
+      "loss": 3.784,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3562393087358644e-05,
+      "loss": 3.7829,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3554007139848124e-05,
+      "loss": 3.785,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3545621192337604e-05,
+      "loss": 3.7687,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3537235244827084e-05,
+      "loss": 3.7884,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3528849297316564e-05,
+      "loss": 3.7718,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3520463349806044e-05,
+      "loss": 3.7744,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351209378109925e-05,
+      "loss": 3.7686,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350370783358874e-05,
+      "loss": 3.7741,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349532188607822e-05,
+      "loss": 3.7821,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348695231737143e-05,
+      "loss": 3.7776,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347856636986091e-05,
+      "loss": 3.7856,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347018042235039e-05,
+      "loss": 3.785,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346179447483987e-05,
+      "loss": 3.7716,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345342490613308e-05,
+      "loss": 3.7804,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344503895862256e-05,
+      "loss": 3.7774,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343665301111204e-05,
+      "loss": 3.7766,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342826706360152e-05,
+      "loss": 3.7693,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419897494894726e-05,
+      "loss": 3.7732,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3411511547384206e-05,
+      "loss": 3.7785,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340314197867742e-05,
+      "loss": 3.7839,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33947560311669e-05,
+      "loss": 3.7726,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338637008365638e-05,
+      "loss": 3.7746,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337798413614586e-05,
+      "loss": 3.7812,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336959818863534e-05,
+      "loss": 3.7767,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336121224112482e-05,
+      "loss": 3.7767,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33528262936143e-05,
+      "loss": 3.7687,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334445672490751e-05,
+      "loss": 3.7472,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333607077739699e-05,
+      "loss": 3.7818,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332768482988647e-05,
+      "loss": 3.7724,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331929888237595e-05,
+      "loss": 3.772,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331092931366916e-05,
+      "loss": 3.7611,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330254336615864e-05,
+      "loss": 3.7555,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329415741864813e-05,
+      "loss": 3.7525,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328577147113761e-05,
+      "loss": 3.7685,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3277385523627087e-05,
+      "loss": 3.7621,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326899957611656e-05,
+      "loss": 3.7666,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326061362860604e-05,
+      "loss": 3.7727,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3252244059899256e-05,
+      "loss": 3.7593,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3243874491192465e-05,
+      "loss": 3.7638,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3235488543681945e-05,
+      "loss": 3.7693,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3227102596171425e-05,
+      "loss": 3.7488,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3218716648660905e-05,
+      "loss": 3.7595,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321033070115038e-05,
+      "loss": 3.7659,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201944753639864e-05,
+      "loss": 3.7683,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193558806129344e-05,
+      "loss": 3.7482,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318518923742256e-05,
+      "loss": 3.7571,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176803289912033e-05,
+      "loss": 3.7502,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3168417342401513e-05,
+      "loss": 3.7717,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3160031394890993e-05,
+      "loss": 3.7684,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315164544738047e-05,
+      "loss": 3.7646,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314325949986995e-05,
+      "loss": 3.778,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313487355235943e-05,
+      "loss": 3.7736,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312648760484891e-05,
+      "loss": 3.7852,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311811803614212e-05,
+      "loss": 3.7623,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31097320886316e-05,
+      "loss": 3.7667,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310134614112108e-05,
+      "loss": 3.7642,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309296019361056e-05,
+      "loss": 3.7553,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308459062490378e-05,
+      "loss": 3.7717,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307620467739326e-05,
+      "loss": 3.7612,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306781872988274e-05,
+      "loss": 3.7715,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305943278237222e-05,
+      "loss": 3.7644,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305106321366543e-05,
+      "loss": 3.7529,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304267726615491e-05,
+      "loss": 3.7595,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303429131864439e-05,
+      "loss": 3.7541,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025921749937596e-05,
+      "loss": 3.7702,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3017535802427076e-05,
+      "loss": 3.7645,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3009149854916556e-05,
+      "loss": 3.7477,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000763907406036e-05,
+      "loss": 3.7703,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2992377959895516e-05,
+      "loss": 3.758,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983992012385e-05,
+      "loss": 3.7508,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297560606487448e-05,
+      "loss": 3.7572,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296722011736396e-05,
+      "loss": 3.7685,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295885054865717e-05,
+      "loss": 3.7622,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295046460114665e-05,
+      "loss": 3.7658,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294207865363613e-05,
+      "loss": 3.7466,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293369270612561e-05,
+      "loss": 3.7622,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292530675861509e-05,
+      "loss": 3.7539,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2916920811104565e-05,
+      "loss": 3.7677,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2908534863594045e-05,
+      "loss": 3.7665,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2900148916083525e-05,
+      "loss": 3.7563,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289177934737674e-05,
+      "loss": 3.759,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288339339986622e-05,
+      "loss": 3.7651,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2875023831159436e-05,
+      "loss": 3.74,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2866637883648916e-05,
+      "loss": 3.7525,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2858251936138396e-05,
+      "loss": 3.7477,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284986598862787e-05,
+      "loss": 3.7586,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2841496419921085e-05,
+      "loss": 3.7456,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2833110472410565e-05,
+      "loss": 3.7537,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282472452490004e-05,
+      "loss": 3.7493,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281633857738952e-05,
+      "loss": 3.75,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807952629879e-05,
+      "loss": 3.754,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279956668236848e-05,
+      "loss": 3.7495,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279118073485796e-05,
+      "loss": 3.7614,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782811166151174e-05,
+      "loss": 3.7585,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2774425218640654e-05,
+      "loss": 3.7607,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2766039271130134e-05,
+      "loss": 3.7514,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757653323619614e-05,
+      "loss": 3.7619,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274928375491282e-05,
+      "loss": 3.7611,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27408978074023e-05,
+      "loss": 3.7451,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273251185989178e-05,
+      "loss": 3.7644,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272412591238126e-05,
+      "loss": 3.7426,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271573996487074e-05,
+      "loss": 3.7599,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270735401736022e-05,
+      "loss": 3.7424,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26989680698497e-05,
+      "loss": 3.7592,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269058212233918e-05,
+      "loss": 3.7403,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268221255363239e-05,
+      "loss": 3.7463,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267382660612187e-05,
+      "loss": 3.7418,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266544065861136e-05,
+      "loss": 3.7631,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265705471110084e-05,
+      "loss": 3.7495,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264868514239405e-05,
+      "loss": 3.7589,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264029919488353e-05,
+      "loss": 3.7444,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263191324737301e-05,
+      "loss": 3.757,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262352729986249e-05,
+      "loss": 3.7604,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261514135235197e-05,
+      "loss": 3.7555,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260675540484145e-05,
+      "loss": 3.7532,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259836945733093e-05,
+      "loss": 3.7527,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2589999888624136e-05,
+      "loss": 3.7535,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2581613941113616e-05,
+      "loss": 3.7605,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2573227993603096e-05,
+      "loss": 3.7458,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2564842046092576e-05,
+      "loss": 3.7545,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2556456098582056e-05,
+      "loss": 3.7464,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2548070151071536e-05,
+      "loss": 3.7515,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2539684203561016e-05,
+      "loss": 3.7503,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253131463485423e-05,
+      "loss": 3.7436,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2522928687343705e-05,
+      "loss": 3.7605,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2514542739833185e-05,
+      "loss": 3.7368,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2506156792322665e-05,
+      "loss": 3.7573,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.870176315307617,
+      "eval_runtime": 348.6512,
+      "eval_samples_per_second": 1094.478,
+      "eval_steps_per_second": 34.203,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2497770844812145e-05,
+      "loss": 3.742,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2489384897301625e-05,
+      "loss": 3.7315,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480998949791105e-05,
+      "loss": 3.7492,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472613002280585e-05,
+      "loss": 3.7459,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2464243433573794e-05,
+      "loss": 3.7579,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245585748606328e-05,
+      "loss": 3.7441,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244747153855276e-05,
+      "loss": 3.7374,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243908559104224e-05,
+      "loss": 3.735,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243069964353172e-05,
+      "loss": 3.7424,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242233007482493e-05,
+      "loss": 3.7498,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241394412731441e-05,
+      "loss": 3.741,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240555817980389e-05,
+      "loss": 3.7412,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239717223229337e-05,
+      "loss": 3.7435,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238878628478285e-05,
+      "loss": 3.7356,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238040033727233e-05,
+      "loss": 3.7394,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237201438976181e-05,
+      "loss": 3.7283,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236362844225129e-05,
+      "loss": 3.7369,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23552588735445e-05,
+      "loss": 3.7392,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234687292603398e-05,
+      "loss": 3.7387,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233848697852346e-05,
+      "loss": 3.7459,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2330117409816674e-05,
+      "loss": 3.7486,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321731462306154e-05,
+      "loss": 3.7385,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2313345514795634e-05,
+      "loss": 3.7401,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304959567285114e-05,
+      "loss": 3.7463,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229657361977459e-05,
+      "loss": 3.7245,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22882040510678e-05,
+      "loss": 3.7519,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227981810355728e-05,
+      "loss": 3.7312,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227143215604676e-05,
+      "loss": 3.7316,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2263046208536236e-05,
+      "loss": 3.7279,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254660261025716e-05,
+      "loss": 3.7351,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2246274313515196e-05,
+      "loss": 3.7385,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223788836600468e-05,
+      "loss": 3.7401,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222950241849416e-05,
+      "loss": 3.7436,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222113284978737e-05,
+      "loss": 3.7437,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221274690227685e-05,
+      "loss": 3.7307,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220437733357006e-05,
+      "loss": 3.7421,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219599138605954e-05,
+      "loss": 3.7411,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218760543854902e-05,
+      "loss": 3.7324,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21792194910385e-05,
+      "loss": 3.7293,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217083354352798e-05,
+      "loss": 3.7343,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216244759601746e-05,
+      "loss": 3.7345,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215407802731067e-05,
+      "loss": 3.742,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214569207980015e-05,
+      "loss": 3.7315,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2137306132289637e-05,
+      "loss": 3.7342,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128920184779116e-05,
+      "loss": 3.7443,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2120534237268596e-05,
+      "loss": 3.734,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2112148289758076e-05,
+      "loss": 3.7393,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103778721051285e-05,
+      "loss": 3.729,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2095392773540765e-05,
+      "loss": 3.7025,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2087006826030245e-05,
+      "loss": 3.7452,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2078620878519725e-05,
+      "loss": 3.7304,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2070234931009205e-05,
+      "loss": 3.7345,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061865362302414e-05,
+      "loss": 3.7219,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2053479414791894e-05,
+      "loss": 3.7123,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2045093467281374e-05,
+      "loss": 3.7179,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036707519770854e-05,
+      "loss": 3.726,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2028321572260334e-05,
+      "loss": 3.7248,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201993562474982e-05,
