diff --git "a/checkpoint-915840/trainer_state.json" "b/checkpoint-915840/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-915840/trainer_state.json"
@@ -0,0 +1,10849 @@
+{
+  "best_metric": 3.840613603591919,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/npi-only/transformer/1/checkpoints/checkpoint-915840",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 915840,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.9232,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8136,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1808,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.976,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8215,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.713,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.61,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5443,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4534,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4147,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3633,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3228,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.2592,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2219,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1884,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1394,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1226,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.089,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0496,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0283,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0114,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.981,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9633,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9298,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9221,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.9022,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8784,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.857,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8314,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756873037409844e-05,
+      "loss": 4.824,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.8069,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.8009,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7843,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7694,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7615,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7503,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.739,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 4.7088,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 4.707,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 4.6818,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 4.6879,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6677,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6699,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.649,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6457,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6376,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6291,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6187,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.5937,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.6004,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.601,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.5946,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.5739,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5624,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 4.5642,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 4.55,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 4.5649,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5261,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5434,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5326,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5126,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5224,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5079,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4979,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4943,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5034,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.4827,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4789,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.471,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.4655,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4722,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940482703000532e-05,
+      "loss": 4.477,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396441082494794e-05,
+      "loss": 4.4526,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388055134984274e-05,
+      "loss": 4.4636,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379669187473754e-05,
+      "loss": 4.4666,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 4.4495,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 4.4365,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 4.4387,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 4.4364,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 4.4224,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 4.4195,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 4.4287,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.4282,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 4.4123,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.3904,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.4031,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.4074,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 4.3974,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926231505873767e-05,
+      "loss": 4.3991,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925394549003088e-05,
+      "loss": 4.3821,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924555954252036e-05,
+      "loss": 4.3804,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923717359500984e-05,
+      "loss": 4.3754,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922878764749932e-05,
+      "loss": 4.3731,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 4.3787,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.3669,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920364618377149e-05,
+      "loss": 4.3841,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919526023626097e-05,
+      "loss": 4.3662,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918687428875045e-05,
+      "loss": 4.3504,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917848834123993e-05,
+      "loss": 4.352,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917010239372941e-05,
+      "loss": 4.3593,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.351,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.3494,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3473,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913657498249106e-05,
+      "loss": 4.337,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912818903498054e-05,
+      "loss": 4.3398,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911980308747002e-05,
+      "loss": 4.3189,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.3246,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.3298,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3238,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.3181,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.3261,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3007,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3122,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052748264993326e-05,
+      "loss": 4.3079,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90443623174828e-05,
+      "loss": 4.3116,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.3183,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.3073,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019237232558704e-05,
+      "loss": 4.2984,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010851285048184e-05,
+      "loss": 4.2929,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900246533753767e-05,
+      "loss": 4.3001,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899407939002715e-05,
+      "loss": 4.2944,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985693442516624e-05,
+      "loss": 4.2926,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977307495006104e-05,
+      "loss": 4.2851,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968921547495584e-05,
+      "loss": 4.2819,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.2825,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2709,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2795,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2758,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.2697,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2613,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2795,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2796,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2674,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.2552,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887670888248733e-05,
+      "loss": 4.2636,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886832293497681e-05,
+      "loss": 4.2639,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.2627,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.2588,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2586,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.2527,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.2502,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2447,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.2488,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.2398,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8792882164989586e-05,
+      "loss": 4.2344,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2476,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877611026996855e-05,
+      "loss": 4.2317,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876772432245803e-05,
+      "loss": 4.2424,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875933837494751e-05,
+      "loss": 4.2295,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875096880624072e-05,
+      "loss": 4.224,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.19596529006958,
+      "eval_runtime": 307.6211,
+      "eval_samples_per_second": 1240.458,
+      "eval_steps_per_second": 38.765,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87425828587302e-05,
+      "loss": 4.2161,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873419691121968e-05,
+      "loss": 4.2145,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872581096370916e-05,
+      "loss": 4.2271,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717425016198635e-05,
+      "loss": 4.2185,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709039068688115e-05,
+      "loss": 4.2223,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870066949998133e-05,
+      "loss": 4.2098,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869228355247081e-05,
+      "loss": 4.2089,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.868389760496029e-05,
+      "loss": 4.1976,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867551165744977e-05,
+      "loss": 4.2139,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866712570993925e-05,
+      "loss": 4.2054,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865875614123246e-05,
+      "loss": 4.2174,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865037019372194e-05,
+      "loss": 4.199,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642000625015155e-05,
+      "loss": 4.1959,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633614677504635e-05,
+      "loss": 4.1895,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862522872999411e-05,
+      "loss": 4.1857,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861684278248359e-05,
+      "loss": 4.1848,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.1926,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860007088746255e-05,
+      "loss": 4.181,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859168493995203e-05,
+      "loss": 4.1791,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.1997,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574929423734724e-05,
+      "loss": 4.1842,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566543476224204e-05,
+      "loss": 4.1846,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558157528713684e-05,
+      "loss": 4.1733,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549771581203164e-05,
+      "loss": 4.186,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854140201249637e-05,
+      "loss": 4.1715,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853301606498585e-05,
+      "loss": 4.1731,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852463011747533e-05,
+      "loss": 4.1706,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851624416996481e-05,
+      "loss": 4.1629,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850787460125802e-05,
+      "loss": 4.159,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84994886537475e-05,
+      "loss": 4.158,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849110270623698e-05,
+      "loss": 4.1673,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848271675872646e-05,
+      "loss": 4.1656,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847434719001968e-05,
+      "loss": 4.1599,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846596124250916e-05,
+      "loss": 4.1656,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845757529499864e-05,
+      "loss": 4.1609,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844918934748812e-05,
+      "loss": 4.1605,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844081977878133e-05,
+      "loss": 4.1478,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843243383127081e-05,
+      "loss": 4.1481,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842404788376029e-05,
+      "loss": 4.1373,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841566193624977e-05,
+      "loss": 4.15,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840727598873925e-05,
+      "loss": 4.1347,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398906420032456e-05,
+      "loss": 4.1489,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8390520472521936e-05,
+      "loss": 4.1351,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382134525011416e-05,
+      "loss": 4.1394,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83737485775009e-05,
+      "loss": 4.1396,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836537900879411e-05,
+      "loss": 4.1371,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835699306128359e-05,
+      "loss": 4.1348,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834860711377307e-05,
+      "loss": 4.1203,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834022116626255e-05,
+      "loss": 4.1246,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833185159755576e-05,
+      "loss": 4.1364,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832346565004524e-05,
+      "loss": 4.1332,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831507970253472e-05,
+      "loss": 4.1238,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83066937550242e-05,
+      "loss": 4.1138,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829832418631741e-05,
+      "loss": 4.1229,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828993823880689e-05,
+      "loss": 4.11,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828155229129637e-05,
+      "loss": 4.1329,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273166343785856e-05,
+      "loss": 4.1041,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264796775079065e-05,
+      "loss": 4.1261,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256410827568545e-05,
+      "loss": 4.1179,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248024880058025e-05,
+      "loss": 4.1028,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239638932547505e-05,
+      "loss": 4.1151,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231269363840714e-05,
+      "loss": 4.1112,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222883416330194e-05,
+      "loss": 4.0991,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214497468819674e-05,
+      "loss": 4.1038,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206111521309154e-05,
+      "loss": 4.1144,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819774195260236e-05,
+      "loss": 4.1019,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818935600509184e-05,
+      "loss": 4.1004,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818097005758132e-05,
+      "loss": 4.0964,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817258411007081e-05,
+      "loss": 4.0942,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816421454136402e-05,
+      "loss": 4.1035,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81558285938535e-05,
+      "loss": 4.1158,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814744264634298e-05,
+      "loss": 4.0961,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813905669883246e-05,
+      "loss": 4.1094,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813068713012567e-05,
+      "loss": 4.1096,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812230118261515e-05,
+      "loss": 4.0974,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811391523510463e-05,
+      "loss": 4.089,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810552928759411e-05,
+      "loss": 4.0998,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809715971888732e-05,
+      "loss": 4.0974,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80887737713768e-05,
+      "loss": 4.088,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808038782386628e-05,
+      "loss": 4.0883,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807200187635576e-05,
+      "loss": 4.0946,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806363230764897e-05,
+      "loss": 4.1017,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805524636013845e-05,
+      "loss": 4.0884,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804686041262793e-05,
+      "loss": 4.0768,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803847446511741e-05,
+      "loss": 4.082,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803010489641062e-05,
+      "loss": 4.0887,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80217189489001e-05,
+      "loss": 4.0863,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801333300138958e-05,
+      "loss": 4.0892,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800496343268279e-05,
+      "loss": 4.0766,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799657748517227e-05,
+      "loss": 4.0722,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798819153766175e-05,
+      "loss": 4.0715,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797980559015123e-05,
+      "loss": 4.0741,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797141964264071e-05,
+      "loss": 4.0767,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796303369513019e-05,
+      "loss": 4.0739,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795464774761967e-05,
+      "loss": 4.094,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794626180010915e-05,
+      "loss": 4.0777,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793789223140236e-05,
+      "loss": 4.061,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792950628389184e-05,
+      "loss": 4.0699,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792112033638132e-05,
+      "loss": 4.0729,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79127343888708e-05,
+      "loss": 4.0726,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790436482016401e-05,
+      "loss": 4.0715,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789597887265349e-05,
+      "loss": 4.068,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788759292514297e-05,
+      "loss": 4.0614,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787920697763245e-05,
+      "loss": 4.0714,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870837408925664e-05,
+      "loss": 4.0442,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862451461415144e-05,
+      "loss": 4.0587,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785408189270836e-05,
+      "loss": 4.0595,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784569594519783e-05,
+      "loss": 4.0604,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783730999768731e-05,
+      "loss": 4.051,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782892405017679e-05,
+      "loss": 4.0618,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782053810266627e-05,