+      "loss": 3.7235,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20115496772393e-05,
+      "loss": 3.7361,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200318010853251e-05,
+      "loss": 3.7185,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199481053982572e-05,
+      "loss": 3.7201,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19864245923152e-05,
+      "loss": 3.7351,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197803864480468e-05,
+      "loss": 3.7078,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196965269729416e-05,
+      "loss": 3.7228,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196126674978364e-05,
+      "loss": 3.7255,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195288080227312e-05,
+      "loss": 3.734,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19444948547626e-05,
+      "loss": 3.7074,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193610890725207e-05,
+      "loss": 3.7222,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192773933854529e-05,
+      "loss": 3.7102,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1919353391034775e-05,
+      "loss": 3.7334,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191096744352425e-05,
+      "loss": 3.7337,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190258149601373e-05,
+      "loss": 3.7268,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189419554850321e-05,
+      "loss": 3.7337,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188580960099269e-05,
+      "loss": 3.735,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187742365348217e-05,
+      "loss": 3.749,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186905408477538e-05,
+      "loss": 3.7232,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186066813726486e-05,
+      "loss": 3.7288,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185228218975434e-05,
+      "loss": 3.7283,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843896242243817e-05,
+      "loss": 3.7161,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1835526673537026e-05,
+      "loss": 3.7289,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182715710483024e-05,
+      "loss": 3.7265,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181877115731972e-05,
+      "loss": 3.7307,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18103852098092e-05,
+      "loss": 3.7323,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180199926229868e-05,
+      "loss": 3.7155,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179361331478816e-05,
+      "loss": 3.715,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178522736727764e-05,
+      "loss": 3.7179,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177685779857085e-05,
+      "loss": 3.7332,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176847185106033e-05,
+      "loss": 3.7238,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176008590354981e-05,
+      "loss": 3.7172,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175169995603929e-05,
+      "loss": 3.7203,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174331400852877e-05,
+      "loss": 3.7287,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173492806101825e-05,
+      "loss": 3.7137,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172654211350773e-05,
+      "loss": 3.718,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1718172544800946e-05,
+      "loss": 3.7357,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1709786597290426e-05,
+      "loss": 3.7247,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1701400649779906e-05,
+      "loss": 3.726,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1693014702269386e-05,
+      "loss": 3.7099,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1684628754758866e-05,
+      "loss": 3.7251,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1676242807248346e-05,
+      "loss": 3.7163,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1667856859737826e-05,
+      "loss": 3.7289,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1659470912227306e-05,
+      "loss": 3.7284,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1651101343520515e-05,
+      "loss": 3.7237,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642715396009995e-05,
+      "loss": 3.7224,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1634329448499475e-05,
+      "loss": 3.7257,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1625959879792684e-05,
+      "loss": 3.7047,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1617573932282164e-05,
+      "loss": 3.7174,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1609187984771644e-05,
+      "loss": 3.7061,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160080203726113e-05,
+      "loss": 3.7251,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159241608975061e-05,
+      "loss": 3.7101,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1584030142240084e-05,
+      "loss": 3.715,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15756605735333e-05,
+      "loss": 3.7102,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156727462602278e-05,
+      "loss": 3.7157,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155888867851226e-05,
+      "loss": 3.7174,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155050273100173e-05,
+      "loss": 3.7116,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154211678349121e-05,
+      "loss": 3.7212,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153374721478443e-05,
+      "loss": 3.7283,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152536126727391e-05,
+      "loss": 3.7191,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151697531976338e-05,
+      "loss": 3.7217,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150858937225287e-05,
+      "loss": 3.7205,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1500219803546084e-05,
+      "loss": 3.7233,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149183385603556e-05,
+      "loss": 3.7131,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148344790852504e-05,
+      "loss": 3.7227,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147506196101452e-05,
+      "loss": 3.7078,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1466676013504e-05,
+      "loss": 3.7231,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145829006599348e-05,
+      "loss": 3.7094,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144990411848296e-05,
+      "loss": 3.7155,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144151817097244e-05,
+      "loss": 3.7097,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1433148602265646e-05,
+      "loss": 3.7084,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424762654755126e-05,
+      "loss": 3.7053,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1416393086048335e-05,
+      "loss": 3.7222,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140800713853782e-05,
+      "loss": 3.7159,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13996211910273e-05,
+      "loss": 3.7241,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139123524351678e-05,
+      "loss": 3.7068,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138286567480999e-05,
+      "loss": 3.7245,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137447972729947e-05,
+      "loss": 3.7226,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136609377978895e-05,
+      "loss": 3.7185,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135770783227843e-05,
+      "loss": 3.7191,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134932188476791e-05,
+      "loss": 3.7163,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134093593725739e-05,
+      "loss": 3.7182,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133254998974687e-05,
+      "loss": 3.7274,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132416404223635e-05,
+      "loss": 3.7084,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131579447352956e-05,
+      "loss": 3.7164,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130740852601904e-05,
+      "loss": 3.7169,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129902257850852e-05,
+      "loss": 3.7153,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1290636630998006e-05,
+      "loss": 3.7082,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1282250683487486e-05,
+      "loss": 3.7072,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1273864735976966e-05,
+      "loss": 3.7235,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1265478788466446e-05,
+      "loss": 3.7042,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125709284095592e-05,
+      "loss": 3.7219,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.860013723373413,
+      "eval_runtime": 310.7249,
+      "eval_samples_per_second": 1228.067,
+      "eval_steps_per_second": 38.378,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1248723272249135e-05,
+      "loss": 3.7073,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1240337324738615e-05,
+      "loss": 3.6964,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1231951377228095e-05,
+      "loss": 3.7173,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122356542971757e-05,
+      "loss": 3.7067,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1215212239814513e-05,
+      "loss": 3.7212,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120682629230399e-05,
+      "loss": 3.7124,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119844034479347e-05,
+      "loss": 3.7047,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119005439728296e-05,
+      "loss": 3.7014,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118168482857617e-05,
+      "loss": 3.7075,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117329888106565e-05,
+      "loss": 3.7131,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116491293355513e-05,
+      "loss": 3.7044,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115652698604461e-05,
+      "loss": 3.7066,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114815741733782e-05,
+      "loss": 3.7045,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11397714698273e-05,
+      "loss": 3.7023,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113138552231678e-05,
+      "loss": 3.7047,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112299957480626e-05,
+      "loss": 3.6956,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111461362729574e-05,
+      "loss": 3.6999,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110622767978521e-05,
+      "loss": 3.7058,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10978417322747e-05,
+      "loss": 3.7047,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108947216356791e-05,
+      "loss": 3.7083,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1081086216057394e-05,
+      "loss": 3.7141,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107270026854687e-05,
+      "loss": 3.7058,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106431432103635e-05,
+      "loss": 3.7038,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105592837352583e-05,
+      "loss": 3.7071,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104755880481904e-05,
+      "loss": 3.6962,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1039172857308516e-05,
+      "loss": 3.7147,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1030786909797996e-05,
+      "loss": 3.6963,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1022400962287476e-05,
+      "loss": 3.7021,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1014015014776956e-05,
+      "loss": 3.6965,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005645446070165e-05,
+      "loss": 3.699,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0997259498559645e-05,
+      "loss": 3.7003,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098887355104913e-05,
+      "loss": 3.7078,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098048760353861e-05,
+      "loss": 3.7074,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097210165602809e-05,
+      "loss": 3.7098,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096371570851757e-05,
+      "loss": 3.6956,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095534613981078e-05,
+      "loss": 3.7111,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094696019230026e-05,
+      "loss": 3.7073,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093857424478974e-05,
+      "loss": 3.696,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093018829727922e-05,
+      "loss": 3.6998,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092181872857243e-05,
+      "loss": 3.7002,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091343278106191e-05,
+      "loss": 3.6984,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090504683355139e-05,
+      "loss": 3.7073,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089666088604087e-05,
+      "loss": 3.7013,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088827493853035e-05,
+      "loss": 3.7007,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087988899101983e-05,
+      "loss": 3.7095,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0871503043509316e-05,
+      "loss": 3.7009,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0863117095998796e-05,
+      "loss": 3.7035,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0854747527292005e-05,
+      "loss": 3.6992,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0846377958585214e-05,
+      "loss": 3.6671,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0837992011074694e-05,
+      "loss": 3.7128,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0829606063564174e-05,
+      "loss": 3.6957,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0821220116053654e-05,
+      "loss": 3.7044,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081285054734686e-05,
+      "loss": 3.6849,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080446459983634e-05,
+      "loss": 3.6828,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079607865232582e-05,
+      "loss": 3.6809,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07876927048153e-05,
+      "loss": 3.6926,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077930675730478e-05,
+      "loss": 3.6953,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077092080979427e-05,
+      "loss": 3.6876,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076253486228375e-05,
+      "loss": 3.7068,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075414891477323e-05,
+      "loss": 3.6884,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074579572487017e-05,
+      "loss": 3.691,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073740977735965e-05,
+      "loss": 3.6999,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072902382984913e-05,
+      "loss": 3.6755,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072063788233861e-05,
+      "loss": 3.6852,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071225193482809e-05,
+      "loss": 3.6941,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070386598731757e-05,
+      "loss": 3.7006,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069548003980705e-05,
+      "loss": 3.6752,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068709409229652e-05,
+      "loss": 3.6884,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0678724523589736e-05,
+      "loss": 3.678,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067033857607922e-05,
+      "loss": 3.7017,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06619526285687e-05,
+      "loss": 3.6958,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0653566681058176e-05,
+      "loss": 3.6959,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0645180733547656e-05,
+      "loss": 3.6998,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636794786037136e-05,
+      "loss": 3.7037,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0628408838526616e-05,
+      "loss": 3.7159,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0620039269819825e-05,
+      "loss": 3.692,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611653322309305e-05,
+      "loss": 3.6961,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0603267374798785e-05,
+      "loss": 3.6977,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0594881427288265e-05,
+      "loss": 3.6832,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0586511858581474e-05,
+      "loss": 3.6935,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057812591107096e-05,
+      "loss": 3.6957,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056973996356044e-05,
+      "loss": 3.6999,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056135401604992e-05,
+      "loss": 3.6989,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05529680685394e-05,
+      "loss": 3.6854,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054458212102888e-05,
+      "loss": 3.6822,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053619617351836e-05,
+      "loss": 3.6864,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052782660481157e-05,
+      "loss": 3.7018,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051944065730105e-05,
+      "loss": 3.6917,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051105470979053e-05,
+      "loss": 3.685,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050266876228001e-05,
+      "loss": 3.6877,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049428281476949e-05,
+      "loss": 3.6968,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048589686725897e-05,
+      "loss": 3.6843,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047751091974845e-05,
+      "loss": 3.6874,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046912497223793e-05,
+      "loss": 3.7013,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046075540353114e-05,
+      "loss": 3.6923,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0452369456020625e-05,
+      "loss": 3.694,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0443983508510105e-05,
+      "loss": 3.6819,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0435597560999585e-05,
+      "loss": 3.6902,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0427211613489065e-05,
+      "loss": 3.6856,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041882566597854e-05,
+      "loss": 3.6994,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041043971846802e-05,
+      "loss": 3.6933,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04020537709575e-05,
+      "loss": 3.6969,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039368420225071e-05,
+      "loss": 3.6913,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038529825474019e-05,
+      "loss": 3.6912,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03769286860334e-05,
+      "loss": 3.6808,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036854273852288e-05,
+      "loss": 3.6808,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036015679101236e-05,
+      "loss": 3.678,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035177084350184e-05,
+      "loss": 3.6953,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034340127479506e-05,
+      "loss": 3.6797,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033501532728454e-05,
+      "loss": 3.6816,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032662937977401e-05,
+      "loss": 3.6766,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031824343226349e-05,
+      "loss": 3.6887,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030985748475297e-05,
+      "loss": 3.6878,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030147153724245e-05,
+      "loss": 3.6783,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029308558973193e-05,
+      "loss": 3.6941,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028471602102514e-05,
+      "loss": 3.6925,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027633007351462e-05,
+      "loss": 3.6868,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02679441260041e-05,
+      "loss": 3.6936,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025955817849358e-05,
+      "loss": 3.6898,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02511886097868e-05,
+      "loss": 3.6892,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024280266227628e-05,
+      "loss": 3.6868,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023441671476576e-05,
+      "loss": 3.6896,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022603076725524e-05,
+      "loss": 3.6791,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021764481974472e-05,
+      "loss": 3.6919,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0209258872234197e-05,
+      "loss": 3.684,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0200872924723677e-05,
+      "loss": 3.6844,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0192486977213157e-05,
+      "loss": 3.6775,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0184117408506366e-05,
+      "loss": 3.6781,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0175731460995846e-05,
+      "loss": 3.6767,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0167345513485326e-05,
+      "loss": 3.6904,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0158959565974805e-05,
+      "loss": 3.6853,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015060637607175e-05,
+      "loss": 3.6935,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014222042856123e-05,
+      "loss": 3.6745,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013383448105071e-05,
+      "loss": 3.6985,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012544853354019e-05,
+      "loss": 3.6876,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011706258602967e-05,
+      "loss": 3.6885,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010867663851915e-05,
+      "loss": 3.6879,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010030706981236e-05,
+      "loss": 3.6822,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009192112230184e-05,
+      "loss": 3.6864,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008353517479132e-05,
+      "loss": 3.6967,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00751492272808e-05,
+      "loss": 3.6782,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006677965857401e-05,
+      "loss": 3.6879,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005839371106349e-05,
+      "loss": 3.6889,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005000776355297e-05,
+      "loss": 3.6817,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0041621816042455e-05,
+      "loss": 3.6792,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0033235868531935e-05,
+      "loss": 3.6792,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0024849921021415e-05,
+      "loss": 3.6918,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0016463973510895e-05,
+      "loss": 3.6766,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000807802600037e-05,
+      "loss": 3.6917,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8540804386138916,
+      "eval_runtime": 310.113,
+      "eval_samples_per_second": 1230.49,
+      "eval_steps_per_second": 38.454,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9999708457293584e-05,
+      "loss": 3.6758,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9991322509783064e-05,
+      "loss": 3.6711,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9982936562272544e-05,
+      "loss": 3.6807,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997455061476202e-05,
+      "loss": 3.6801,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996618104605523e-05,
+      "loss": 3.6925,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995779509854471e-05,
+      "loss": 3.6803,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994940915103419e-05,
+      "loss": 3.6759,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994102320352367e-05,
+      "loss": 3.6732,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993263725601315e-05,
+      "loss": 3.6727,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992426768730637e-05,
+      "loss": 3.6859,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991588173979584e-05,
+      "loss": 3.6745,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990749579228532e-05,
+      "loss": 3.6748,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98991098447748e-05,
+      "loss": 3.6837,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989072389726428e-05,
+      "loss": 3.6653,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988233794975376e-05,
+      "loss": 3.673,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987395200224324e-05,
+      "loss": 3.6686,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986556605473272e-05,
+      "loss": 3.6687,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985719648602593e-05,
+      "loss": 3.6766,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984881053851541e-05,
+      "loss": 3.6751,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9840440969808626e-05,
+      "loss": 3.6745,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9832055022298106e-05,
+      "loss": 3.6884,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823669074787586e-05,
+      "loss": 3.6779,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9815283127277066e-05,
+      "loss": 3.6778,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9806897179766546e-05,
+      "loss": 3.6733,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9798511232256026e-05,
+      "loss": 3.6712,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9790141663549235e-05,
+      "loss": 3.6835,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781755716038715e-05,
+      "loss": 3.6678,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9773369768528195e-05,
+      "loss": 3.6763,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9764983821017675e-05,
+      "loss": 3.6624,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9756597873507155e-05,
+      "loss": 3.6722,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9748211925996635e-05,
+      "loss": 3.6687,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9739825978486115e-05,
+      "loss": 3.6769,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9731456409779324e-05,
+      "loss": 3.6815,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972307046226881e-05,
+      "loss": 3.681,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971468451475829e-05,
+      "loss": 3.6705,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970629856724777e-05,
+      "loss": 3.6739,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969792899854098e-05,
+      "loss": 3.6819,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968954305103046e-05,
+      "loss": 3.6687,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968115710351994e-05,
+      "loss": 3.6696,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967277115600942e-05,
+      "loss": 3.6696,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96643852084989e-05,
+      "loss": 3.6744,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965601563979211e-05,
+      "loss": 3.672,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964762969228159e-05,
+      "loss": 3.6778,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963924374477107e-05,
+      "loss": 3.6727,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963085779726055e-05,
+      "loss": 3.6789,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962247184975003e-05,
+      "loss": 3.6771,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9614102281043244e-05,
+      "loss": 3.674,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9605716333532724e-05,
+      "loss": 3.6718,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9597330386022204e-05,
+      "loss": 3.637,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958894443851168e-05,
+      "loss": 3.6824,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958055849100116e-05,
+      "loss": 3.6645,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957217254349064e-05,
+      "loss": 3.6767,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956380297478385e-05,
+      "loss": 3.6616,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9555417027273326e-05,
+      "loss": 3.6553,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9547031079762806e-05,
+      "loss": 3.6488,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9538645132252286e-05,
+      "loss": 3.6627,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9530259184741766e-05,
+      "loss": 3.6682,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9521873237231246e-05,
+      "loss": 3.6615,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951350366852446e-05,
+      "loss": 3.6747,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950511772101394e-05,
+      "loss": 3.6612,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949673177350342e-05,
+      "loss": 3.6613,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94883458259929e-05,
+      "loss": 3.671,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947995987848238e-05,
+      "loss": 3.6486,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947157393097186e-05,
+      "loss": 3.6566,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946320436226507e-05,
+      "loss": 3.6675,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945481841475455e-05,
+      "loss": 3.6698,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944643246724403e-05,
+      "loss": 3.6509,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943804651973351e-05,
+      "loss": 3.6602,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942966057222299e-05,
+      "loss": 3.6503,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942127462471247e-05,
+      "loss": 3.6731,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941288867720195e-05,
+      "loss": 3.6703,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940451910849517e-05,
+      "loss": 3.6674,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939613316098465e-05,
+      "loss": 3.6717,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9387747213474127e-05,
+      "loss": 3.6746,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9379361265963607e-05,
+      "loss": 3.6887,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9370975318453086e-05,
+      "loss": 3.6656,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9362589370942566e-05,
+      "loss": 3.6681,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935420342343204e-05,
+      "loss": 3.6637,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934581747592152e-05,
+      "loss": 3.6599,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9337447907214735e-05,
+      "loss": 3.6644,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9329061959704215e-05,
+      "loss": 3.6696,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932067601219369e-05,
+      "loss": 3.6692,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931229006468317e-05,
+      "loss": 3.674,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930390411717265e-05,
+      "loss": 3.655,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9295518169662135e-05,
+      "loss": 3.6541,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9287132222151615e-05,
+      "loss": 3.6586,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9278762653444824e-05,
+      "loss": 3.6756,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9270376705934304e-05,
+      "loss": 3.6661,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9261990758423784e-05,
+      "loss": 3.655,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9253604810913264e-05,
+      "loss": 3.6609,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9245218863402744e-05,
+      "loss": 3.6666,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9236832915892224e-05,
+      "loss": 3.6591,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9228446968381704e-05,
+      "loss": 3.6585,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922007739967491e-05,
+      "loss": 3.6736,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921169145216439e-05,
+      "loss": 3.6671,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920330550465387e-05,
+      "loss": 3.6661,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919491955714335e-05,
+      "loss": 3.6562,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918653360963283e-05,
+      "loss": 3.66,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917814766212232e-05,
+      "loss": 3.66,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916977809341553e-05,
+      "loss": 3.6703,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916139214590501e-05,
+      "loss": 3.6675,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915300619839449e-05,
+      "loss": 3.6657,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914462025088397e-05,
+      "loss": 3.6675,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913623430337345e-05,
+      "loss": 3.66,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912786473466666e-05,
+      "loss": 3.6589,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911947878715614e-05,
+      "loss": 3.652,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911109283964562e-05,
+      "loss": 3.6479,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91027068921351e-05,
+      "loss": 3.6695,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909432094462458e-05,
+      "loss": 3.656,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908593499711406e-05,
+      "loss": 3.651,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907756542840727e-05,
+      "loss": 3.6508,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906917948089675e-05,
+      "loss": 3.6538,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9060793533386226e-05,
+      "loss": 3.6646,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9052407585875706e-05,
+      "loss": 3.6519,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9044021638365186e-05,
+      "loss": 3.6644,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90356520696584e-05,
+      "loss": 3.6659,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9027266122147875e-05,
+      "loss": 3.6622,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9018880174637355e-05,
+      "loss": 3.6659,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9010494227126835e-05,
+      "loss": 3.6619,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900212465842005e-05,
+      "loss": 3.6614,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8993738710909524e-05,
+      "loss": 3.6604,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898535276339901e-05,
+      "loss": 3.6603,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897696681588849e-05,
+      "loss": 3.6563,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896858086837797e-05,
+      "loss": 3.6634,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896019492086745e-05,
+      "loss": 3.6583,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895180897335693e-05,
+      "loss": 3.657,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894342302584641e-05,
+      "loss": 3.6509,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893505345713962e-05,
+      "loss": 3.6511,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892668388843283e-05,