+      "loss": 4.0482,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781215215515575e-05,
+      "loss": 4.052,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780376620764523e-05,
+      "loss": 4.0537,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779539663893844e-05,
+      "loss": 4.054,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778701069142792e-05,
+      "loss": 4.0698,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77786247439174e-05,
+      "loss": 4.0596,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777025517521062e-05,
+      "loss": 4.0532,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77618692277001e-05,
+      "loss": 4.048,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775348328018958e-05,
+      "loss": 4.0512,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774509733267906e-05,
+      "loss": 4.0502,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773671138516854e-05,
+      "loss": 4.0467,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772832543765802e-05,
+      "loss": 4.046,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77199394901475e-05,
+      "loss": 4.0423,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771155354263698e-05,
+      "loss": 4.0469,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703183973930186e-05,
+      "loss": 4.035,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694798026419666e-05,
+      "loss": 4.0426,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686412078909146e-05,
+      "loss": 4.0393,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7678042510202355e-05,
+      "loss": 4.0414,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766965656269184e-05,
+      "loss": 4.0281,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766127061518132e-05,
+      "loss": 4.0519,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76528846676708e-05,
+      "loss": 4.0558,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764451509896401e-05,
+      "loss": 4.0378,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763612915145349e-05,
+      "loss": 4.0316,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762774320394297e-05,
+      "loss": 4.0392,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761935725643245e-05,
+      "loss": 4.0414,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761098768772566e-05,
+      "loss": 4.042,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760260174021514e-05,
+      "loss": 4.038,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759421579270462e-05,
+      "loss": 4.0379,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758584622399783e-05,
+      "loss": 4.0331,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757746027648731e-05,
+      "loss": 4.0389,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569074328976796e-05,
+      "loss": 4.0273,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560688381466276e-05,
+      "loss": 4.0357,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552302433955756e-05,
+      "loss": 4.0321,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543932865248965e-05,
+      "loss": 4.0209,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535546917738445e-05,
+      "loss": 4.0379,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527160970227925e-05,
+      "loss": 4.0249,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518775022717405e-05,
+      "loss": 4.0343,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510389075206885e-05,
+      "loss": 4.0203,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7502003127696365e-05,
+      "loss": 4.021,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.025944232940674,
+      "eval_runtime": 312.8015,
+      "eval_samples_per_second": 1219.914,
+      "eval_steps_per_second": 38.123,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493617180185845e-05,
+      "loss": 4.0152,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748523123267532e-05,
+      "loss": 4.0082,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74768452851648e-05,
+      "loss": 4.0288,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746845933765428e-05,
+      "loss": 4.0195,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460073390143764e-05,
+      "loss": 4.0285,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451687442633244e-05,
+      "loss": 4.0096,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443301495122724e-05,
+      "loss": 4.0167,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434915547612204e-05,
+      "loss": 4.0008,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426529600101684e-05,
+      "loss": 4.0183,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418143652591164e-05,
+      "loss": 4.0191,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409757705080644e-05,
+      "loss": 4.0196,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740138813637385e-05,
+      "loss": 4.0122,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739301856766706e-05,
+      "loss": 4.004,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738463262015654e-05,
+      "loss": 4.0029,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737624667264602e-05,
+      "loss": 3.9995,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73678607251355e-05,
+      "loss": 4.0008,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735947477762498e-05,
+      "loss": 4.0031,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735108883011446e-05,
+      "loss": 3.9981,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734270288260394e-05,
+      "loss": 3.9991,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733433331389716e-05,
+      "loss": 4.0189,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732594736638664e-05,
+      "loss": 4.0015,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731756141887612e-05,
+      "loss": 4.0076,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730919185016933e-05,
+      "loss": 3.9948,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730080590265881e-05,
+      "loss": 4.0107,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729241995514829e-05,
+      "loss": 3.9874,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728403400763777e-05,
+      "loss": 3.9989,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727564806012725e-05,
+      "loss": 3.9967,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726726211261673e-05,
+      "loss": 3.9934,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258892543909936e-05,
+      "loss": 3.9809,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250506596399416e-05,
+      "loss": 3.9872,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7242120648888896e-05,
+      "loss": 3.9963,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723373470137838e-05,
+      "loss": 3.9972,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225348753867856e-05,
+      "loss": 3.9931,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216962806357336e-05,
+      "loss": 3.9947,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208576858846816e-05,
+      "loss": 3.9912,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7200190911336296e-05,
+      "loss": 3.9972,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191821342629505e-05,
+      "loss": 3.9815,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183435395118985e-05,
+      "loss": 3.9819,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175049447608465e-05,
+      "loss": 3.9722,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716667987890168e-05,
+      "loss": 3.9896,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158293931391154e-05,
+      "loss": 3.9758,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149907983880634e-05,
+      "loss": 3.9858,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714152203637012e-05,
+      "loss": 3.9783,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71331360888596e-05,
+      "loss": 3.9749,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712476652015281e-05,
+      "loss": 3.9874,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711638057264229e-05,
+      "loss": 3.9756,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710799462513177e-05,
+      "loss": 3.9774,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709960867762125e-05,
+      "loss": 3.9596,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709122273011073e-05,
+      "loss": 3.9712,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708283678260021e-05,
+      "loss": 3.9803,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707445083508969e-05,
+      "loss": 3.9751,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70660812663829e-05,
+      "loss": 3.9738,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705769531887238e-05,
+      "loss": 3.9593,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704930937136186e-05,
+      "loss": 3.9722,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704092342385134e-05,
+      "loss": 3.9577,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703253747634082e-05,
+      "loss": 3.9814,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024151528830305e-05,
+      "loss": 3.955,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015781960123514e-05,
+      "loss": 3.9744,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007396012612994e-05,
+      "loss": 3.972,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6999010065102474e-05,
+      "loss": 3.953,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990624117591954e-05,
+      "loss": 3.963,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982238170081434e-05,
+      "loss": 3.9669,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973852222570914e-05,
+      "loss": 3.953,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696548265386412e-05,
+      "loss": 3.9589,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69570967063536e-05,
+      "loss": 3.9676,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694871075884308e-05,
+      "loss": 3.964,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694032481133256e-05,
+      "loss": 3.9518,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693193886382204e-05,
+      "loss": 3.9563,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692355291631152e-05,
+      "loss": 3.9486,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915166968801e-05,
+      "loss": 3.9609,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690678102129048e-05,
+      "loss": 3.9739,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689842783138743e-05,
+      "loss": 3.9575,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689004188387691e-05,
+      "loss": 3.9697,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688165593636639e-05,
+      "loss": 3.968,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687326998885587e-05,
+      "loss": 3.9614,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686488404134534e-05,
+      "loss": 3.9488,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856514472638556e-05,
+      "loss": 3.9646,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848128525128036e-05,
+      "loss": 3.9557,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683974257761751e-05,
+      "loss": 3.9564,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831356630106996e-05,
+      "loss": 3.9518,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822970682596476e-05,
+      "loss": 3.9574,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681460111388969e-05,
+      "loss": 3.9695,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806215166379165e-05,
+      "loss": 3.9526,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6797829218868645e-05,
+      "loss": 3.9405,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789443271358125e-05,
+      "loss": 3.9459,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6781057323847605e-05,
+      "loss": 3.9561,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772671376337085e-05,
+      "loss": 3.9559,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764285428826565e-05,
+      "loss": 3.9581,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755915860119774e-05,
+      "loss": 3.9426,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747529912609254e-05,
+      "loss": 3.9461,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739143965098734e-05,
+      "loss": 3.9371,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730758017588214e-05,
+      "loss": 3.943,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672238844888143e-05,
+      "loss": 3.9485,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671400250137091e-05,
+      "loss": 3.9487,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670561655386039e-05,
+      "loss": 3.9659,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669723060634987e-05,
+      "loss": 3.9461,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668884465883935e-05,
+      "loss": 3.9381,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668045871132883e-05,
+      "loss": 3.9426,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667207276381831e-05,
+      "loss": 3.9435,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666370319511152e-05,
+      "loss": 3.9499,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655317247601e-05,
+      "loss": 3.9459,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664693130009048e-05,
+      "loss": 3.9419,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663854535257996e-05,
+      "loss": 3.9377,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663015940506944e-05,
+      "loss": 3.9463,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662177345755892e-05,
+      "loss": 3.9211,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661340388885213e-05,
+      "loss": 3.9387,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6605017941341614e-05,
+      "loss": 3.9345,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596631993831094e-05,
+      "loss": 3.9378,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588246046320574e-05,
+      "loss": 3.9303,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657987647761378e-05,
+      "loss": 3.9396,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657149053010326e-05,
+      "loss": 3.925,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656310458259274e-05,
+      "loss": 3.931,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655471863508222e-05,
+      "loss": 3.9383,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65463326875717e-05,
+      "loss": 3.9313,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653796311886491e-05,
+      "loss": 3.9512,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652957717135439e-05,
+      "loss": 3.9432,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652119122384387e-05,
+      "loss": 3.9377,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651280527633335e-05,
+      "loss": 3.9264,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650443570762657e-05,
+      "loss": 3.9306,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649604976011605e-05,
+      "loss": 3.9343,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648766381260553e-05,
+      "loss": 3.931,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479277865095e-05,
+      "loss": 3.9295,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647090829638822e-05,
+      "loss": 3.9216,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64625223488777e-05,
+      "loss": 3.9317,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645413640136718e-05,
+      "loss": 3.9188,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644575045385665e-05,
+      "loss": 3.9276,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643736450634613e-05,
+      "loss": 3.9241,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6428994937639346e-05,
+      "loss": 3.9254,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642060899012882e-05,
+      "loss": 3.9145,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412223042618306e-05,
+      "loss": 3.9399,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403837095107786e-05,
+      "loss": 3.9408,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395467526401e-05,
+      "loss": 3.9287,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387081578890475e-05,
+      "loss": 3.9144,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378695631379955e-05,
+      "loss": 3.9268,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370309683869435e-05,
+      "loss": 3.9281,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361923736358915e-05,
+      "loss": 3.9319,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353554167652124e-05,
+      "loss": 3.9251,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345168220141604e-05,
+      "loss": 3.9242,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336782272631084e-05,
+      "loss": 3.9244,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328396325120564e-05,
+      "loss": 3.9314,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6320010377610043e-05,
+      "loss": 3.9206,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631164080890326e-05,
+      "loss": 3.9259,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630325486139274e-05,
+      "loss": 3.9222,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629486891388222e-05,
+      "loss": 3.914,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62864829663717e-05,
+      "loss": 3.9269,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627811339766491e-05,
+      "loss": 3.9242,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626972745015439e-05,
+      "loss": 3.9223,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626134150264387e-05,
+      "loss": 3.9137,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625295555513335e-05,
+      "loss": 3.9128,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.951106309890747,
+      "eval_runtime": 307.338,
+      "eval_samples_per_second": 1241.6,
+      "eval_steps_per_second": 38.801,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624456960762283e-05,
+      "loss": 3.9148,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623620003891604e-05,
+      "loss": 3.9056,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622781409140552e-05,
+      "loss": 3.9222,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219428143895e-05,
+      "loss": 3.9141,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621104219638448e-05,
+      "loss": 3.9257,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620267262767769e-05,
+      "loss": 3.9068,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619428668016717e-05,
+      "loss": 3.9103,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618590073265665e-05,
+      "loss": 3.8988,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617751478514613e-05,
+      "loss": 3.9153,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616912883763561e-05,
+      "loss": 3.9135,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616075926892882e-05,
+      "loss": 3.9169,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61523733214183e-05,
+      "loss": 3.9094,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614398737390778e-05,
+      "loss": 3.9027,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613560142639726e-05,
+      "loss": 3.9024,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612721547888674e-05,
+      "loss": 3.8978,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611884591017995e-05,
+      "loss": 3.9008,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611045996266943e-05,
+      "loss": 3.9021,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610207401515891e-05,
+      "loss": 3.8975,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609368806764839e-05,
+      "loss": 3.8975,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608530212013788e-05,
+      "loss": 3.9214,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076932551431087e-05,
+      "loss": 3.9018,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068546603920566e-05,
+      "loss": 3.9045,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6060160656410046e-05,
+      "loss": 3.8985,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051774708899526e-05,
+      "loss": 3.9081,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043405140192735e-05,
+      "loss": 3.8913,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6035019192682215e-05,
+      "loss": 3.906,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026633245171695e-05,
+      "loss": 3.897,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018247297661175e-05,
+      "loss": 3.8936,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009877728954384e-05,
+      "loss": 3.8808,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001491781443864e-05,
+      "loss": 3.8913,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5993105833933344e-05,