+      "loss": 3.6503,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891829794092231e-05,
+      "loss": 3.6626,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890991199341179e-05,
+      "loss": 3.6565,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890152604590127e-05,
+      "loss": 3.6705,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889315647719448e-05,
+      "loss": 3.6484,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8884770529683965e-05,
+      "loss": 3.6687,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8876384582173445e-05,
+      "loss": 3.6619,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8867998634662925e-05,
+      "loss": 3.6635,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8859629065956134e-05,
+      "loss": 3.6623,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8851243118445614e-05,
+      "loss": 3.6557,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8842857170935094e-05,
+      "loss": 3.6628,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8834471223424574e-05,
+      "loss": 3.6695,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8826085275914054e-05,
+      "loss": 3.6511,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881771570720726e-05,
+      "loss": 3.6625,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880932975969674e-05,
+      "loss": 3.6584,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880094381218622e-05,
+      "loss": 3.6543,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.87925578646757e-05,
+      "loss": 3.6572,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878417191716518e-05,
+      "loss": 3.6495,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877578596965466e-05,
+      "loss": 3.6652,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876740002214414e-05,
+      "loss": 3.6529,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875901407463363e-05,
+      "loss": 3.6619,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.849538803100586,
+      "eval_runtime": 307.5313,
+      "eval_samples_per_second": 1240.82,
+      "eval_steps_per_second": 38.777,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.875064450592684e-05,
+      "loss": 3.6499,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874225855841632e-05,
+      "loss": 3.6451,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.873388898970953e-05,
+      "loss": 3.652,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.872550304219901e-05,
+      "loss": 3.6533,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871711709468849e-05,
+      "loss": 3.6674,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870873114717797e-05,
+      "loss": 3.6533,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8700361578471176e-05,
+      "loss": 3.6515,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8691975630960656e-05,
+      "loss": 3.6499,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8683589683450136e-05,
+      "loss": 3.645,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8675203735939616e-05,
+      "loss": 3.6604,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8666817788429096e-05,
+      "loss": 3.6478,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865844821972231e-05,
+      "loss": 3.6497,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865006227221179e-05,
+      "loss": 3.6582,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864167632470127e-05,
+      "loss": 3.6407,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.863329037719075e-05,
+      "loss": 3.646,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862492080848396e-05,
+      "loss": 3.6418,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861653486097344e-05,
+      "loss": 3.6454,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.860814891346292e-05,
+      "loss": 3.6439,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85997629659524e-05,
+      "loss": 3.6533,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859139339724561e-05,
+      "loss": 3.6512,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858300744973509e-05,
+      "loss": 3.6613,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857462150222457e-05,
+      "loss": 3.6497,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.856623555471405e-05,
+      "loss": 3.6526,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8557849607203537e-05,
+      "loss": 3.6447,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854946365969301e-05,
+      "loss": 3.6493,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8541094090986226e-05,
+      "loss": 3.6556,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8532708143475706e-05,
+      "loss": 3.642,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8524322195965185e-05,
+      "loss": 3.6531,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851593624845466e-05,
+      "loss": 3.636,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.850755030094414e-05,
+      "loss": 3.6446,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849916435343362e-05,
+      "loss": 3.6453,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849079478472683e-05,
+      "loss": 3.65,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848240883721631e-05,
+      "loss": 3.6591,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847402288970579e-05,
+      "loss": 3.6555,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8465636942195274e-05,
+      "loss": 3.6463,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8457250994684754e-05,
+      "loss": 3.6424,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.844888142597796e-05,
+      "loss": 3.659,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.844049547846744e-05,
+      "loss": 3.6447,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843210953095692e-05,
+      "loss": 3.6444,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84237235834464e-05,
+      "loss": 3.6437,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.841535401473961e-05,
+      "loss": 3.6491,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840696806722909e-05,
+      "loss": 3.6477,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839858211971857e-05,
+      "loss": 3.65,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839019617220805e-05,
+      "loss": 3.65,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838181022469753e-05,
+      "loss": 3.649,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837342427718701e-05,
+      "loss": 3.653,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836505470848023e-05,
+      "loss": 3.6481,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835666876096971e-05,
+      "loss": 3.6478,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834828281345919e-05,
+      "loss": 3.6143,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833989686594867e-05,
+      "loss": 3.6543,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833151091843815e-05,
+      "loss": 3.6397,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832312497092763e-05,
+      "loss": 3.6544,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831475540222084e-05,
+      "loss": 3.6359,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830636945471032e-05,
+      "loss": 3.6296,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.82979835071998e-05,
+      "loss": 3.6244,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828959755968928e-05,
+      "loss": 3.6362,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8281227990982486e-05,
+      "loss": 3.6487,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8272842043471966e-05,
+      "loss": 3.6308,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8264456095961446e-05,
+      "loss": 3.6528,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8256070148450926e-05,
+      "loss": 3.6413,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824770057974414e-05,
+      "loss": 3.6305,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823931463223362e-05,
+      "loss": 3.6465,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.82309286847231e-05,
+      "loss": 3.6302,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822254273721258e-05,
+      "loss": 3.6293,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821417316850579e-05,
+      "loss": 3.6443,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820578722099527e-05,
+      "loss": 3.6461,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819740127348475e-05,
+      "loss": 3.6274,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818901532597423e-05,
+      "loss": 3.6338,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818062937846371e-05,
+      "loss": 3.6248,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817224343095319e-05,
+      "loss": 3.647,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8163857483442663e-05,
+      "loss": 3.6459,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815548791473588e-05,
+      "loss": 3.6413,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814710196722536e-05,
+      "loss": 3.6476,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8138716019714846e-05,
+      "loss": 3.6502,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813033007220432e-05,
+      "loss": 3.6645,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81219441246938e-05,
+      "loss": 3.6424,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811355817718328e-05,
+      "loss": 3.6449,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810517222967276e-05,
+      "loss": 3.6393,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.809678628216224e-05,
+      "loss": 3.6368,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808841671345545e-05,
+      "loss": 3.6385,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808003076594493e-05,
+      "loss": 3.6462,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807164481843441e-05,
+      "loss": 3.6434,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806325887092389e-05,
+      "loss": 3.6527,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805487292341337e-05,
+      "loss": 3.6296,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804648697590285e-05,
+      "loss": 3.63,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803810102839233e-05,
+      "loss": 3.6367,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8029731459685544e-05,
+      "loss": 3.6457,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8021345512175024e-05,
+      "loss": 3.6434,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8012959564664504e-05,
+      "loss": 3.6348,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8004573617153984e-05,
+      "loss": 3.6362,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7996187669643464e-05,
+      "loss": 3.6435,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798781810093667e-05,
+      "loss": 3.6345,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797943215342615e-05,
+      "loss": 3.632,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797104620591563e-05,
+      "loss": 3.6481,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796266025840511e-05,
+      "loss": 3.6413,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.795427431089459e-05,
+      "loss": 3.645,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79459047421878e-05,
+      "loss": 3.6309,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793751879467728e-05,
+      "loss": 3.6364,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792913284716677e-05,
+      "loss": 3.6345,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792074689965625e-05,
+      "loss": 3.649,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791236095214573e-05,
+      "loss": 3.6456,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.790399138343894e-05,
+      "loss": 3.6384,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789560543592842e-05,
+      "loss": 3.6424,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78872194884179e-05,
+      "loss": 3.636,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787883354090738e-05,
+      "loss": 3.64,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7870463972200586e-05,
+      "loss": 3.6268,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7862078024690066e-05,
+      "loss": 3.6278,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7853692077179546e-05,
+      "loss": 3.6404,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7845306129669026e-05,
+      "loss": 3.6367,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7836936560962235e-05,
+      "loss": 3.6228,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782855061345172e-05,
+      "loss": 3.6286,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78201646659412e-05,
+      "loss": 3.6306,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781177871843068e-05,
+      "loss": 3.6389,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7803392770920155e-05,
+      "loss": 3.6283,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7795006823409635e-05,
+      "loss": 3.6427,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778663725470285e-05,
+      "loss": 3.6436,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7778251307192324e-05,
+      "loss": 3.6391,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7769865359681804e-05,
+      "loss": 3.6369,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7761479412171284e-05,
+      "loss": 3.6435,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77531098434645e-05,
+      "loss": 3.6374,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.774472389595397e-05,
+      "loss": 3.64,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.773633794844346e-05,
+      "loss": 3.636,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772795200093294e-05,
+      "loss": 3.6295,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771956605342242e-05,
+      "loss": 3.6389,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77111801059119e-05,
+      "loss": 3.6397,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770279415840138e-05,
+      "loss": 3.6308,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769440821089086e-05,
+      "loss": 3.6317,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768603864218407e-05,
+      "loss": 3.6275,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767766907347728e-05,
+      "loss": 3.6221,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766928312596676e-05,
+      "loss": 3.6414,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766089717845624e-05,
+      "loss": 3.6316,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765251123094572e-05,
+      "loss": 3.6465,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764414166223893e-05,
+      "loss": 3.6247,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7635755714728413e-05,
+      "loss": 3.6441,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.762736976721789e-05,
+      "loss": 3.6411,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761898381970737e-05,
+      "loss": 3.6405,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761061425100058e-05,
+      "loss": 3.637,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760222830349006e-05,
+      "loss": 3.6294,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759384235597954e-05,
+      "loss": 3.6423,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758545640846902e-05,
+      "loss": 3.6434,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75770704609585e-05,
+      "loss": 3.6296,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756870089225171e-05,
+      "loss": 3.6391,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756031494474119e-05,
+      "loss": 3.637,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755192899723067e-05,
+      "loss": 3.6323,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754354304972015e-05,
+      "loss": 3.6319,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753515710220963e-05,
+      "loss": 3.6278,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752677115469911e-05,
+      "loss": 3.6411,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751838520718859e-05,
+      "loss": 3.6339,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.750999925967808e-05,
+      "loss": 3.6396,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.846451759338379,
+      "eval_runtime": 304.7916,
+      "eval_samples_per_second": 1251.974,
+      "eval_steps_per_second": 39.125,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750162969097129e-05,
+      "loss": 3.6263,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749324374346077e-05,
+      "loss": 3.6228,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748485779595025e-05,
+      "loss": 3.629,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747647184843973e-05,
+      "loss": 3.6308,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7468102279732936e-05,
+      "loss": 3.6445,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7459716332222416e-05,
+      "loss": 3.6329,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7451330384711896e-05,
+      "loss": 3.626,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7442944437201376e-05,
+      "loss": 3.6276,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7434558489690856e-05,
+      "loss": 3.6224,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7426188920984065e-05,
+      "loss": 3.6326,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7417802973473545e-05,
+      "loss": 3.6271,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740941702596303e-05,
+      "loss": 3.6318,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740103107845251e-05,
+      "loss": 3.6352,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7392645130941985e-05,
+      "loss": 3.6138,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73842755622352e-05,
+      "loss": 3.6252,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737588961472468e-05,
+      "loss": 3.6144,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736750366721416e-05,
+      "loss": 3.6283,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7359117719703634e-05,
+      "loss": 3.6251,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7350731772193114e-05,
+      "loss": 3.6281,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734236220348633e-05,
+      "loss": 3.6261,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733397625597581e-05,
+      "loss": 3.6386,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732559030846528e-05,
+      "loss": 3.6261,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731720436095477e-05,
+      "loss": 3.6317,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730881841344425e-05,
+      "loss": 3.6258,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730044884473746e-05,
+      "loss": 3.6269,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729206289722694e-05,
+      "loss": 3.632,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728367694971642e-05,
+      "loss": 3.6162,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.72752910022059e-05,
+      "loss": 3.6284,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726690505469538e-05,
+      "loss": 3.6162,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725851910718486e-05,
+      "loss": 3.6212,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.725013315967434e-05,
+      "loss": 3.6261,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724174721216382e-05,
+      "loss": 3.6287,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723337764345703e-05,
+      "loss": 3.6273,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722499169594651e-05,
+      "loss": 3.639,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.721660574843599e-05,
+      "loss": 3.6205,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720821980092547e-05,
+      "loss": 3.6213,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719985023221868e-05,
+      "loss": 3.6393,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719146428470816e-05,
+      "loss": 3.6221,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.718307833719764e-05,
+      "loss": 3.6222,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717469238968712e-05,
+      "loss": 3.6228,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.716632282098033e-05,
+      "loss": 3.621,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.715793687346981e-05,
+      "loss": 3.6258,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714955092595929e-05,
+      "loss": 3.6257,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714116497844877e-05,
+      "loss": 3.6303,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.713277903093825e-05,
+      "loss": 3.6243,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.712439308342773e-05,
+      "loss": 3.6328,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711602351472094e-05,
+      "loss": 3.6242,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.710763756721042e-05,
+      "loss": 3.6286,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709925161969991e-05,
+      "loss": 3.5931,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709086567218939e-05,
+      "loss": 3.6316,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.708247972467887e-05,
+      "loss": 3.6142,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7074110155972076e-05,
+      "loss": 3.6306,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7065724208461556e-05,
+      "loss": 3.6134,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7057338260951036e-05,
+      "loss": 3.6096,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7048952313440516e-05,
+      "loss": 3.6026,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7040582744733725e-05,
+      "loss": 3.6152,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7032196797223205e-05,
+      "loss": 3.6269,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7023810849712685e-05,
+      "loss": 3.6084,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7015424902202165e-05,
+      "loss": 3.6296,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7007038954691645e-05,
+      "loss": 3.6215,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699866938598486e-05,
+      "loss": 3.6061,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699028343847434e-05,
+      "loss": 3.6242,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698189749096382e-05,
+      "loss": 3.6073,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6973511543453294e-05,
+      "loss": 3.6093,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6965125595942774e-05,
+      "loss": 3.6216,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695675602723599e-05,
+      "loss": 3.6235,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694837007972547e-05,
+      "loss": 3.6027,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693998413221494e-05,
+      "loss": 3.6162,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693159818470442e-05,
+      "loss": 3.605,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69232122371939e-05,
+      "loss": 3.6182,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691482628968338e-05,
+      "loss": 3.6259,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69064567209766e-05,
+      "loss": 3.6206,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689807077346608e-05,
+      "loss": 3.6286,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688968482595556e-05,
+      "loss": 3.626,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688129887844504e-05,
+      "loss": 3.6419,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.687291293093452e-05,
+      "loss": 3.6225,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686454336222773e-05,
+      "loss": 3.6268,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685615741471721e-05,
+      "loss": 3.6125,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684777146720669e-05,
+      "loss": 3.6144,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683938551969617e-05,
+      "loss": 3.6176,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683101595098938e-05,
+      "loss": 3.6219,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682263000347886e-05,
+      "loss": 3.6248,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6814244055968337e-05,
+      "loss": 3.6285,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6805858108457817e-05,
+      "loss": 3.6073,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6797472160947297e-05,
+      "loss": 3.6106,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6789086213436776e-05,
+      "loss": 3.6121,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678071664472999e-05,
+      "loss": 3.6266,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677233069721947e-05,
+      "loss": 3.6186,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676394474970895e-05,
+      "loss": 3.6144,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675555880219843e-05,
+      "loss": 3.6185,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.674717285468791e-05,
+      "loss": 3.6198,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673880328598112e-05,
+      "loss": 3.6102,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.67304173384706e-05,
+      "loss": 3.6125,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672203139096008e-05,
+      "loss": 3.6254,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.671364544344956e-05,
+      "loss": 3.625,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.670525949593904e-05,
+      "loss": 3.6222,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669688992723225e-05,
+      "loss": 3.6078,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668850397972173e-05,
+      "loss": 3.6142,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668011803221122e-05,
+      "loss": 3.6146,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66717320847007e-05,
+      "loss": 3.6205,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.666334613719018e-05,
+      "loss": 3.6277,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6654976568483386e-05,
+      "loss": 3.6144,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6646590620972866e-05,
+      "loss": 3.6216,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6638204673462346e-05,
+      "loss": 3.6179,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6629818725951826e-05,
+      "loss": 3.615,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6621449157245035e-05,
+      "loss": 3.6075,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6613063209734515e-05,
+      "loss": 3.6062,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6604677262223995e-05,
+      "loss": 3.6179,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6596291314713475e-05,
+      "loss": 3.6135,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6587905367202955e-05,
+      "loss": 3.6042,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657953579849617e-05,
+      "loss": 3.6092,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657114985098565e-05,
+      "loss": 3.6028,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656276390347513e-05,
+      "loss": 3.6185,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6554377955964604e-05,
+      "loss": 3.6108,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654600838725782e-05,
+      "loss": 3.6181,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65376224397473e-05,
+      "loss": 3.6218,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652923649223678e-05,
+      "loss": 3.6182,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652085054472625e-05,
+      "loss": 3.616,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651248097601947e-05,
+      "loss": 3.626,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650409502850895e-05,
+      "loss": 3.6165,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649570908099842e-05,
+      "loss": 3.6186,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648732313348791e-05,
+      "loss": 3.6122,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647893718597739e-05,
+      "loss": 3.6117,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647055123846687e-05,
+      "loss": 3.6157,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646218166976008e-05,
+      "loss": 3.6162,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.645379572224956e-05,
+      "loss": 3.6073,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644540977473904e-05,
+      "loss": 3.6115,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643702382722852e-05,
+      "loss": 3.6032,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6428654258521726e-05,
+      "loss": 3.6015,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6420268311011206e-05,
+      "loss": 3.6196,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6411882363500686e-05,
+      "loss": 3.6128,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6403496415990166e-05,
+      "loss": 3.6234,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6395126847283375e-05,
+      "loss": 3.6065,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.638674089977286e-05,
+      "loss": 3.6211,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637835495226234e-05,
+      "loss": 3.6197,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636996900475182e-05,
+      "loss": 3.6167,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636159943604503e-05,
+      "loss": 3.6165,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635321348853451e-05,
+      "loss": 3.6089,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634482754102399e-05,
+      "loss": 3.6259,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633644159351347e-05,
+      "loss": 3.6246,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632805564600295e-05,
+      "loss": 3.608,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631968607729616e-05,
+      "loss": 3.6157,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631130012978564e-05,
+      "loss": 3.6161,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630291418227512e-05,
+      "loss": 3.6107,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.62945282347646e-05,
+      "loss": 3.6089,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628614228725408e-05,
+      "loss": 3.6112,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627775633974356e-05,
+      "loss": 3.6157,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6269386771036776e-05,
+      "loss": 3.6161,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6261000823526256e-05,
+      "loss": 3.6165,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.845015525817871,
+      "eval_runtime": 304.9308,
+      "eval_samples_per_second": 1251.402,
+      "eval_steps_per_second": 39.107,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6252614876015735e-05,
+      "loss": 3.6053,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6244228928505215e-05,
+      "loss": 3.6004,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6235859359798425e-05,
+      "loss": 3.6084,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6227473412287904e-05,
+      "loss": 3.6126,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6219087464777384e-05,
+      "loss": 3.6228,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6210701517266864e-05,
+      "loss": 3.6125,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6202315569756344e-05,
+      "loss": 3.6094,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6193946001049553e-05,
+      "loss": 3.6072,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6185560053539033e-05,
+      "loss": 3.5983,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617717410602851e-05,
+      "loss": 3.6086,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616878815851799e-05,
+      "loss": 3.6126,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616041858981121e-05,
+      "loss": 3.6082,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615204902110442e-05,
+      "loss": 3.6127,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61436630735939e-05,
+      "loss": 3.5962,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.613527712608338e-05,
+      "loss": 3.6031,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612689117857286e-05,
+      "loss": 3.5916,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611850523106234e-05,
+      "loss": 3.6078,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611011928355182e-05,
+      "loss": 3.6038,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61017333360413e-05,
+      "loss": 3.606,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609334738853078e-05,
+      "loss": 3.605,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608497781982399e-05,
+      "loss": 3.6193,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607659187231347e-05,
+      "loss": 3.6044,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606820592480295e-05,
+      "loss": 3.6118,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6059819977292434e-05,
+      "loss": 3.603,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605145040858564e-05,
+      "loss": 3.6089,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.604306446107512e-05,
+      "loss": 3.6138,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60346785135646e-05,
+      "loss": 3.595,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602629256605408e-05,
+      "loss": 3.607,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6017906618543556e-05,
+      "loss": 3.598,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600953704983677e-05,
+      "loss": 3.6014,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600115110232625e-05,
+      "loss": 3.6065,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599276515481573e-05,
+      "loss": 3.6097,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5984379207305205e-05,
+      "loss": 3.6022,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597600963859842e-05,
+      "loss": 3.6216,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59676236910879e-05,
+      "loss": 3.6027,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595923774357739e-05,
+      "loss": 3.5997,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595085179606686e-05,
+      "loss": 3.616,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5942482227360076e-05,
+      "loss": 3.5991,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5934096279849556e-05,
+      "loss": 3.6019,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.592571033233903e-05,
+      "loss": 3.6064,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591732438482851e-05,
+      "loss": 3.6005,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590893843731799e-05,
+      "loss": 3.6064,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5900568868611205e-05,
+      "loss": 3.612,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589218292110068e-05,
+      "loss": 3.6059,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588379697359016e-05,
+      "loss": 3.6076,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587541102607964e-05,
+      "loss": 3.608,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5867041457372854e-05,
+      "loss": 3.6051,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5858655509862334e-05,