+      "loss": 3.8972,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598471988642283e-05,
+      "loss": 3.9008,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597635031771604e-05,
+      "loss": 3.9015,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596796437020552e-05,
+      "loss": 3.8986,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959578422695e-05,
+      "loss": 3.8896,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595119247518448e-05,
+      "loss": 3.9044,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594282290647769e-05,
+      "loss": 3.8846,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593443695896717e-05,
+      "loss": 3.8915,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592605101145665e-05,
+      "loss": 3.8767,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591766506394613e-05,
+      "loss": 3.8927,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590929549523934e-05,
+      "loss": 3.8842,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590090954772882e-05,
+      "loss": 3.8934,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58925236002183e-05,
+      "loss": 3.8895,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5884137652707785e-05,
+      "loss": 3.8803,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587575170519726e-05,
+      "loss": 3.8951,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867382136490474e-05,
+      "loss": 3.8847,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858996188979954e-05,
+      "loss": 3.8823,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585061024146943e-05,
+      "loss": 3.8693,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584222429395891e-05,
+      "loss": 3.8783,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583383834644839e-05,
+      "loss": 3.8902,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58254687777416e-05,
+      "loss": 3.8851,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5817082830231076e-05,
+      "loss": 3.8829,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808696882720556e-05,
+      "loss": 3.871,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800310935210036e-05,
+      "loss": 3.8829,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579194136650325e-05,
+      "loss": 3.8675,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578355541899273e-05,
+      "loss": 3.8869,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577516947148221e-05,
+      "loss": 3.872,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576678352397169e-05,
+      "loss": 3.8795,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57584139552649e-05,
+      "loss": 3.8841,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575002800775438e-05,
+      "loss": 3.8673,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574164206024386e-05,
+      "loss": 3.8718,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573325611273334e-05,
+      "loss": 3.8785,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572487016522282e-05,
+      "loss": 3.865,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571650059651603e-05,
+      "loss": 3.8738,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570811464900551e-05,
+      "loss": 3.8762,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569972870149499e-05,
+      "loss": 3.8798,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691342753984476e-05,
+      "loss": 3.8647,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5682973185277685e-05,
+      "loss": 3.8704,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674587237767165e-05,
+      "loss": 3.861,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5666201290256645e-05,
+      "loss": 3.8716,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657815342746125e-05,
+      "loss": 3.8828,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649445774039334e-05,
+      "loss": 3.8733,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5641059826528814e-05,
+      "loss": 3.8848,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632673879018294e-05,
+      "loss": 3.8836,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624287931507774e-05,
+      "loss": 3.8812,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561591836280098e-05,
+      "loss": 3.8599,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560753241529046e-05,
+      "loss": 3.8781,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559914646777994e-05,
+      "loss": 3.8771,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559076052026943e-05,
+      "loss": 3.867,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558239095156264e-05,
+      "loss": 3.872,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557400500405212e-05,
+      "loss": 3.8748,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55656190565416e-05,
+      "loss": 3.8861,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555723310903108e-05,
+      "loss": 3.8684,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554886354032429e-05,
+      "loss": 3.8562,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554047759281377e-05,
+      "loss": 3.8665,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553209164530325e-05,
+      "loss": 3.8689,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552370569779273e-05,
+      "loss": 3.8735,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551533612908594e-05,
+      "loss": 3.8738,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550695018157542e-05,
+      "loss": 3.8629,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54985642340649e-05,
+      "loss": 3.8656,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5490178286554384e-05,
+      "loss": 3.8555,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548180871784759e-05,
+      "loss": 3.8606,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547342277033707e-05,
+      "loss": 3.8611,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546503682282655e-05,
+      "loss": 3.8743,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545665087531603e-05,
+      "loss": 3.8857,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544828130660924e-05,
+      "loss": 3.862,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543989535909872e-05,
+      "loss": 3.8545,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54315094115882e-05,
+      "loss": 3.8642,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542312346407768e-05,
+      "loss": 3.8616,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541475389537089e-05,
+      "loss": 3.8731,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540636794786037e-05,
+      "loss": 3.8625,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539798200034985e-05,
+      "loss": 3.8651,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538959605283933e-05,
+      "loss": 3.8589,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5381226484132546e-05,
+      "loss": 3.8696,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372840536622026e-05,
+      "loss": 3.8422,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364454589111506e-05,
+      "loss": 3.8566,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5356068641600986e-05,
+      "loss": 3.8522,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347699072894195e-05,
+      "loss": 3.8643,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339313125383675e-05,
+      "loss": 3.8503,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5330927177873155e-05,
+      "loss": 3.8594,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5322541230362635e-05,
+      "loss": 3.8493,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314171661655844e-05,
+      "loss": 3.8534,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305785714145324e-05,
+      "loss": 3.8572,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297399766634804e-05,
+      "loss": 3.8583,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5289013819124284e-05,
+      "loss": 3.8695,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52806442504175e-05,
+      "loss": 3.8698,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527227468171071e-05,
+      "loss": 3.8594,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526388873420019e-05,
+      "loss": 3.8515,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525550278668967e-05,
+      "loss": 3.8528,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524711683917915e-05,
+      "loss": 3.8587,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523873089166863e-05,
+      "loss": 3.8499,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523034494415811e-05,
+      "loss": 3.8544,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522195899664759e-05,
+      "loss": 3.8528,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521357304913707e-05,
+      "loss": 3.8521,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520520348043028e-05,
+      "loss": 3.8478,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519681753291976e-05,
+      "loss": 3.8487,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518843158540924e-05,
+      "loss": 3.8479,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5180045637898725e-05,
+      "loss": 3.8523,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5171676069191934e-05,
+      "loss": 3.8383,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5163290121681414e-05,
+      "loss": 3.8654,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154904174170894e-05,
+      "loss": 3.8632,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146518226660373e-05,
+      "loss": 3.8585,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513814865795358e-05,
+      "loss": 3.8364,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512976271044306e-05,
+      "loss": 3.8536,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512137676293254e-05,
+      "loss": 3.8575,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511299081542202e-05,
+      "loss": 3.8537,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510462124671523e-05,
+      "loss": 3.8519,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509623529920471e-05,
+      "loss": 3.8519,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508784935169419e-05,
+      "loss": 3.8501,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507946340418368e-05,
+      "loss": 3.8574,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507109383547689e-05,
+      "loss": 3.8471,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506270788796637e-05,
+      "loss": 3.8504,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505432194045585e-05,
+      "loss": 3.8501,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504593599294533e-05,
+      "loss": 3.8388,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5037566424238536e-05,
+      "loss": 3.8563,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5029180476728016e-05,
+      "loss": 3.8523,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020794529217496e-05,
+      "loss": 3.85,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012408581706976e-05,
+      "loss": 3.8396,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500402263419645e-05,
+      "loss": 3.8414,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.909769058227539,
+      "eval_runtime": 307.4618,
+      "eval_samples_per_second": 1241.101,
+      "eval_steps_per_second": 38.785,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499563668668593e-05,
+      "loss": 3.8395,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4987250739175416e-05,
+      "loss": 3.8349,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978864791664896e-05,
+      "loss": 3.8475,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970478844154376e-05,
+      "loss": 3.8428,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4962092896643856e-05,
+      "loss": 3.859,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953706949133336e-05,
+      "loss": 3.8336,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945321001622816e-05,
+      "loss": 3.8396,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936935054112296e-05,
+      "loss": 3.8267,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928565485405505e-05,
+      "loss": 3.8468,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4920179537894985e-05,
+      "loss": 3.8432,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911793590384465e-05,
+      "loss": 3.8415,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903407642873945e-05,
+      "loss": 3.8414,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4895038074167154e-05,
+      "loss": 3.8298,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886652126656634e-05,
+      "loss": 3.8381,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878266179146114e-05,
+      "loss": 3.8258,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48698802316356e-05,
+      "loss": 3.8339,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486151066292881e-05,
+      "loss": 3.83,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485312471541829e-05,
+      "loss": 3.8305,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484473876790777e-05,
+      "loss": 3.8343,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483635282039725e-05,
+      "loss": 3.851,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482798325169046e-05,
+      "loss": 3.8323,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481959730417994e-05,
+      "loss": 3.838,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481121135666942e-05,
+      "loss": 3.8305,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48028254091589e-05,
+      "loss": 3.8404,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479445584045211e-05,
+      "loss": 3.8263,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478606989294159e-05,
+      "loss": 3.8371,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477768394543107e-05,
+      "loss": 3.8295,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476931437672428e-05,
+      "loss": 3.8248,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476092842921376e-05,
+      "loss": 3.8184,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475254248170324e-05,
+      "loss": 3.8242,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474415653419272e-05,
+      "loss": 3.8305,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473578696548593e-05,
+      "loss": 3.8347,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472740101797541e-05,
+      "loss": 3.8355,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471901507046489e-05,
+      "loss": 3.8294,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471062912295437e-05,
+      "loss": 3.8239,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470225955424758e-05,
+      "loss": 3.8382,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469387360673706e-05,
+      "loss": 3.825,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468548765922654e-05,
+      "loss": 3.8249,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467710171171602e-05,
+      "loss": 3.8101,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466873214300924e-05,
+      "loss": 3.8257,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466034619549872e-05,
+      "loss": 3.8205,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46519602479882e-05,
+      "loss": 3.831,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464357430047768e-05,
+      "loss": 3.8252,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4635204731770886e-05,
+      "loss": 3.8145,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626818784260366e-05,
+      "loss": 3.8298,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618432836749846e-05,
+      "loss": 3.8195,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4610046889239326e-05,
+      "loss": 3.8168,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4601677320532535e-05,
+      "loss": 3.8102,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4593291373022015e-05,
+      "loss": 3.81,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584905425511495e-05,
+      "loss": 3.825,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576519478000975e-05,
+      "loss": 3.8188,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456814990929419e-05,
+      "loss": 3.818,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455976396178367e-05,
+      "loss": 3.8096,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455137801427315e-05,
+      "loss": 3.8194,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454299206676263e-05,
+      "loss": 3.8044,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453462249805584e-05,
+      "loss": 3.822,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452623655054532e-05,
+      "loss": 3.8055,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45178506030348e-05,
+      "loss": 3.8193,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450946465552428e-05,
+      "loss": 3.8177,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450109508681749e-05,
+      "loss": 3.8108,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449270913930697e-05,
+      "loss": 3.8061,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448432319179645e-05,
+      "loss": 3.817,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447593724428593e-05,
+      "loss": 3.8024,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467567675579144e-05,
+      "loss": 3.8109,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4459181728068624e-05,
+      "loss": 3.8181,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4450795780558104e-05,
+      "loss": 3.8145,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442409833047584e-05,
+      "loss": 3.8041,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443404026434079e-05,
+      "loss": 3.8068,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442565431683027e-05,
+      "loss": 3.8032,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441726836931975e-05,
+      "loss": 3.8069,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440888242180923e-05,
+      "loss": 3.8191,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440051285310244e-05,
+      "loss": 3.8137,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439212690559192e-05,
+      "loss": 3.8248,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43837409580814e-05,
+      "loss": 3.8228,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437535501057088e-05,
+      "loss": 3.821,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43669854418641e-05,
+      "loss": 3.8013,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435859949435358e-05,
+      "loss": 3.8114,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435021354684306e-05,
+      "loss": 3.8204,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434184397813627e-05,
+      "loss": 3.8072,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433345803062575e-05,
+      "loss": 3.8088,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432507208311523e-05,
+      "loss": 3.8115,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431668613560471e-05,
+      "loss": 3.8268,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430830018809418e-05,
+      "loss": 3.809,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429991424058366e-05,
+      "loss": 3.7949,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429152829307314e-05,
+      "loss": 3.8046,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428314234556262e-05,
+      "loss": 3.8092,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274772776855836e-05,
+      "loss": 3.8144,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266386829345316e-05,
+      "loss": 3.8139,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4258000881834796e-05,
+      "loss": 3.8014,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249614934324276e-05,
+      "loss": 3.8053,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4241245365617485e-05,
+      "loss": 3.8018,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232859418106965e-05,
+      "loss": 3.7986,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224473470596445e-05,
+      "loss": 3.805,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216087523085925e-05,
+      "loss": 3.8118,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207717954379134e-05,
+      "loss": 3.8278,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4199332006868614e-05,
+      "loss": 3.8089,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190946059358094e-05,
+      "loss": 3.7936,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182560111847573e-05,
+      "loss": 3.8048,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417419054314079e-05,
+      "loss": 3.8005,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416580459563027e-05,
+      "loss": 3.8146,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415741864811975e-05,