+      "loss": 3.6083,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5850269562351814e-05,
+      "loss": 3.5758,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5841883614841294e-05,
+      "loss": 3.6079,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5833497667330774e-05,
+      "loss": 3.5972,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5825111719820254e-05,
+      "loss": 3.6095,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581674215111346e-05,
+      "loss": 3.5964,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580835620360294e-05,
+      "loss": 3.5851,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579997025609242e-05,
+      "loss": 3.5856,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57915843085819e-05,
+      "loss": 3.5921,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.578321473987511e-05,
+      "loss": 3.6073,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577482879236459e-05,
+      "loss": 3.5842,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.576644284485408e-05,
+      "loss": 3.6128,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.575805689734356e-05,
+      "loss": 3.6043,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574968732863677e-05,
+      "loss": 3.583,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574130138112625e-05,
+      "loss": 3.6056,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573291543361573e-05,
+      "loss": 3.5867,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572452948610521e-05,
+      "loss": 3.5865,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571614353859469e-05,
+      "loss": 3.5954,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57077739698879e-05,
+      "loss": 3.6068,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569938802237738e-05,
+      "loss": 3.5815,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569100207486686e-05,
+      "loss": 3.5998,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568261612735634e-05,
+      "loss": 3.5809,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5674246558649546e-05,
+      "loss": 3.5998,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566586061113903e-05,
+      "loss": 3.6044,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565747466362851e-05,
+      "loss": 3.6039,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564908871611799e-05,
+      "loss": 3.6042,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56407191474112e-05,
+      "loss": 3.606,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563233319990068e-05,
+      "loss": 3.6228,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562394725239016e-05,
+      "loss": 3.6048,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561556130487964e-05,
+      "loss": 3.6081,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.560717535736912e-05,
+      "loss": 3.5917,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55987894098586e-05,
+      "loss": 3.5953,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559040346234808e-05,
+      "loss": 3.5968,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558203389364129e-05,
+      "loss": 3.6016,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557364794613077e-05,
+      "loss": 3.6061,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556526199862025e-05,
+      "loss": 3.6085,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555687605110973e-05,
+      "loss": 3.591,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5548506482402946e-05,
+      "loss": 3.5917,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5540120534892426e-05,
+      "loss": 3.5871,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5531734587381906e-05,
+      "loss": 3.6089,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5523348639871386e-05,
+      "loss": 3.5992,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5514962692360866e-05,
+      "loss": 3.5923,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550657674485034e-05,
+      "loss": 3.6025,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5498207176143555e-05,
+      "loss": 3.5962,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5489821228633035e-05,
+      "loss": 3.5921,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5481435281122515e-05,
+      "loss": 3.5934,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.547304933361199e-05,
+      "loss": 3.6059,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5464679764905204e-05,
+      "loss": 3.603,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5456293817394684e-05,
+      "loss": 3.6047,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5447907869884164e-05,
+      "loss": 3.5889,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5439521922373644e-05,
+      "loss": 3.5963,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543115235366686e-05,
+      "loss": 3.5956,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.542276640615634e-05,
+      "loss": 3.6037,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.541438045864581e-05,
+      "loss": 3.6051,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.540599451113529e-05,
+      "loss": 3.6008,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539760856362477e-05,
+      "loss": 3.6009,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538922261611425e-05,
+      "loss": 3.6,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538085304740746e-05,
+      "loss": 3.5977,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537246709989694e-05,
+      "loss": 3.5859,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536408115238642e-05,
+      "loss": 3.5898,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53556952048759e-05,
+      "loss": 3.5948,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534732563616912e-05,
+      "loss": 3.5942,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53389396886586e-05,
+      "loss": 3.5839,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533055374114808e-05,
+      "loss": 3.5863,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532216779363756e-05,
+      "loss": 3.5866,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5313798224930766e-05,
+      "loss": 3.5952,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5305412277420246e-05,
+      "loss": 3.5948,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5297026329909726e-05,
+      "loss": 3.5962,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5288640382399206e-05,
+      "loss": 3.6028,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5280270813692415e-05,
+      "loss": 3.6013,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5271884866181895e-05,
+      "loss": 3.5941,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5263498918671375e-05,
+      "loss": 3.6051,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5255112971160855e-05,
+      "loss": 3.597,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524672702365034e-05,
+      "loss": 3.6001,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523835745494355e-05,
+      "loss": 3.5941,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522997150743303e-05,
+      "loss": 3.5947,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522158555992251e-05,
+      "loss": 3.5958,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521319961241199e-05,
+      "loss": 3.5991,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520481366490147e-05,
+      "loss": 3.5898,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519644409619468e-05,
+      "loss": 3.5925,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518805814868416e-05,
+      "loss": 3.581,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517967220117364e-05,
+      "loss": 3.5877,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517128625366312e-05,
+      "loss": 3.597,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516291668495633e-05,
+      "loss": 3.5931,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515453073744581e-05,
+      "loss": 3.6046,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5146144789935296e-05,
+      "loss": 3.5826,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5137758842424776e-05,
+      "loss": 3.6043,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5129372894914256e-05,
+      "loss": 3.6014,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5120986947403736e-05,
+      "loss": 3.6009,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5112617378696945e-05,
+      "loss": 3.588,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5104231431186425e-05,
+      "loss": 3.5919,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5095845483675905e-05,
+      "loss": 3.6071,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5087459536165385e-05,
+      "loss": 3.6028,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5079073588654864e-05,
+      "loss": 3.5913,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5070704019948074e-05,
+      "loss": 3.5926,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5062318072437554e-05,
+      "loss": 3.5951,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5053932124927034e-05,
+      "loss": 3.5978,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5045546177416513e-05,
+      "loss": 3.5869,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.503717660870973e-05,
+      "loss": 3.5971,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502879066119921e-05,
+      "loss": 3.5925,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502040471368869e-05,
+      "loss": 3.5947,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501201876617817e-05,
+      "loss": 3.5984,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8442368507385254,
+      "eval_runtime": 305.2159,
+      "eval_samples_per_second": 1250.233,
+      "eval_steps_per_second": 39.071,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.500363281866765e-05,
+      "loss": 3.5936,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499524687115712e-05,
+      "loss": 3.577,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.49868609236466e-05,
+      "loss": 3.5914,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497847497613608e-05,
+      "loss": 3.5951,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497010540742929e-05,
+      "loss": 3.5953,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496171945991877e-05,
+      "loss": 3.5978,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495333351240825e-05,
+      "loss": 3.5922,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494494756489773e-05,
+      "loss": 3.5873,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493656161738722e-05,
+      "loss": 3.578,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492819204868043e-05,
+      "loss": 3.591,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491980610116991e-05,
+      "loss": 3.5935,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491142015365939e-05,
+      "loss": 3.5894,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4903050584952596e-05,
+      "loss": 3.5988,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4894664637442076e-05,
+      "loss": 3.5726,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4886278689931556e-05,
+      "loss": 3.5834,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4877892742421036e-05,
+      "loss": 3.5761,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4869506794910516e-05,
+      "loss": 3.5872,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4861120847399996e-05,
+      "loss": 3.5855,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4852734899889476e-05,
+      "loss": 3.5867,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4844348952378956e-05,
+      "loss": 3.5816,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4835979383672165e-05,
+      "loss": 3.6035,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482759343616165e-05,
+      "loss": 3.587,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481920748865113e-05,
+      "loss": 3.5925,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481082154114061e-05,
+      "loss": 3.5851,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480245197243382e-05,
+      "loss": 3.5942,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.47940660249233e-05,
+      "loss": 3.5916,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478568007741278e-05,
+      "loss": 3.5768,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477729412990226e-05,
+      "loss": 3.5883,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476890818239174e-05,
+      "loss": 3.5778,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476053861368495e-05,
+      "loss": 3.5809,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.475215266617443e-05,
+      "loss": 3.5918,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.474376671866391e-05,
+      "loss": 3.5879,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.473538077115339e-05,
+      "loss": 3.5848,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.472699482364287e-05,
+      "loss": 3.601,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4718625254936085e-05,
+      "loss": 3.5844,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4710239307425565e-05,
+      "loss": 3.5841,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4701853359915045e-05,
+      "loss": 3.5969,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4693467412404525e-05,
+      "loss": 3.578,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4685097843697734e-05,
+      "loss": 3.5848,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4676711896187214e-05,
+      "loss": 3.5853,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4668325948676694e-05,
+      "loss": 3.5808,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4659940001166174e-05,
+      "loss": 3.5898,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4651554053655654e-05,
+      "loss": 3.5944,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464318448494886e-05,
+      "loss": 3.5845,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463479853743834e-05,
+      "loss": 3.5901,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462641258992782e-05,
+      "loss": 3.5911,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46180266424173e-05,
+      "loss": 3.5849,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460965707371052e-05,
+      "loss": 3.5895,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46012711262e-05,
+      "loss": 3.5586,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459288517868948e-05,
+      "loss": 3.5851,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.458449923117895e-05,
+      "loss": 3.582,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.457612966247217e-05,
+      "loss": 3.591,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.456774371496165e-05,
+      "loss": 3.579,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455935776745113e-05,
+      "loss": 3.5691,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45509718199406e-05,
+      "loss": 3.5671,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.454260225123382e-05,
+      "loss": 3.5714,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45342163037233e-05,
+      "loss": 3.5905,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.452583035621278e-05,
+      "loss": 3.5644,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4517444408702257e-05,
+      "loss": 3.5936,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450907483999547e-05,
+      "loss": 3.5849,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450068889248495e-05,
+      "loss": 3.5695,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4492302944974426e-05,
+      "loss": 3.5868,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4483916997463906e-05,
+      "loss": 3.5692,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447554742875712e-05,
+      "loss": 3.5714,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44671614812466e-05,
+      "loss": 3.574,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4458775533736075e-05,
+      "loss": 3.5881,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4450389586225555e-05,
+      "loss": 3.5697,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4442003638715034e-05,
+      "loss": 3.5807,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443363407000825e-05,
+      "loss": 3.5604,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442524812249773e-05,
+      "loss": 3.5811,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.441686217498721e-05,
+      "loss": 3.5861,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440847622747669e-05,
+      "loss": 3.5868,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44001066587699e-05,
+      "loss": 3.5862,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439172071125938e-05,
+      "loss": 3.5883,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.438333476374886e-05,
+      "loss": 3.6023,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.437494881623834e-05,
+      "loss": 3.5874,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436657924753155e-05,
+      "loss": 3.591,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435819330002103e-05,
+      "loss": 3.5743,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434980735251051e-05,
+      "loss": 3.5767,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434142140499999e-05,
+      "loss": 3.5789,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433303545748947e-05,
+      "loss": 3.5812,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4324665888782684e-05,
+      "loss": 3.5875,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4316279941272164e-05,
+      "loss": 3.5913,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4307893993761644e-05,