+      "loss": 3.807,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414903270060923e-05,
+      "loss": 3.8059,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414066313190244e-05,
+      "loss": 3.7972,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413227718439192e-05,
+      "loss": 3.8102,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41238912368814e-05,
+      "loss": 3.7856,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411550528937088e-05,
+      "loss": 3.8016,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410713572066409e-05,
+      "loss": 3.79,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409874977315357e-05,
+      "loss": 3.81,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409036382564305e-05,
+      "loss": 3.7944,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408197787813253e-05,
+      "loss": 3.7985,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407360830942574e-05,
+      "loss": 3.7936,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406522236191522e-05,
+      "loss": 3.8004,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40568364144047e-05,
+      "loss": 3.8017,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404845046689418e-05,
+      "loss": 3.7993,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404008089818739e-05,
+      "loss": 3.8116,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40317113294806e-05,
+      "loss": 3.8123,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402332538197008e-05,
+      "loss": 3.8021,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401493943445956e-05,
+      "loss": 3.7955,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400655348694904e-05,
+      "loss": 3.7963,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399818391824225e-05,
+      "loss": 3.8061,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398979797073173e-05,
+      "loss": 3.7874,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398141202322122e-05,
+      "loss": 3.8025,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39730260757107e-05,
+      "loss": 3.7911,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396464012820018e-05,
+      "loss": 3.7969,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3956254180689657e-05,
+      "loss": 3.7942,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947868233179137e-05,
+      "loss": 3.7944,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3939498664472346e-05,
+      "loss": 3.7882,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3931112716961826e-05,
+      "loss": 3.8007,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922726769451306e-05,
+      "loss": 3.786,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3914340821940786e-05,
+      "loss": 3.8077,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905971253233995e-05,
+      "loss": 3.8018,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897585305723475e-05,
+      "loss": 3.8092,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3889199358212955e-05,
+      "loss": 3.779,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880813410702434e-05,
+      "loss": 3.8012,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387244384199565e-05,
+      "loss": 3.8003,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386405789448513e-05,
+      "loss": 3.7979,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385567194697461e-05,
+      "loss": 3.7961,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384728599946409e-05,
+      "loss": 3.7993,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38389164307573e-05,
+      "loss": 3.7937,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383053048324678e-05,
+      "loss": 3.8051,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382214453573626e-05,
+      "loss": 3.7871,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381377496702947e-05,
+      "loss": 3.7971,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380538901951895e-05,
+      "loss": 3.7951,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379700307200843e-05,
+      "loss": 3.786,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378861712449791e-05,
+      "loss": 3.7979,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378023117698739e-05,
+      "loss": 3.8018,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771861608280604e-05,
+      "loss": 3.7938,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763475660770084e-05,
+      "loss": 3.7841,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3755089713259564e-05,
+      "loss": 3.7915,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.884277105331421,
+      "eval_runtime": 308.3892,
+      "eval_samples_per_second": 1237.368,
+      "eval_steps_per_second": 38.669,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3746703765749044e-05,
+      "loss": 3.7934,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3738317818238524e-05,
+      "loss": 3.7809,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3729931870728004e-05,
+      "loss": 3.7909,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3721545923217484e-05,
+      "loss": 3.7934,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371315997570696e-05,
+      "loss": 3.8025,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370479040700017e-05,
+      "loss": 3.7871,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369640445948965e-05,
+      "loss": 3.784,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368801851197913e-05,
+      "loss": 3.7751,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3679632564468606e-05,
+      "loss": 3.7896,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367126299576182e-05,
+      "loss": 3.7927,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36628770482513e-05,
+      "loss": 3.7875,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365449110074079e-05,
+      "loss": 3.7893,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364610515323026e-05,
+      "loss": 3.7781,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363773558452348e-05,
+      "loss": 3.7866,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362934963701296e-05,
+      "loss": 3.7753,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362096368950243e-05,
+      "loss": 3.7822,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361257774199191e-05,
+      "loss": 3.7743,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3604208173285126e-05,
+      "loss": 3.7827,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3595822225774606e-05,
+      "loss": 3.7776,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358743627826408e-05,
+      "loss": 3.7967,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357905033075356e-05,
+      "loss": 3.7826,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357069714085051e-05,
+      "loss": 3.791,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356231119333999e-05,
+      "loss": 3.7763,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355392524582947e-05,
+      "loss": 3.788,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354553929831895e-05,
+      "loss": 3.773,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353716972961216e-05,
+      "loss": 3.7869,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352878378210164e-05,
+      "loss": 3.7815,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352039783459112e-05,
+      "loss": 3.7691,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35120118870806e-05,
+      "loss": 3.7652,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350362593957008e-05,
+      "loss": 3.7726,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349523999205955e-05,
+      "loss": 3.7805,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348685404454903e-05,
+      "loss": 3.7823,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347846809703851e-05,
+      "loss": 3.7852,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347009852833173e-05,
+      "loss": 3.7844,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346171258082121e-05,
+      "loss": 3.7743,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345332663331069e-05,
+      "loss": 3.7848,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3444957064603905e-05,
+      "loss": 3.776,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343657111709338e-05,
+      "loss": 3.7729,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342818516958286e-05,
+      "loss": 3.7594,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341979922207234e-05,
+      "loss": 3.7764,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3411429653365554e-05,
+      "loss": 3.7706,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340304370585503e-05,
+      "loss": 3.7793,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339465775834451e-05,
+      "loss": 3.773,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338627181083399e-05,
+      "loss": 3.7668,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337788586332347e-05,
+      "loss": 3.7804,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336951629461668e-05,
+      "loss": 3.7681,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336113034710616e-05,
+      "loss": 3.7695,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335274439959564e-05,
+      "loss": 3.7622,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334435845208512e-05,
+      "loss": 3.7597,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33359725045746e-05,
+      "loss": 3.7738,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332758655706408e-05,
+      "loss": 3.7713,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331921698835729e-05,
+      "loss": 3.7685,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331083104084677e-05,
+      "loss": 3.7617,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330244509333625e-05,
+      "loss": 3.7713,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329405914582573e-05,
+      "loss": 3.7514,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328568957711894e-05,
+      "loss": 3.7745,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327730362960842e-05,
+      "loss": 3.759,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32689176820979e-05,
+      "loss": 3.7694,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326053173458739e-05,
+      "loss": 3.7679,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325214578707687e-05,
+      "loss": 3.7645,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324375983956635e-05,
+      "loss": 3.7584,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323537389205583e-05,
+      "loss": 3.7672,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322698794454531e-05,
+      "loss": 3.7599,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3218618375838516e-05,
+      "loss": 3.7587,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3210232428327996e-05,
+      "loss": 3.7699,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201846480817476e-05,
+      "loss": 3.7657,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193460533306956e-05,
+      "loss": 3.7535,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3185074585796436e-05,
+      "loss": 3.7646,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176688638285916e-05,
+      "loss": 3.7524,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316830269077539e-05,
+      "loss": 3.758,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315991674326487e-05,
+      "loss": 3.7734,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151547174558085e-05,
+      "loss": 3.7688,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31431776058513e-05,
+      "loss": 3.7739,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313479165834078e-05,
+      "loss": 3.7706,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312640571083026e-05,
+      "loss": 3.7787,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311801976331974e-05,
+      "loss": 3.7576,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3109633815809214e-05,
+      "loss": 3.7618,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3101247868298694e-05,
+      "loss": 3.7733,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3092861920788174e-05,
+      "loss": 3.7581,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308449235208139e-05,
+      "loss": 3.7649,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307610640457086e-05,
+      "loss": 3.7684,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306772045706034e-05,
+      "loss": 3.7719,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305933450954982e-05,
+      "loss": 3.7622,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305096494084304e-05,
+      "loss": 3.7514,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304257899333252e-05,
+      "loss": 3.7581,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3034193045822e-05,
+      "loss": 3.7619,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302580709831148e-05,
+      "loss": 3.7681,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301742115080096e-05,
+      "loss": 3.7651,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300903520329044e-05,
+      "loss": 3.758,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300064925577992e-05,
+      "loss": 3.7617,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299227968707313e-05,
+      "loss": 3.7538,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.298389373956261e-05,
+      "loss": 3.7505,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975524170855817e-05,
+      "loss": 3.7559,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2967138223345296e-05,
+      "loss": 3.7663,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2958752275834776e-05,
+      "loss": 3.7809,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2950366328324256e-05,
+      "loss": 3.7657,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294198038081374e-05,
+      "loss": 3.7454,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293359443330322e-05,
+      "loss": 3.7622,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29252084857927e-05,
+      "loss": 3.7567,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291682253828218e-05,
+      "loss": 3.7672,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290845296957539e-05,
+      "loss": 3.7574,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290006702206487e-05,
+      "loss": 3.7618,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289169745335808e-05,
+      "loss": 3.7554,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288331150584756e-05,
+      "loss": 3.7588,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287492555833704e-05,
+      "loss": 3.7482,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286653961082652e-05,
+      "loss": 3.7545,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2858153663316e-05,
+      "loss": 3.7443,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284976771580548e-05,
+      "loss": 3.7676,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284138176829496e-05,
+      "loss": 3.7488,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283299582078444e-05,
+      "loss": 3.7523,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282462625207766e-05,
+      "loss": 3.7449,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2816256683370866e-05,
+      "loss": 3.7584,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807870735860346e-05,
+      "loss": 3.7557,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2799484788349826e-05,
+      "loss": 3.7536,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2791098840839306e-05,
+      "loss": 3.7661,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782712893328786e-05,
+      "loss": 3.7679,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2774343324621995e-05,
+      "loss": 3.7587,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2765957377111475e-05,
+      "loss": 3.7537,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757571429600955e-05,
+      "loss": 3.7474,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2749185482090435e-05,
+      "loss": 3.7611,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274081591338365e-05,
+      "loss": 3.7443,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273242996587313e-05,
+      "loss": 3.757,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272404401836261e-05,
+      "loss": 3.7444,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271565807085209e-05,
+      "loss": 3.7564,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270727212334157e-05,
+      "loss": 3.7434,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269888617583105e-05,
+      "loss": 3.7513,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269051660712426e-05,
+      "loss": 3.7496,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268213065961374e-05,
+      "loss": 3.7495,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267374471210322e-05,
+      "loss": 3.7454,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26653587645927e-05,
+      "loss": 3.7619,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265697281708217e-05,
+      "loss": 3.7604,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264858686957165e-05,
+      "loss": 3.7619,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264021730086487e-05,
+      "loss": 3.7351,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263183135335435e-05,
+      "loss": 3.7577,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262344540584383e-05,
+      "loss": 3.756,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261505945833331e-05,
+      "loss": 3.7578,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260667351082279e-05,
+      "loss": 3.7501,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259828756331227e-05,
+      "loss": 3.7565,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258991799460548e-05,
+      "loss": 3.7519,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258153204709496e-05,
+      "loss": 3.763,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257314609958444e-05,
+      "loss": 3.7418,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256476015207392e-05,
+      "loss": 3.7574,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2556390583367126e-05,
+      "loss": 3.7487,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2548004635856606e-05,
+      "loss": 3.7465,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2539618688346086e-05,
+      "loss": 3.7539,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253123274083557e-05,
+      "loss": 3.7533,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252286317212878e-05,
+      "loss": 3.7571,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251447722461826e-05,
+      "loss": 3.7365,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250609127710774e-05,
+      "loss": 3.7489,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.868356943130493,
+      "eval_runtime": 304.7992,
+      "eval_samples_per_second": 1251.942,
+      "eval_steps_per_second": 39.124,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249770532959722e-05,
+      "loss": 3.7443,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24893193820867e-05,
+      "loss": 3.7377,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248093343457618e-05,
+      "loss": 3.7491,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247254748706566e-05,
+      "loss": 3.7522,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246417791835887e-05,
+      "loss": 3.7612,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245579197084835e-05,
+      "loss": 3.7411,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244740602333783e-05,
+      "loss": 3.7368,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243902007582731e-05,
+      "loss": 3.7381,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243065050712052e-05,
+      "loss": 3.741,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2422264559610006e-05,
+      "loss": 3.7515,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2413878612099486e-05,
+      "loss": 3.7455,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2405492664588966e-05,
+      "loss": 3.7472,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2397106717078446e-05,
+      "loss": 3.7398,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2388737148371655e-05,
+      "loss": 3.7397,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2380351200861135e-05,
+      "loss": 3.7363,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2371965253350615e-05,
+      "loss": 3.7401,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363579305840095e-05,
+      "loss": 3.7327,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2355209737133304e-05,
+      "loss": 3.7374,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346823789622784e-05,
+      "loss": 3.7382,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2338437842112264e-05,