+      "loss": 3.5764,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4299508046251124e-05,
+      "loss": 3.5709,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429113847754433e-05,
+      "loss": 3.57,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.428275253003381e-05,
+      "loss": 3.5902,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427436658252329e-05,
+      "loss": 3.5797,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.426598063501277e-05,
+      "loss": 3.5753,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425759468750225e-05,
+      "loss": 3.5838,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424920873999173e-05,
+      "loss": 3.5803,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424083917128494e-05,
+      "loss": 3.5732,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423245322377442e-05,
+      "loss": 3.5729,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.42240672762639e-05,
+      "loss": 3.5874,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421568132875338e-05,
+      "loss": 3.5887,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.42073117600466e-05,
+      "loss": 3.5848,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419892581253608e-05,
+      "loss": 3.5687,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419053986502556e-05,
+      "loss": 3.5787,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418215391751504e-05,
+      "loss": 3.5792,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417376797000452e-05,
+      "loss": 3.5821,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4165398401297726e-05,
+      "loss": 3.5896,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4157012453787206e-05,
+      "loss": 3.5805,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4148626506276686e-05,
+      "loss": 3.5808,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4140240558766166e-05,
+      "loss": 3.5815,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4131870990059375e-05,
+      "loss": 3.581,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4123485042548855e-05,
+      "loss": 3.5697,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4115099095038335e-05,
+      "loss": 3.5731,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.410671314752782e-05,
+      "loss": 3.5787,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409834357882103e-05,
+      "loss": 3.5772,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408995763131051e-05,
+      "loss": 3.5617,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408157168379999e-05,
+      "loss": 3.5726,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407318573628947e-05,
+      "loss": 3.5616,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406481616758268e-05,
+      "loss": 3.5819,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405643022007216e-05,
+      "loss": 3.5758,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404804427256164e-05,
+      "loss": 3.5785,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403965832505112e-05,
+      "loss": 3.5782,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403128875634433e-05,
+      "loss": 3.5848,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402290280883381e-05,
+      "loss": 3.5787,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401451686132329e-05,
+      "loss": 3.5811,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4006130913812776e-05,
+      "loss": 3.5835,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3997761345105985e-05,
+      "loss": 3.5806,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3989375397595465e-05,
+      "loss": 3.576,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3980989450084945e-05,
+      "loss": 3.5784,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3972603502574425e-05,
+      "loss": 3.5786,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3964233933867634e-05,
+      "loss": 3.5783,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3955847986357114e-05,
+      "loss": 3.5726,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3947462038846594e-05,
+      "loss": 3.5753,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3939076091336074e-05,
+      "loss": 3.5631,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3930690143825554e-05,
+      "loss": 3.5715,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392232057511876e-05,
+      "loss": 3.5781,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.391393462760824e-05,
+      "loss": 3.5785,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.390554868009773e-05,
+      "loss": 3.5848,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.389716273258721e-05,
+      "loss": 3.5717,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.388879316388042e-05,
+      "loss": 3.5847,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38804072163699e-05,
+      "loss": 3.5845,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387202126885938e-05,
+      "loss": 3.5777,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.386363532134886e-05,
+      "loss": 3.5752,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.385526575264207e-05,
+      "loss": 3.5759,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.384687980513155e-05,
+      "loss": 3.5902,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383849385762103e-05,
+      "loss": 3.5833,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38301079101105e-05,
+      "loss": 3.5762,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3821738341403716e-05,
+      "loss": 3.5766,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3813352393893196e-05,
+      "loss": 3.5768,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380496644638268e-05,
+      "loss": 3.5764,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3796580498872156e-05,
+      "loss": 3.5718,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3788194551361636e-05,
+      "loss": 3.5768,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377982498265485e-05,
+      "loss": 3.5741,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377143903514433e-05,
+      "loss": 3.5798,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3763053087633805e-05,
+      "loss": 3.5791,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8442249298095703,
+      "eval_runtime": 302.6004,
+      "eval_samples_per_second": 1261.039,
+      "eval_steps_per_second": 39.408,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3754667140123285e-05,
+      "loss": 3.5751,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3746281192612765e-05,
+      "loss": 3.5613,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3737911623905974e-05,
+      "loss": 3.5697,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3729525676395454e-05,
+      "loss": 3.5817,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3721139728884934e-05,
+      "loss": 3.5788,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.371275378137442e-05,
+      "loss": 3.5814,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.370438421266763e-05,
+      "loss": 3.5703,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.369599826515711e-05,
+      "loss": 3.5711,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.368761231764659e-05,
+      "loss": 3.5607,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367922637013607e-05,
+      "loss": 3.5766,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367084042262555e-05,
+      "loss": 3.5721,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366247085391876e-05,
+      "loss": 3.5743,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3654101285211975e-05,
+      "loss": 3.5801,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.364571533770145e-05,
+      "loss": 3.558,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363732939019093e-05,
+      "loss": 3.5684,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362894344268041e-05,
+      "loss": 3.5544,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362055749516989e-05,
+      "loss": 3.5724,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3612171547659375e-05,
+      "loss": 3.5653,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3603785600148854e-05,
+      "loss": 3.5735,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3595399652638334e-05,
+      "loss": 3.5633,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3587030083931544e-05,
+      "loss": 3.5854,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3578644136421024e-05,
+      "loss": 3.5718,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3570258188910503e-05,
+      "loss": 3.5744,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3561872241399983e-05,
+      "loss": 3.5698,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.355350267269319e-05,
+      "loss": 3.578,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.354511672518267e-05,
+      "loss": 3.5736,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.353673077767215e-05,
+      "loss": 3.5626,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.352834483016163e-05,
+      "loss": 3.5723,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351995888265111e-05,
+      "loss": 3.5586,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351158931394433e-05,
+      "loss": 3.5626,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.350320336643381e-05,
+      "loss": 3.5707,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.349481741892329e-05,
+      "loss": 3.5679,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.348643147141277e-05,
+      "loss": 3.5688,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.347806190270598e-05,
+      "loss": 3.582,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346967595519546e-05,
+      "loss": 3.566,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346129000768494e-05,
+      "loss": 3.5719,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.345290406017442e-05,
+      "loss": 3.5739,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3444534491467626e-05,
+      "loss": 3.5624,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3436148543957106e-05,
+      "loss": 3.5717,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3427762596446586e-05,
+      "loss": 3.5647,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3419376648936066e-05,
+      "loss": 3.5626,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3410990701425546e-05,
+      "loss": 3.5716,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.340262113271876e-05,
+      "loss": 3.5799,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339423518520824e-05,
+      "loss": 3.5647,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.338584923769772e-05,
+      "loss": 3.5726,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.33774632901872e-05,
+      "loss": 3.5748,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336909372148041e-05,
+      "loss": 3.5685,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336070777396989e-05,
+      "loss": 3.5731,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335232182645937e-05,
+      "loss": 3.5391,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334393587894885e-05,
+      "loss": 3.5695,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.333556631024206e-05,
+      "loss": 3.5643,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.332718036273154e-05,
+      "loss": 3.5755,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331879441522102e-05,
+      "loss": 3.5622,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.33104084677105e-05,
+      "loss": 3.5522,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3302038899003715e-05,
+      "loss": 3.5472,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3293652951493195e-05,
+      "loss": 3.5509,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3285267003982675e-05,
+      "loss": 3.5763,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3276881056472155e-05,
+      "loss": 3.5503,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3268511487765364e-05,
+      "loss": 3.574,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3260125540254844e-05,
+      "loss": 3.5658,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3251739592744324e-05,
+      "loss": 3.5561,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3243353645233804e-05,
+      "loss": 3.5687,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3234984076527013e-05,
+      "loss": 3.5482,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.322659812901649e-05,
+      "loss": 3.5545,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.321821218150597e-05,
+      "loss": 3.5566,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320982623399545e-05,
+      "loss": 3.5727,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320144028648493e-05,
+      "loss": 3.5536,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319307071777815e-05,
+      "loss": 3.5595,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318468477026763e-05,
+      "loss": 3.5489,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317629882275711e-05,
+      "loss": 3.5612,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316791287524658e-05,
+      "loss": 3.5706,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.31595433065398e-05,
+      "loss": 3.5682,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315115735902928e-05,
+      "loss": 3.5722,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.314277141151876e-05,
+      "loss": 3.5707,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.313438546400823e-05,
+      "loss": 3.5845,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.312601589530145e-05,
+      "loss": 3.5695,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.311762994779093e-05,
+      "loss": 3.5738,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310924400028041e-05,
+      "loss": 3.56,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310085805276989e-05,
+      "loss": 3.5616,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30924884840631e-05,
+      "loss": 3.5632,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.308410253655258e-05,
+      "loss": 3.5609,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3075716589042056e-05,
+      "loss": 3.5694,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3067330641531536e-05,
+      "loss": 3.5731,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.305896107282475e-05,
+      "loss": 3.5677,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.305057512531423e-05,
+      "loss": 3.5484,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3042189177803705e-05,
+      "loss": 3.5544,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3033803230293185e-05,
+      "loss": 3.57,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3025417282782665e-05,
+      "loss": 3.5662,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3017031335272145e-05,
+      "loss": 3.5596,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.300866176656536e-05,
+      "loss": 3.5663,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.300027581905484e-05,
+      "loss": 3.5652,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299188987154432e-05,
+      "loss": 3.5561,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29835039240338e-05,
+      "loss": 3.5603,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.297513435532701e-05,
+      "loss": 3.5681,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.296674840781649e-05,
+      "loss": 3.5708,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295836246030597e-05,
+      "loss": 3.5694,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294997651279545e-05,
+      "loss": 3.5517,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294160694408866e-05,
+      "loss": 3.5598,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.293322099657814e-05,
+      "loss": 3.5649,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.292483504906762e-05,
+      "loss": 3.5681,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29164491015571e-05,
+      "loss": 3.5701,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2908079532850314e-05,
+      "loss": 3.5651,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2899693585339794e-05,
+      "loss": 3.5628,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2891307637829274e-05,
+      "loss": 3.5628,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2882921690318754e-05,
+      "loss": 3.5656,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2874535742808234e-05,
+      "loss": 3.5533,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.286616617410144e-05,
+      "loss": 3.5519,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.285778022659092e-05,
+      "loss": 3.5632,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.28493942790804e-05,
+      "loss": 3.5627,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284100833156988e-05,
+      "loss": 3.5472,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.283263876286309e-05,
+      "loss": 3.5548,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.282425281535257e-05,
+      "loss": 3.5496,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.281586686784205e-05,
+      "loss": 3.5606,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.280748092033153e-05,
+      "loss": 3.5598,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279911135162475e-05,
+      "loss": 3.5592,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279072540411423e-05,
+      "loss": 3.5638,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278233945660371e-05,
+      "loss": 3.5703,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.277395350909319e-05,