+      "loss": 3.7542,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2330051894601744e-05,
+      "loss": 3.7419,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232168232589496e-05,
+      "loss": 3.7497,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231329637838444e-05,
+      "loss": 3.7341,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230491043087392e-05,
+      "loss": 3.7469,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22965244833634e-05,
+      "loss": 3.7317,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228815491465661e-05,
+      "loss": 3.747,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227976896714609e-05,
+      "loss": 3.7372,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227138301963557e-05,
+      "loss": 3.7277,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226299707212505e-05,
+      "loss": 3.729,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225462750341826e-05,
+      "loss": 3.7317,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224624155590774e-05,
+      "loss": 3.7391,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223785560839722e-05,
+      "loss": 3.7406,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22294696608867e-05,
+      "loss": 3.7437,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222108371337618e-05,
+      "loss": 3.7478,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2212714144669394e-05,
+      "loss": 3.7283,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2204328197158874e-05,
+      "loss": 3.747,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2195942249648353e-05,
+      "loss": 3.7392,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2187556302137833e-05,
+      "loss": 3.7279,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217917035462731e-05,
+      "loss": 3.7219,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217080078592052e-05,
+      "loss": 3.7311,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216241483841e-05,
+      "loss": 3.7344,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2154028890899476e-05,
+      "loss": 3.7386,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2145642943388956e-05,
+      "loss": 3.7319,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213727337468217e-05,
+      "loss": 3.7265,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212888742717165e-05,
+      "loss": 3.744,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212050147966113e-05,
+      "loss": 3.7306,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211211553215061e-05,
+      "loss": 3.7287,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210374596344383e-05,
+      "loss": 3.7208,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209536001593331e-05,
+      "loss": 3.7186,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208697406842278e-05,
+      "loss": 3.7358,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207858812091226e-05,
+      "loss": 3.7333,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2070218552205476e-05,
+      "loss": 3.7276,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206183260469495e-05,
+      "loss": 3.7202,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205344665718443e-05,
+      "loss": 3.7339,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204506070967391e-05,
+      "loss": 3.7096,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036691140967125e-05,
+      "loss": 3.7363,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2028305193456605e-05,
+      "loss": 3.7192,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2019919245946085e-05,
+      "loss": 3.7232,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011533298435565e-05,
+      "loss": 3.7328,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200316372972878e-05,
+      "loss": 3.7229,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1994777782218254e-05,
+      "loss": 3.719,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1986391834707734e-05,
+      "loss": 3.7336,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1978005887197214e-05,
+      "loss": 3.7181,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1969619939686694e-05,
+      "loss": 3.7212,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19612503709799e-05,
+      "loss": 3.7293,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195286442346938e-05,
+      "loss": 3.7291,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194447847595886e-05,
+      "loss": 3.7157,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193609252844834e-05,
+      "loss": 3.7228,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192772295974156e-05,
+      "loss": 3.7117,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191933701223104e-05,
+      "loss": 3.7176,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191095106472052e-05,
+      "loss": 3.7313,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190256511721e-05,
+      "loss": 3.7277,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189419554850321e-05,
+      "loss": 3.7361,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188580960099269e-05,
+      "loss": 3.7285,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187742365348217e-05,
+      "loss": 3.7432,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186903770597165e-05,
+      "loss": 3.7167,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186065175846113e-05,
+      "loss": 3.7278,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185228218975434e-05,
+      "loss": 3.7331,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843896242243817e-05,
+      "loss": 3.7174,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1835510294733297e-05,
+      "loss": 3.7266,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1827124347222776e-05,
+      "loss": 3.7289,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181875477851599e-05,
+      "loss": 3.7349,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181036883100547e-05,
+      "loss": 3.7235,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180198288349495e-05,
+      "loss": 3.7149,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179359693598443e-05,
+      "loss": 3.7157,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178522736727764e-05,
+      "loss": 3.723,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177684141976712e-05,
+      "loss": 3.7305,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17684554722566e-05,
+      "loss": 3.7267,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176006952474608e-05,
+      "loss": 3.7247,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175168357723556e-05,
+      "loss": 3.715,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174329762972504e-05,
+      "loss": 3.7188,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173492806101825e-05,
+      "loss": 3.7157,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172654211350773e-05,
+      "loss": 3.7197,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171815616599721e-05,
+      "loss": 3.7306,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170977021848669e-05,
+      "loss": 3.74,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1701400649779906e-05,
+      "loss": 3.7291,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1693014702269386e-05,
+      "loss": 3.7119,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1684628754758866e-05,
+      "loss": 3.7211,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1676242807248346e-05,
+      "loss": 3.7188,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1667856859737826e-05,
+      "loss": 3.7281,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1659487291031035e-05,
+      "loss": 3.7193,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1651101343520515e-05,
+      "loss": 3.7272,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642715396009995e-05,
+      "loss": 3.7195,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1634329448499475e-05,
+      "loss": 3.7169,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1625959879792684e-05,
+      "loss": 3.7137,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1617573932282164e-05,
+      "loss": 3.7162,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1609187984771644e-05,
+      "loss": 3.7074,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160080203726113e-05,
+      "loss": 3.7286,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159241608975061e-05,
+      "loss": 3.7132,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158404652104382e-05,
+      "loss": 3.7137,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15756605735333e-05,
+      "loss": 3.7119,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156727462602278e-05,
+      "loss": 3.7167,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155888867851226e-05,
+      "loss": 3.7228,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155051910980547e-05,
+      "loss": 3.7158,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154213316229495e-05,
+      "loss": 3.7316,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153374721478443e-05,
+      "loss": 3.7272,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152536126727391e-05,
+      "loss": 3.7243,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151699169856712e-05,
+      "loss": 3.7198,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15086057510566e-05,
+      "loss": 3.7067,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1500219803546084e-05,
+      "loss": 3.7294,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149185023483929e-05,
+      "loss": 3.714,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148346428732877e-05,
+      "loss": 3.7146,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147507833981825e-05,
+      "loss": 3.7084,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146669239230773e-05,
+      "loss": 3.718,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145832282360094e-05,
+      "loss": 3.7088,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144993687609042e-05,
+      "loss": 3.7145,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14415509285799e-05,
+      "loss": 3.7114,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143316498106938e-05,
+      "loss": 3.7171,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424779033558855e-05,
+      "loss": 3.7075,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141640946485207e-05,
+      "loss": 3.7262,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140802351734155e-05,
+      "loss": 3.7208,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139963756983103e-05,
+      "loss": 3.7265,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139125162232051e-05,
+      "loss": 3.702,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138288205361373e-05,
+      "loss": 3.7221,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137449610610321e-05,
+      "loss": 3.7185,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136611015859268e-05,
+      "loss": 3.7195,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135772421108216e-05,
+      "loss": 3.7166,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134933826357164e-05,
+      "loss": 3.7209,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340968694864856e-05,
+      "loss": 3.7105,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133258274735433e-05,
+      "loss": 3.7326,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132419679984381e-05,
+      "loss": 3.7015,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131581085233329e-05,
+      "loss": 3.7219,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1307441283626505e-05,
+      "loss": 3.7139,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1299055336115985e-05,
+      "loss": 3.712,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1290669388605465e-05,
+      "loss": 3.7169,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1282283441094945e-05,
+      "loss": 3.7153,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1273913872388154e-05,
+      "loss": 3.7221,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1265527924877634e-05,
+      "loss": 3.6987,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1257141977367114e-05,
+      "loss": 3.7172,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.857776165008545,
+      "eval_runtime": 331.1709,
+      "eval_samples_per_second": 1152.248,
+      "eval_steps_per_second": 36.009,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1248756029856594e-05,
+      "loss": 3.7143,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1240370082346074e-05,
+      "loss": 3.7012,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123200051363928e-05,
+      "loss": 3.7128,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122361456612876e-05,
+      "loss": 3.7133,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121522861861824e-05,
+      "loss": 3.7241,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120684267110773e-05,
+      "loss": 3.7115,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119847310240094e-05,
+      "loss": 3.7025,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119008715489042e-05,
+      "loss": 3.7042,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11817012073799e-05,
+      "loss": 3.7093,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117331525986938e-05,
+      "loss": 3.7148,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116494569116259e-05,
+      "loss": 3.7108,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115655974365207e-05,
+      "loss": 3.7122,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114817379614155e-05,
+      "loss": 3.7064,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113978784863103e-05,
+      "loss": 3.7061,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113140190112051e-05,
+      "loss": 3.6985,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1123032332413716e-05,
+      "loss": 3.7009,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1114646384903196e-05,
+      "loss": 3.7012,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110626043739268e-05,
+      "loss": 3.702,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109787448988216e-05,
+      "loss": 3.7052,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108948854237164e-05,
+      "loss": 3.7148,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108111897366485e-05,
+      "loss": 3.7095,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107273302615433e-05,
+      "loss": 3.7124,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106434707864381e-05,
+      "loss": 3.703,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105596113113329e-05,
+      "loss": 3.71,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10475915624265e-05,
+      "loss": 3.6961,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103920561491598e-05,
+      "loss": 3.7168,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103081966740546e-05,
+      "loss": 3.6997,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102243371989494e-05,
+      "loss": 3.6961,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101406415118815e-05,
+      "loss": 3.6968,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005678203677637e-05,
+      "loss": 3.6995,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0997292256167117e-05,
+      "loss": 3.7002,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988906308656597e-05,
+      "loss": 3.7102,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0980536739949806e-05,
+      "loss": 3.7111,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0972150792439286e-05,
+      "loss": 3.7075,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0963764844928766e-05,
+      "loss": 3.6999,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0955378897418245e-05,
+      "loss": 3.7105,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0947009328711455e-05,
+      "loss": 3.709,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0938623381200935e-05,
+      "loss": 3.69,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0930237433690414e-05,
+      "loss": 3.6917,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0921851486179894e-05,
+      "loss": 3.6949,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0913481917473104e-05,
+      "loss": 3.6997,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0905095969962583e-05,
+      "loss": 3.703,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089671002245207e-05,
+      "loss": 3.699,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088832407494155e-05,
+      "loss": 3.695,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087995450623476e-05,
+      "loss": 3.7077,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087156855872424e-05,
+      "loss": 3.698,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086318261121372e-05,
+      "loss": 3.6991,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08547966637032e-05,
+      "loss": 3.6843,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084642709499641e-05,
+      "loss": 3.6822,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083804114748589e-05,
+      "loss": 3.7076,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082965519997537e-05,
+      "loss": 3.6956,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082126925246485e-05,
+      "loss": 3.6998,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081289968375806e-05,
+      "loss": 3.6915,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080451373624754e-05,
+      "loss": 3.6938,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0796127788737024e-05,
+      "loss": 3.6796,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0787741841226504e-05,
+      "loss": 3.6982,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077937227251971e-05,
+      "loss": 3.6894,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077098632500919e-05,
+      "loss": 3.6905,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076260037749867e-05,
+      "loss": 3.7025,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075421442998815e-05,
+      "loss": 3.6901,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074584486128136e-05,
+      "loss": 3.6836,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073745891377084e-05,
+      "loss": 3.7005,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072907296626032e-05,
+      "loss": 3.685,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07206870187498e-05,
+      "loss": 3.6846,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071231745004301e-05,
+      "loss": 3.7004,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070393150253249e-05,
+      "loss": 3.6943,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069554555502198e-05,
+      "loss": 3.68,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068715960751146e-05,
+      "loss": 3.6936,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0678790038804667e-05,
+      "loss": 3.6815,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0670404091294147e-05,
+      "loss": 3.6837,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0662018143783627e-05,
+      "loss": 3.7005,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0653632196273106e-05,
+      "loss": 3.6977,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0645262627566316e-05,
+      "loss": 3.6999,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636876680055796e-05,
+      "loss": 3.6961,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0628490732545275e-05,
+      "loss": 3.709,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0620104785034755e-05,
+      "loss": 3.6872,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611735216327965e-05,
+      "loss": 3.6955,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0603349268817444e-05,
+      "loss": 3.7014,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059496332130693e-05,
+      "loss": 3.6831,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0586577373796404e-05,
+      "loss": 3.6915,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0578191426285884e-05,
+      "loss": 3.7007,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05698218575791e-05,
+      "loss": 3.7008,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056143591006858e-05,
+      "loss": 3.6948,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055304996255805e-05,
+      "loss": 3.6829,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054466401504753e-05,
+      "loss": 3.6777,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053629444634075e-05,
+      "loss": 3.6917,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052790849883023e-05,
+      "loss": 3.7,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05195225513197e-05,
+      "loss": 3.6939,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051113660380918e-05,
+      "loss": 3.6922,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.05027670351024e-05,
+      "loss": 3.679,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049438108759188e-05,
+      "loss": 3.689,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048599514008136e-05,
+      "loss": 3.6833,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047760919257084e-05,
+      "loss": 3.6909,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0469239623864054e-05,
+      "loss": 3.6939,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046085367635353e-05,
+      "loss": 3.7074,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045246772884301e-05,
+      "loss": 3.6961,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044408178133249e-05,
+      "loss": 3.6816,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04357122126257e-05,
+      "loss": 3.688,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0427326265115176e-05,
+      "loss": 3.687,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0418940317604656e-05,
+      "loss": 3.6985,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0410554370094136e-05,
+      "loss": 3.6875,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040218480138735e-05,
+      "loss": 3.6991,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039379885387683e-05,
+      "loss": 3.689,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038541290636631e-05,
+      "loss": 3.6771,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037702695885579e-05,
+      "loss": 3.6884,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0368657390149e-05,
+      "loss": 3.6805,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036027144263848e-05,
+      "loss": 3.6802,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035188549512796e-05,
+      "loss": 3.6932,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034349954761744e-05,
+      "loss": 3.686,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033512997891065e-05,
+      "loss": 3.68,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032674403140013e-05,
+      "loss": 3.6837,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031835808388961e-05,
+      "loss": 3.6801,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030997213637909e-05,
+      "loss": 3.6914,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0301602567672305e-05,
+      "loss": 3.6829,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0293216620161785e-05,
+      "loss": 3.7019,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284830672651265e-05,
+      "loss": 3.6966,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0276444725140745e-05,
+      "loss": 3.6938,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0268075156433954e-05,
+      "loss": 3.6875,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0259689208923434e-05,
+      "loss": 3.6763,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0251303261412914e-05,
+      "loss": 3.6898,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0242917313902394e-05,
+      "loss": 3.6876,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0234547745195603e-05,
+      "loss": 3.6803,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022616179768508e-05,
+      "loss": 3.6769,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021777585017456e-05,
+      "loss": 3.6897,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020940628146778e-05,
+      "loss": 3.6774,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020102033395726e-05,
+      "loss": 3.6865,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019263438644674e-05,
+      "loss": 3.6773,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018424843893622e-05,
+      "loss": 3.6908,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01758624914257e-05,
+      "loss": 3.6745,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016749292271891e-05,
+      "loss": 3.6979,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015910697520839e-05,
+      "loss": 3.6866,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015072102769787e-05,
+      "loss": 3.6958,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014233508018735e-05,
+      "loss": 3.672,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013396551148056e-05,
+      "loss": 3.6924,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012557956397004e-05,
+      "loss": 3.6868,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011719361645952e-05,
+      "loss": 3.6926,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0108807668949e-05,
+      "loss": 3.6856,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010043810024221e-05,
+      "loss": 3.6855,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009205215273169e-05,
+      "loss": 3.6859,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008366620522117e-05,
+      "loss": 3.6989,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007528025771065e-05,
+      "loss": 3.6743,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006691068900386e-05,
+      "loss": 3.6887,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005852474149334e-05,
+      "loss": 3.6842,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005013879398282e-05,
+      "loss": 3.6799,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00417528464723e-05,
+      "loss": 3.6863,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003338327776551e-05,
+      "loss": 3.6821,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002499733025499e-05,
+      "loss": 3.6945,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001661138274447e-05,
+      "loss": 3.6675,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000822543523395e-05,
+      "loss": 3.6886,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.850329637527466,
+      "eval_runtime": 330.7284,
+      "eval_samples_per_second": 1153.79,
+      "eval_steps_per_second": 36.057,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9999855866527166e-05,
+      "loss": 3.6789,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9991469919016646e-05,
+      "loss": 3.6738,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9983083971506126e-05,
+      "loss": 3.68,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9974698023995606e-05,
+      "loss": 3.6814,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9966328455288815e-05,
+      "loss": 3.6972,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9957942507778295e-05,
+      "loss": 3.6773,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9949556560267775e-05,
+      "loss": 3.6776,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9941170612757255e-05,
+      "loss": 3.6751,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9932801044050464e-05,
+      "loss": 3.6751,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9924415096539944e-05,
+      "loss": 3.6874,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9916029149029424e-05,
+      "loss": 3.678,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9907643201518904e-05,
+      "loss": 3.6844,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989927363281212e-05,
+      "loss": 3.6743,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98908876853016e-05,
+      "loss": 3.6741,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988250173779108e-05,
+      "loss": 3.6724,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987411579028056e-05,
+      "loss": 3.6718,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986574622157377e-05,
+      "loss": 3.6693,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985736027406325e-05,
+      "loss": 3.6736,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984897432655273e-05,
+      "loss": 3.6758,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984058837904221e-05,
+      "loss": 3.6831,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983221881033542e-05,
+      "loss": 3.6815,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98238328628249e-05,
+      "loss": 3.6862,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981544691531438e-05,
+      "loss": 3.6697,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980706096780386e-05,
+      "loss": 3.6788,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979867502029334e-05,
+      "loss": 3.6705,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9790305451586554e-05,
+      "loss": 3.6843,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781919504076034e-05,
+      "loss": 3.673,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9773533556565514e-05,
+      "loss": 3.6702,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9765147609054994e-05,
+      "loss": 3.6657,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97567780403482e-05,
+      "loss": 3.6674,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974839209283768e-05,
+      "loss": 3.6667,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974000614532716e-05,
+      "loss": 3.6823,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973162019781664e-05,
+      "loss": 3.678,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972325062910985e-05,
+      "loss": 3.6798,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971486468159933e-05,
+      "loss": 3.6701,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970647873408881e-05,
+      "loss": 3.6817,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969809278657829e-05,
+      "loss": 3.68,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968972321787151e-05,
+      "loss": 3.6628,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968133727036099e-05,
+      "loss": 3.6644,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967295132285047e-05,
+      "loss": 3.6659,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966456537533995e-05,
+      "loss": 3.6682,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9656195806633156e-05,
+      "loss": 3.6731,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9647809859122636e-05,
+      "loss": 3.6701,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9639423911612116e-05,
+      "loss": 3.6655,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9631037964101596e-05,
+      "loss": 3.6785,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9622668395394805e-05,
+      "loss": 3.6664,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9614282447884285e-05,
+      "loss": 3.6717,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9605896500373765e-05,
+      "loss": 3.6574,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9597510552863245e-05,
+      "loss": 3.6472,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958914098415646e-05,
+      "loss": 3.6797,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958075503664594e-05,
+      "loss": 3.6678,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957236908913542e-05,
+      "loss": 3.6723,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95639831416249e-05,
+      "loss": 3.6633,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955561357291811e-05,
+      "loss": 3.661,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954722762540759e-05,
+      "loss": 3.656,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953884167789707e-05,
+      "loss": 3.6687,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953045573038655e-05,
+      "loss": 3.6595,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952208616167976e-05,
+      "loss": 3.6658,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951370021416924e-05,
+      "loss": 3.6744,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950531426665872e-05,
+      "loss": 3.6574,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94969283191482e-05,
+      "loss": 3.6567,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9488558750441415e-05,
+      "loss": 3.6733,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9480172802930895e-05,
+      "loss": 3.6563,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9471786855420375e-05,
+      "loss": 3.6569,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946340090790985e-05,
+      "loss": 3.6735,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945501496039933e-05,
+      "loss": 3.6618,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9446645391692544e-05,
+      "loss": 3.6543,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9438259444182024e-05,
+      "loss": 3.6641,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94298734966715e-05,
+      "loss": 3.6528,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942148754916098e-05,
+      "loss": 3.654,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941311798045419e-05,
+      "loss": 3.6746,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940473203294367e-05,
+      "loss": 3.6666,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939634608543315e-05,
+      "loss": 3.6712,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938796013792263e-05,
+      "loss": 3.6664,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937959056921585e-05,
+      "loss": 3.6845,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937120462170532e-05,
+      "loss": 3.6574,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93628186741948e-05,
+      "loss": 3.6703,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935443272668428e-05,
+      "loss": 3.6694,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93460631579775e-05,
+      "loss": 3.6591,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933767721046697e-05,
+      "loss": 3.6592,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932929126295645e-05,
+      "loss": 3.6766,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932090531544593e-05,
+      "loss": 3.6699,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9312535746739146e-05,
+      "loss": 3.6716,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9304149799228626e-05,
+      "loss": 3.653,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9295763851718106e-05,
+      "loss": 3.647,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9287377904207586e-05,
+      "loss": 3.667,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9279008335500795e-05,
+      "loss": 3.6731,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9270622387990275e-05,
+      "loss": 3.6676,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9262236440479755e-05,
+      "loss": 3.6653,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9253850492969235e-05,
+      "loss": 3.6515,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9245464545458715e-05,
+      "loss": 3.6613,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9237094976751924e-05,
+      "loss": 3.6564,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9228709029241404e-05,
+      "loss": 3.6585,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9220323081730884e-05,
+      "loss": 3.6706,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211937134220364e-05,
+      "loss": 3.6775,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920356756551358e-05,
+      "loss": 3.6698,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919518161800306e-05,
+      "loss": 3.6577,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918679567049254e-05,
+      "loss": 3.6568,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917840972298202e-05,
+      "loss": 3.6618,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917004015427523e-05,
+      "loss": 3.6693,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916165420676471e-05,
+      "loss": 3.6595,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915326825925419e-05,
+      "loss": 3.6717,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914488231174367e-05,
+      "loss": 3.6612,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913651274303688e-05,
+      "loss": 3.6533,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912812679552636e-05,
+      "loss": 3.6603,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911974084801584e-05,
+      "loss": 3.6524,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911135490050532e-05,
+      "loss": 3.6538,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9102985331798534e-05,
+      "loss": 3.6653,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9094599384288014e-05,
+      "loss": 3.6601,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9086213436777494e-05,
+      "loss": 3.6529,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9077827489266973e-05,
+      "loss": 3.6531,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906945792056018e-05,
+      "loss": 3.652,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906107197304966e-05,
+      "loss": 3.6668,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905268602553914e-05,
+      "loss": 3.6533,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904430007802862e-05,
+      "loss": 3.6768,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903593050932183e-05,
+      "loss": 3.672,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902754456181131e-05,
+      "loss": 3.6658,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901915861430079e-05,
+      "loss": 3.6579,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901077266679027e-05,
+      "loss": 3.6524,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900240309808349e-05,
+      "loss": 3.664,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899401715057297e-05,
+      "loss": 3.6587,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898563120306245e-05,
+      "loss": 3.6537,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897724525555193e-05,
+      "loss": 3.6523,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8968875686845136e-05,
+      "loss": 3.6655,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8960489739334616e-05,
+      "loss": 3.6532,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8952103791824096e-05,
+      "loss": 3.6571,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8943717844313576e-05,
+      "loss": 3.6492,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8935348275606785e-05,
+      "loss": 3.6649,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8926962328096265e-05,
+      "loss": 3.6482,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8918576380585745e-05,
+      "loss": 3.6658,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8910190433075225e-05,
+      "loss": 3.6617,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890182086436844e-05,
+      "loss": 3.6678,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889343491685792e-05,
+      "loss": 3.649,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88850489693474e-05,
+      "loss": 3.6651,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887666302183688e-05,
+      "loss": 3.6554,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886829345313009e-05,
+      "loss": 3.664,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885990750561957e-05,
+      "loss": 3.6637,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885152155810905e-05,
+      "loss": 3.6547,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884313561059853e-05,
+      "loss": 3.6613,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883476604189174e-05,
+      "loss": 3.6711,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882638009438122e-05,
+      "loss": 3.6478,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88179941468707e-05,
+      "loss": 3.6598,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880960819936018e-05,
+      "loss": 3.6623,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8801238630653395e-05,
+      "loss": 3.6573,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8792852683142875e-05,
+      "loss": 3.6509,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8784466735632355e-05,
+      "loss": 3.6566,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8776080788121834e-05,
+      "loss": 3.6659,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8767711219415044e-05,
+      "loss": 3.644,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8759325271904524e-05,
+      "loss": 3.6592,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8463475704193115,
+      "eval_runtime": 309.9084,
+      "eval_samples_per_second": 1231.302,
+      "eval_steps_per_second": 38.479,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8750939324394004e-05,
+      "loss": 3.653,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8742553376883483e-05,
+      "loss": 3.647,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.873418380817669e-05,
+      "loss": 3.6512,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.872579786066617e-05,
+      "loss": 3.6583,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871741191315565e-05,
+      "loss": 3.6721,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870902596564513e-05,
+      "loss": 3.6506,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870065639693835e-05,
+      "loss": 3.6515,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869227044942783e-05,
+      "loss": 3.6514,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.868388450191731e-05,