+      "loss": 3.5617,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.27655839403864e-05,
+      "loss": 3.5647,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.275719799287588e-05,
+      "loss": 3.5693,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.274881204536536e-05,
+      "loss": 3.5619,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.274042609785484e-05,
+      "loss": 3.5575,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2732056529148046e-05,
+      "loss": 3.5642,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2723670581637526e-05,
+      "loss": 3.5617,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2715284634127006e-05,
+      "loss": 3.5593,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2706898686616486e-05,
+      "loss": 3.554,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26985291179097e-05,
+      "loss": 3.5606,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269014317039918e-05,
+      "loss": 3.5525,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268175722288866e-05,
+      "loss": 3.5539,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.267337127537814e-05,
+      "loss": 3.5546,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.266500170667135e-05,
+      "loss": 3.563,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.265661575916083e-05,
+      "loss": 3.5648,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.264822981165031e-05,
+      "loss": 3.5556,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263984386413979e-05,
+      "loss": 3.5685,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2631474295433e-05,
+      "loss": 3.5695,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.262308834792248e-05,
+      "loss": 3.5669,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.261470240041196e-05,
+      "loss": 3.5605,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.260631645290144e-05,
+      "loss": 3.5566,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2597946884194655e-05,
+      "loss": 3.5744,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2589560936684135e-05,
+      "loss": 3.5579,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2581174989173615e-05,
+      "loss": 3.5666,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2572789041663095e-05,
+      "loss": 3.5597,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2564419472956304e-05,
+      "loss": 3.5612,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2556033525445784e-05,
+      "loss": 3.5584,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2547647577935264e-05,
+      "loss": 3.5564,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2539261630424744e-05,
+      "loss": 3.5642,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253089206171795e-05,
+      "loss": 3.5575,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252250611420743e-05,
+      "loss": 3.5634,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.251412016669691e-05,
+      "loss": 3.5605,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.845308303833008,
+      "eval_runtime": 308.404,
+      "eval_samples_per_second": 1237.309,
+      "eval_steps_per_second": 38.667,
+      "step": 1068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.250573421918639e-05,
+      "loss": 3.5535,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.249734827167587e-05,
+      "loss": 3.547,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.248896232416535e-05,
+      "loss": 3.5514,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.248057637665484e-05,
+      "loss": 3.5664,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247220680794805e-05,
+      "loss": 3.5596,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.246382086043753e-05,
+      "loss": 3.5663,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.245543491292701e-05,
+      "loss": 3.5521,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.244704896541649e-05,
+      "loss": 3.5548,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.24386793967097e-05,
+      "loss": 3.545,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.243029344919918e-05,
+      "loss": 3.5572,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242190750168866e-05,
+      "loss": 3.5584,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.241352155417813e-05,
+      "loss": 3.5596,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2405168364275076e-05,
+      "loss": 3.5639,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.239678241676456e-05,
+      "loss": 3.5458,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238839646925404e-05,
+      "loss": 3.5457,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238001052174352e-05,
+      "loss": 3.5437,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2371624574233e-05,
+      "loss": 3.5528,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.236323862672248e-05,
+      "loss": 3.5487,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.235485267921196e-05,
+      "loss": 3.5585,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2346466731701436e-05,
+      "loss": 3.5478,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.233809716299465e-05,
+      "loss": 3.5679,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232971121548413e-05,
+      "loss": 3.5554,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2321325267973605e-05,
+      "loss": 3.553,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2312939320463085e-05,
+      "loss": 3.5561,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.23045697517563e-05,
+      "loss": 3.5658,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.229618380424578e-05,
+      "loss": 3.5541,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.228779785673526e-05,
+      "loss": 3.5475,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227941190922474e-05,
+      "loss": 3.5551,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227102596171422e-05,
+      "loss": 3.541,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2262656393007436e-05,
+      "loss": 3.546,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.225427044549691e-05,
+      "loss": 3.5528,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.224588449798639e-05,
+      "loss": 3.5531,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223749855047587e-05,
+      "loss": 3.5563,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222912898176908e-05,
+      "loss": 3.5627,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222074303425856e-05,
+      "loss": 3.5546,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221235708674804e-05,
+      "loss": 3.5548,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220397113923752e-05,
+      "loss": 3.557,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2195601570530734e-05,
+      "loss": 3.5474,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2187215623020214e-05,
+      "loss": 3.5579,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2178829675509694e-05,
+      "loss": 3.5479,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2170443727999174e-05,
+      "loss": 3.5479,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216207415929238e-05,
+      "loss": 3.5531,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.215368821178186e-05,
+      "loss": 3.5647,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.214530226427134e-05,
+      "loss": 3.547,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.213691631676082e-05,
+      "loss": 3.5627,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212854674805403e-05,
+      "loss": 3.5559,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212016080054351e-05,
+      "loss": 3.5548,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211177485303299e-05,
+      "loss": 3.5557,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.210338890552247e-05,
+      "loss": 3.5243,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.209500295801195e-05,
+      "loss": 3.5504,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.208663338930517e-05,
+      "loss": 3.5552,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207824744179465e-05,
+      "loss": 3.5546,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206986149428413e-05,
+      "loss": 3.5458,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206147554677361e-05,
+      "loss": 3.5394,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2053105978066817e-05,
+      "loss": 3.5311,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2044720030556297e-05,
+      "loss": 3.5363,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2036334083045776e-05,
+      "loss": 3.5631,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2027948135535256e-05,
+      "loss": 3.533,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2019578566828466e-05,
+      "loss": 3.5567,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2011192619317946e-05,
+      "loss": 3.549,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2002806671807425e-05,
+      "loss": 3.5378,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1994420724296905e-05,
+      "loss": 3.5525,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.198605115559012e-05,
+      "loss": 3.5377,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.19776652080796e-05,
+      "loss": 3.5417,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.196927926056908e-05,
+      "loss": 3.5376,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.196089331305856e-05,
+      "loss": 3.5558,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.195252374435177e-05,
+      "loss": 3.5424,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.194413779684125e-05,
+      "loss": 3.5421,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.193575184933073e-05,
+      "loss": 3.5351,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.192736590182021e-05,
+      "loss": 3.5449,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191899633311342e-05,
+      "loss": 3.5538,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.19106103856029e-05,
+      "loss": 3.5511,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190222443809238e-05,
+      "loss": 3.5526,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.189383849058186e-05,
+      "loss": 3.5574,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1885452543071346e-05,
+      "loss": 3.5681,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1877082974364555e-05,
+      "loss": 3.5545,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1868697026854035e-05,
+      "loss": 3.5573,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1860311079343515e-05,
+      "loss": 3.5463,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1851925131832995e-05,
+      "loss": 3.5463,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1843555563126204e-05,
+      "loss": 3.5446,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1835169615615684e-05,
+      "loss": 3.5428,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1826783668105164e-05,
+      "loss": 3.556,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1818397720594644e-05,
+      "loss": 3.5592,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181002815188785e-05,
+      "loss": 3.5525,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180164220437733e-05,
+      "loss": 3.5291,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.179325625686681e-05,
+      "loss": 3.5406,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.178487030935629e-05,
+      "loss": 3.5523,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.177648436184578e-05,
+      "loss": 3.5515,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.176809841433526e-05,
+      "loss": 3.5408,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.175972884562847e-05,
+      "loss": 3.5525,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.175134289811795e-05,
+      "loss": 3.5462,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.174295695060743e-05,
+      "loss": 3.5396,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.173457100309691e-05,
+      "loss": 3.5428,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.172620143439012e-05,
+      "loss": 3.5552,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.17178154868796e-05,
+      "loss": 3.5469,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170942953936908e-05,
+      "loss": 3.5566,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170104359185856e-05,
+      "loss": 3.5362,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169265764434804e-05,
+      "loss": 3.5435,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1684288075641246e-05,
+      "loss": 3.5519,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.167590212813073e-05,
+      "loss": 3.5505,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.166751618062021e-05,
+      "loss": 3.5551,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1659130233109686e-05,
+      "loss": 3.5495,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16507606644029e-05,
+      "loss": 3.5474,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164237471689238e-05,
+      "loss": 3.5473,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.163398876938186e-05,
+      "loss": 3.5534,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1625602821871335e-05,
+      "loss": 3.5333,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.161723325316455e-05,
+      "loss": 3.5409,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160884730565403e-05,
+      "loss": 3.5432,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160046135814351e-05,
+      "loss": 3.5448,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1592075410632984e-05,
+      "loss": 3.5362,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.15837058419262e-05,
+      "loss": 3.5379,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157531989441569e-05,
+      "loss": 3.5302,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.156693394690516e-05,
+      "loss": 3.5496,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.155854799939464e-05,
+      "loss": 3.5428,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1550178430687856e-05,
+      "loss": 3.5442,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1541792483177336e-05,
+      "loss": 3.548,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.153340653566681e-05,
+      "loss": 3.5526,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.152502058815629e-05,
+      "loss": 3.5469,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1516651019449505e-05,
+      "loss": 3.5458,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1508265071938985e-05,
+      "loss": 3.5561,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.149987912442846e-05,
+      "loss": 3.5459,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.149149317691794e-05,
+      "loss": 3.5396,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1483107229407425e-05,
+      "loss": 3.5476,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1474737660700634e-05,
+      "loss": 3.5452,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1466351713190114e-05,
+      "loss": 3.5474,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1457965765679594e-05,
+      "loss": 3.5366,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1449579818169074e-05,
+      "loss": 3.5472,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144121024946228e-05,
+      "loss": 3.5373,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.143282430195176e-05,
+      "loss": 3.5408,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.142443835444124e-05,
+      "loss": 3.5332,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.141605240693072e-05,
+      "loss": 3.553,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.140768283822393e-05,
+      "loss": 3.5498,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.139929689071341e-05,
+      "loss": 3.5342,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.139091094320289e-05,
+      "loss": 3.5535,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138252499569238e-05,
+      "loss": 3.5511,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.137415542698559e-05,
+      "loss": 3.5511,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.136576947947507e-05,
+      "loss": 3.5428,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.135738353196455e-05,
+      "loss": 3.5449,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.134899758445403e-05,
+      "loss": 3.5598,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.134061163694351e-05,
+      "loss": 3.5426,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1332242068236716e-05,
+      "loss": 3.5512,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1323856120726196e-05,
+      "loss": 3.5428,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1315470173215676e-05,
+      "loss": 3.5462,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1307084225705156e-05,
+      "loss": 3.5412,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1298714656998365e-05,
+      "loss": 3.5385,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1290328709487845e-05,
+      "loss": 3.5467,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128194276197733e-05,
+      "loss": 3.5434,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.127355681446681e-05,
+      "loss": 3.5505,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.126518724576002e-05,
+      "loss": 3.5437,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.845208168029785,
+      "eval_runtime": 303.7344,
+      "eval_samples_per_second": 1256.331,
+      "eval_steps_per_second": 39.261,
+      "step": 1144800
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 7.961627666393856e+17,
+  "trial_name": null,
+  "trial_params": null
+}