+      "loss": 3.6417,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.867549855440679e-05,
+      "loss": 3.6646,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86671289857e-05,
+      "loss": 3.6502,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865874303818948e-05,
+      "loss": 3.6538,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865035709067896e-05,
+      "loss": 3.6579,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864197114316844e-05,
+      "loss": 3.6431,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8633601574461646e-05,
+      "loss": 3.6449,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8625215626951126e-05,
+      "loss": 3.6461,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8616829679440606e-05,
+      "loss": 3.6443,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8608443731930086e-05,
+      "loss": 3.6512,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86000741632233e-05,
+      "loss": 3.6451,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859168821571278e-05,
+      "loss": 3.6554,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858330226820226e-05,
+      "loss": 3.6562,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857491632069174e-05,
+      "loss": 3.6605,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.856654675198495e-05,
+      "loss": 3.6488,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855816080447443e-05,
+      "loss": 3.646,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854977485696391e-05,
+      "loss": 3.6462,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854138890945339e-05,
+      "loss": 3.6582,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85330193407466e-05,
+      "loss": 3.6487,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852463339323608e-05,
+      "loss": 3.6443,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851624744572556e-05,
+      "loss": 3.6418,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.850786149821504e-05,
+      "loss": 3.6404,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8499491929508256e-05,
+      "loss": 3.6396,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8491105981997736e-05,
+      "loss": 3.658,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8482720034487216e-05,
+      "loss": 3.6561,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8474334086976695e-05,
+      "loss": 3.6549,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8465964518269905e-05,
+      "loss": 3.645,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8457578570759385e-05,
+      "loss": 3.6488,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8449192623248865e-05,
+      "loss": 3.6588,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8440806675738344e-05,
+      "loss": 3.639,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8432437107031554e-05,
+      "loss": 3.6392,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8424051159521034e-05,
+      "loss": 3.6367,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8415665212010513e-05,
+      "loss": 3.6423,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8407279264499993e-05,
+      "loss": 3.6507,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839890969579321e-05,
+      "loss": 3.6439,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839052374828269e-05,
+      "loss": 3.6403,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838213780077217e-05,
+      "loss": 3.6523,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837375185326164e-05,
+      "loss": 3.6446,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836538228455486e-05,
+      "loss": 3.6447,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835699633704434e-05,
+      "loss": 3.635,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834861038953382e-05,
+      "loss": 3.6164,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834022444202329e-05,
+      "loss": 3.6561,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833185487331651e-05,
+      "loss": 3.6397,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832346892580599e-05,
+      "loss": 3.65,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831508297829547e-05,
+      "loss": 3.6357,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830669703078495e-05,
+      "loss": 3.6384,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.829832746207816e-05,
+      "loss": 3.6304,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828994151456764e-05,
+      "loss": 3.6413,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8281555567057116e-05,
+      "loss": 3.6347,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8273169619546596e-05,
+      "loss": 3.64,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826480005083981e-05,
+      "loss": 3.6515,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.825641410332929e-05,
+      "loss": 3.6329,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8248028155818765e-05,
+      "loss": 3.6299,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8239642208308245e-05,
+      "loss": 3.6503,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8231256260797725e-05,
+      "loss": 3.6331,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822288669209094e-05,
+      "loss": 3.631,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8214500744580414e-05,
+      "loss": 3.6455,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.82061147970699e-05,
+      "loss": 3.64,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819772884955938e-05,
+      "loss": 3.6289,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818935928085259e-05,
+      "loss": 3.6392,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818097333334207e-05,
+      "loss": 3.627,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817258738583155e-05,
+      "loss": 3.6331,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816420143832103e-05,
+      "loss": 3.6461,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815583186961424e-05,
+      "loss": 3.6421,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814744592210372e-05,
+      "loss": 3.6484,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81390599745932e-05,
+      "loss": 3.6415,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813067402708268e-05,
+      "loss": 3.6586,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812230445837589e-05,
+      "loss": 3.6297,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811391851086537e-05,
+      "loss": 3.6498,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810553256335485e-05,
+      "loss": 3.6374,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8097146615844334e-05,
+      "loss": 3.6366,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8088777047137543e-05,
+      "loss": 3.6372,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8080391099627023e-05,
+      "loss": 3.6524,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80720051521165e-05,
+      "loss": 3.6443,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806361920460598e-05,
+      "loss": 3.6492,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805524963589919e-05,
+      "loss": 3.6271,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804686368838867e-05,
+      "loss": 3.6263,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803847774087815e-05,
+      "loss": 3.6392,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803009179336763e-05,
+      "loss": 3.6479,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802172222466084e-05,
+      "loss": 3.6374,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.801333627715032e-05,
+      "loss": 3.6425,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.80049503296398e-05,
+      "loss": 3.6289,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.799656438212929e-05,
+      "loss": 3.6362,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79881948134225e-05,
+      "loss": 3.6335,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797980886591198e-05,
+      "loss": 3.6312,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797142291840146e-05,
+      "loss": 3.6462,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796303697089094e-05,
+      "loss": 3.652,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7954667402184146e-05,
+      "loss": 3.6471,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7946281454673626e-05,
+      "loss": 3.6315,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7937895507163106e-05,
+      "loss": 3.633,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7929509559652586e-05,
+      "loss": 3.638,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7921139990945795e-05,
+      "loss": 3.6444,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7912754043435275e-05,
+      "loss": 3.6371,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7904368095924755e-05,
+      "loss": 3.6458,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789599852721797e-05,
+      "loss": 3.6376,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788761257970745e-05,
+      "loss": 3.6319,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787922663219693e-05,
+      "loss": 3.6346,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787084068468641e-05,
+      "loss": 3.6271,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786247111597962e-05,
+      "loss": 3.6304,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78540851684691e-05,
+      "loss": 3.6416,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.784569922095858e-05,
+      "loss": 3.6371,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.783731327344806e-05,
+      "loss": 3.6276,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782894370474127e-05,
+      "loss": 3.6309,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782055775723075e-05,
+      "loss": 3.622,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781217180972023e-05,
+      "loss": 3.6412,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780378586220971e-05,
+      "loss": 3.63,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7795416293502925e-05,
+      "loss": 3.6494,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7787030345992404e-05,
+      "loss": 3.6438,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7778644398481884e-05,
+      "loss": 3.6475,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7770258450971364e-05,
+      "loss": 3.6317,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7761888882264573e-05,
+      "loss": 3.6291,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7753502934754053e-05,
+      "loss": 3.6404,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7745116987243533e-05,
+      "loss": 3.639,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.773673103973301e-05,
+      "loss": 3.6275,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772834509222249e-05,
+      "loss": 3.6274,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77199755235157e-05,
+      "loss": 3.639,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771158957600518e-05,
+      "loss": 3.6308,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770320362849466e-05,
+      "loss": 3.6334,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769483405978788e-05,
+      "loss": 3.6286,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768644811227736e-05,
+      "loss": 3.6346,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767806216476684e-05,
+      "loss": 3.6266,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766967621725632e-05,
+      "loss": 3.6394,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766130664854953e-05,
+      "loss": 3.6367,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765292070103901e-05,
+      "loss": 3.6495,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764453475352849e-05,
+      "loss": 3.6238,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763614880601797e-05,
+      "loss": 3.6409,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7627779237311176e-05,
+      "loss": 3.6311,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7619393289800656e-05,
+      "loss": 3.6414,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7611007342290136e-05,
+      "loss": 3.6361,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7602621394779616e-05,
+      "loss": 3.6309,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759425182607283e-05,
+      "loss": 3.6363,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758586587856231e-05,
+      "loss": 3.6518,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.757747993105179e-05,
+      "loss": 3.623,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756909398354127e-05,
+      "loss": 3.6393,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756072441483448e-05,
+      "loss": 3.6348,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755233846732396e-05,
+      "loss": 3.6329,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754395251981344e-05,
+      "loss": 3.6267,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753556657230292e-05,
+      "loss": 3.6336,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752719700359613e-05,
+      "loss": 3.6376,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751881105608561e-05,
+      "loss": 3.6295,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751042510857509e-05,
+      "loss": 3.635,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8439929485321045,
+      "eval_runtime": 310.3463,
+      "eval_samples_per_second": 1229.565,
+      "eval_steps_per_second": 38.425,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750203916106457e-05,
+      "loss": 3.6269,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7493653213554056e-05,
+      "loss": 3.6231,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7485267266043536e-05,
+      "loss": 3.6276,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7476881318533016e-05,
+      "loss": 3.6366,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7468511749826225e-05,
+      "loss": 3.647,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7460125802315705e-05,
+      "loss": 3.6303,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7451739854805185e-05,
+      "loss": 3.6294,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7443353907294665e-05,
+      "loss": 3.627,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7434984338587874e-05,
+      "loss": 3.6216,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7426598391077354e-05,
+      "loss": 3.6382,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7418212443566834e-05,
+      "loss": 3.6265,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7409826496056314e-05,
+      "loss": 3.6341,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740145692734952e-05,
+      "loss": 3.6338,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7393070979839e-05,
+      "loss": 3.6205,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738468503232849e-05,
+      "loss": 3.6226,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737629908481796e-05,
+      "loss": 3.6194,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736792951611118e-05,
+      "loss": 3.6227,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735954356860066e-05,
+      "loss": 3.6237,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735115762109014e-05,
+      "loss": 3.6239,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734277167357961e-05,
+      "loss": 3.6337,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733440210487283e-05,
+      "loss": 3.6313,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732601615736231e-05,
+      "loss": 3.6358,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731763020985179e-05,
+      "loss": 3.6264,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730924426234126e-05,
+      "loss": 3.6225,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730087469363448e-05,
+      "loss": 3.6266,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729248874612396e-05,
+      "loss": 3.6334,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728410279861344e-05,
+      "loss": 3.6216,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727571685110292e-05,
+      "loss": 3.6246,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726734728239613e-05,
+      "loss": 3.6178,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725896133488561e-05,
+      "loss": 3.6155,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7250575387375086e-05,
+      "loss": 3.6176,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7242189439864566e-05,
+      "loss": 3.6296,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723381987115778e-05,
+      "loss": 3.6358,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722543392364726e-05,
+      "loss": 3.6276,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7217047976136735e-05,
+      "loss": 3.6252,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7208662028626215e-05,
+      "loss": 3.6207,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720029245991943e-05,
+      "loss": 3.6384,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719190651240891e-05,
+      "loss": 3.6189,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.718352056489839e-05,
+      "loss": 3.6165,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717513461738787e-05,
+      "loss": 3.6134,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7166765048681086e-05,
+      "loss": 3.6226,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.715837910117056e-05,
+      "loss": 3.6217,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714999315366004e-05,
+      "loss": 3.6234,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714160720614952e-05,
+      "loss": 3.6203,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7133237637442735e-05,
+      "loss": 3.627,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.712485168993221e-05,
+      "loss": 3.6188,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711646574242169e-05,
+      "loss": 3.6184,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.710807979491117e-05,
+      "loss": 3.6156,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7099710226204384e-05,
+      "loss": 3.5961,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7091324278693864e-05,
+      "loss": 3.6337,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7082938331183344e-05,
+      "loss": 3.6151,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7074552383672824e-05,
+      "loss": 3.6279,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.706618281496603e-05,
+      "loss": 3.6156,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.705779686745551e-05,
+      "loss": 3.6146,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704941091994499e-05,
+      "loss": 3.605,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704102497243447e-05,
+      "loss": 3.6181,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703265540372768e-05,
+      "loss": 3.6157,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702426945621716e-05,
+      "loss": 3.6105,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.701588350870664e-05,
+      "loss": 3.6308,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.700749756119612e-05,
+      "loss": 3.6154,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699912799248934e-05,
+      "loss": 3.6035,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699074204497882e-05,
+      "loss": 3.6247,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69823560974683e-05,
+      "loss": 3.6124,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697397014995778e-05,
+      "loss": 3.6102,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696560058125099e-05,
+      "loss": 3.6228,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695721463374047e-05,
+      "loss": 3.6157,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694882868622995e-05,
+      "loss": 3.6084,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694044273871943e-05,
+      "loss": 3.6187,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6932073170012636e-05,
+      "loss": 3.6052,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6923687222502116e-05,
+      "loss": 3.6087,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6915301274991596e-05,
+      "loss": 3.6224,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6906915327481076e-05,
+      "loss": 3.6201,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689854575877429e-05,
+      "loss": 3.6248,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689015981126377e-05,
+      "loss": 3.6164,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688177386375325e-05,
+      "loss": 3.638,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.687340429504646e-05,
+      "loss": 3.6119,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686501834753594e-05,
+      "loss": 3.6262,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685663240002542e-05,
+      "loss": 3.6162,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68482464525149e-05,
+      "loss": 3.6132,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683987688380811e-05,
+      "loss": 3.6143,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683149093629759e-05,
+      "loss": 3.6286,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682310498878707e-05,
+      "loss": 3.6179,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681471904127655e-05,
+      "loss": 3.6304,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680633309376603e-05,
+      "loss": 3.6037,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679794714625551e-05,
+      "loss": 3.6031,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6789561198744996e-05,
+      "loss": 3.6176,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6781175251234476e-05,
+      "loss": 3.6267,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6772805682527685e-05,
+      "loss": 3.6141,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6764419735017165e-05,
+      "loss": 3.6176,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6756033787506645e-05,
+      "loss": 3.6127,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6747647839996125e-05,
+      "loss": 3.6108,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6739278271289334e-05,
+      "loss": 3.6142,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6730892323778814e-05,
+      "loss": 3.6085,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6722506376268294e-05,
+      "loss": 3.623,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6714120428757774e-05,
+      "loss": 3.633,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.670575086005098e-05,
+      "loss": 3.6275,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669736491254046e-05,
+      "loss": 3.6081,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668897896502995e-05,
+      "loss": 3.6082,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668059301751943e-05,
+      "loss": 3.6147,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667222344881264e-05,
+      "loss": 3.6199,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.666383750130212e-05,
+      "loss": 3.6184,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66554515537916e-05,
+      "loss": 3.6215,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664706560628108e-05,
+      "loss": 3.619,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663869603757429e-05,
+      "loss": 3.6092,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663031009006377e-05,
+      "loss": 3.6125,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662192414255325e-05,
+      "loss": 3.6058,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.661353819504273e-05,
+      "loss": 3.6089,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.660516862633594e-05,
+      "loss": 3.6205,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.659678267882542e-05,
+      "loss": 3.6148,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6588396731314903e-05,
+      "loss": 3.6053,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6580010783804383e-05,
+      "loss": 3.6107,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657164121509759e-05,
+      "loss": 3.5986,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656325526758707e-05,
+      "loss": 3.6199,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655486932007655e-05,
+      "loss": 3.6137,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654648337256603e-05,
+      "loss": 3.6223,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.653813018266297e-05,
+      "loss": 3.6242,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652974423515245e-05,
+      "loss": 3.6264,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652135828764193e-05,
+      "loss": 3.611,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651297234013141e-05,
+      "loss": 3.6104,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650458639262089e-05,
+      "loss": 3.616,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649620044511037e-05,
+      "loss": 3.6177,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648781449759985e-05,
+      "loss": 3.6072,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647942855008934e-05,
+      "loss": 3.6036,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6471058981382546e-05,
+      "loss": 3.6182,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6462673033872026e-05,
+      "loss": 3.608,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6454287086361506e-05,
+      "loss": 3.6095,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6445901138850986e-05,
+      "loss": 3.6068,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6437531570144195e-05,
+      "loss": 3.6171,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6429145622633675e-05,
+      "loss": 3.6019,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6420759675123155e-05,
+      "loss": 3.6198,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6412373727612635e-05,
+      "loss": 3.6155,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6404004158905844e-05,
+      "loss": 3.6278,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6395618211395324e-05,
+      "loss": 3.5994,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6387232263884804e-05,
+      "loss": 3.6179,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637884631637429e-05,
+      "loss": 3.6181,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63704767476675e-05,
+      "loss": 3.6157,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636209080015698e-05,
+      "loss": 3.6166,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635370485264646e-05,
+      "loss": 3.6091,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634531890513593e-05,
+      "loss": 3.6114,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633694933642915e-05,
+      "loss": 3.6299,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632856338891863e-05,
+      "loss": 3.6048,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632017744140811e-05,
+      "loss": 3.6152,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631179149389758e-05,
+      "loss": 3.6141,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63034219251908e-05,
+      "loss": 3.6094,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629503597768028e-05,
+      "loss": 3.6099,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628665003016976e-05,
+      "loss": 3.6133,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627826408265924e-05,
+      "loss": 3.6168,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6269894513952454e-05,
+      "loss": 3.6132,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6261508566441933e-05,
+      "loss": 3.6089,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8413844108581543,
+      "eval_runtime": 311.9356,
+      "eval_samples_per_second": 1223.3,
+      "eval_steps_per_second": 38.229,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.625312261893141e-05,
+      "loss": 3.607,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6244736671420887e-05,
+      "loss": 3.5995,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.62363671027141e-05,
+      "loss": 3.608,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622798115520358e-05,
+      "loss": 3.6159,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6219595207693056e-05,
+      "loss": 3.6223,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621122563898627e-05,
+      "loss": 3.6101,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620283969147575e-05,
+      "loss": 3.6091,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.619445374396523e-05,
+      "loss": 3.6079,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.618606779645471e-05,
+      "loss": 3.5986,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617769822774793e-05,
+      "loss": 3.6152,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616931228023741e-05,
+      "loss": 3.6072,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616092633272688e-05,
+      "loss": 3.614,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615254038521636e-05,
+      "loss": 3.6118,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6144170816509576e-05,
+      "loss": 3.5973,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6135784868999056e-05,
+      "loss": 3.6016,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612739892148853e-05,
+      "loss": 3.5952,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611901297397801e-05,
+      "loss": 3.6072,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6110643405271225e-05,
+      "loss": 3.6037,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6102257457760705e-05,
+      "loss": 3.6011,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6093871510250185e-05,
+      "loss": 3.6094,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6085485562739665e-05,
+      "loss": 3.611,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607711599403288e-05,
+      "loss": 3.6156,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6068730046522354e-05,
+      "loss": 3.6073,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6060344099011834e-05,
+      "loss": 3.6049,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6051958151501314e-05,
+      "loss": 3.6062,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.604358858279453e-05,
+      "loss": 3.6109,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6035202635284e-05,
+      "loss": 3.6002,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602681668777348e-05,
+      "loss": 3.6049,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601843074026296e-05,
+      "loss": 3.5994,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601006117155617e-05,
+      "loss": 3.595,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600167522404566e-05,
+      "loss": 3.5989,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599328927653514e-05,
+      "loss": 3.6066,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598490332902462e-05,
+      "loss": 3.6075,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597653376031783e-05,
+      "loss": 3.6169,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596814781280731e-05,
+      "loss": 3.6002,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595976186529679e-05,
+      "loss": 3.6051,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595137591778627e-05,
+      "loss": 3.6178,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.594300634907948e-05,
+      "loss": 3.5949,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.593462040156896e-05,
+      "loss": 3.5967,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.592623445405844e-05,
+      "loss": 3.5941,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591784850654792e-05,
+      "loss": 3.6009,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5909478937841126e-05,
+      "loss": 3.6013,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590109299033061e-05,
+      "loss": 3.6007,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589270704282009e-05,
+      "loss": 3.6034,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588432109530957e-05,
+      "loss": 3.603,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587595152660278e-05,
+      "loss": 3.6064,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586756557909226e-05,
+      "loss": 3.5936,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585917963158174e-05,
+      "loss": 3.5964,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585079368407122e-05,
+      "loss": 3.5732,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584242411536443e-05,
+      "loss": 3.6113,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583403816785391e-05,
+      "loss": 3.5967,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582565222034339e-05,
+      "loss": 3.6047,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581726627283287e-05,
+      "loss": 3.5987,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580889670412608e-05,
+      "loss": 3.5944,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5800510756615566e-05,
+      "loss": 3.5844,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5792124809105046e-05,
+      "loss": 3.5975,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5783738861594526e-05,
+      "loss": 3.5993,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5775369292887735e-05,
+      "loss": 3.5887,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5766983345377215e-05,
+      "loss": 3.6105,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5758597397866695e-05,
+      "loss": 3.5989,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5750211450356175e-05,
+      "loss": 3.5779,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5741841881649384e-05,
+      "loss": 3.6061,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5733455934138864e-05,
+      "loss": 3.5959,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5725069986628344e-05,
+      "loss": 3.5904,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5716684039117824e-05,
+      "loss": 3.5995,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570831447041103e-05,
+      "loss": 3.5959,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569992852290052e-05,
+      "loss": 3.5867,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569154257539e-05,
+      "loss": 3.6009,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568317300668321e-05,
+      "loss": 3.5846,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567478705917269e-05,
+      "loss": 3.586,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566640111166217e-05,
+      "loss": 3.6052,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565801516415165e-05,
+      "loss": 3.5969,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564964559544486e-05,
+      "loss": 3.6059,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564125964793434e-05,
+      "loss": 3.6,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563287370042382e-05,
+      "loss": 3.6113,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56244877529133e-05,
+      "loss": 3.5942,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561611818420651e-05,
+      "loss": 3.6091,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.560773223669599e-05,
+      "loss": 3.5931,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5599346289185473e-05,
+      "loss": 3.5958,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5590960341674953e-05,
+      "loss": 3.5956,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558259077296816e-05,
+      "loss": 3.6052,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557420482545764e-05,
+      "loss": 3.6024,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556581887794712e-05,
+      "loss": 3.6111,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55574329304366e-05,
+      "loss": 3.5833,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554906336172981e-05,
+      "loss": 3.5835,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554067741421929e-05,
+      "loss": 3.5976,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553229146670877e-05,
+      "loss": 3.6068,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.552390551919825e-05,
+      "loss": 3.5929,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551553595049146e-05,
+      "loss": 3.6015,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550715000298094e-05,
+      "loss": 3.5932,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549876405547042e-05,
+      "loss": 3.5867,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549037810795991e-05,
+      "loss": 3.5927,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548199216044939e-05,
+      "loss": 3.5895,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5473622591742596e-05,
+      "loss": 3.6033,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5465236644232076e-05,
+      "loss": 3.61,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5456850696721556e-05,
+      "loss": 3.6111,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5448481128014765e-05,
+      "loss": 3.5853,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5440095180504245e-05,
+      "loss": 3.5923,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5431709232993725e-05,
+      "loss": 3.5952,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5423323285483205e-05,
+      "loss": 3.6026,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5414953716776414e-05,
+      "loss": 3.5958,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5406567769265894e-05,
+      "loss": 3.6028,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5398181821755374e-05,
+      "loss": 3.5969,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538979587424486e-05,
+      "loss": 3.5905,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538142630553807e-05,
+      "loss": 3.5935,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537304035802755e-05,
+      "loss": 3.586,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536465441051703e-05,
+      "loss": 3.5906,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535626846300651e-05,
+      "loss": 3.5987,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534789889429972e-05,
+      "loss": 3.5984,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53395129467892e-05,
+      "loss": 3.5841,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533112699927868e-05,
+      "loss": 3.5926,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532274105176816e-05,
+      "loss": 3.5766,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531437148306137e-05,
+      "loss": 3.605,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.530598553555085e-05,
+      "loss": 3.5917,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.529759958804033e-05,
+      "loss": 3.6031,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5289213640529814e-05,
+      "loss": 3.6035,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5280844071823024e-05,
+      "loss": 3.6059,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5272458124312503e-05,
+      "loss": 3.5941,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5264072176801983e-05,
+      "loss": 3.5891,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.525568622929146e-05,
+      "loss": 3.5985,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524731666058467e-05,
+      "loss": 3.5957,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523893071307415e-05,
+      "loss": 3.591,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523054476556363e-05,
+      "loss": 3.5851,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522215881805311e-05,
+      "loss": 3.5942,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521378924934632e-05,
+      "loss": 3.5897,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52054033018358e-05,
+      "loss": 3.5917,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519701735432528e-05,
+      "loss": 3.5868,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518863140681477e-05,
+      "loss": 3.5956,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518026183810798e-05,
+      "loss": 3.5819,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517187589059746e-05,
+      "loss": 3.6015,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516348994308694e-05,
+      "loss": 3.5923,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515510399557642e-05,
+      "loss": 3.6095,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5146734426869626e-05,
+      "loss": 3.5836,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5138348479359106e-05,
+      "loss": 3.5971,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5129962531848586e-05,
+      "loss": 3.5937,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5121576584338066e-05,
+      "loss": 3.5968,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5113207015631275e-05,
+      "loss": 3.5976,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5104821068120755e-05,
+      "loss": 3.5901,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5096435120610235e-05,
+      "loss": 3.5949,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.508804917309972e-05,
+      "loss": 3.6071,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507967960439293e-05,
+      "loss": 3.5903,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507129365688241e-05,
+      "loss": 3.5955,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506290770937189e-05,
+      "loss": 3.5946,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505452176186137e-05,
+      "loss": 3.5918,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504615219315458e-05,
+      "loss": 3.591,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.503776624564406e-05,
+      "loss": 3.5924,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502938029813354e-05,
+      "loss": 3.5947,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502099435062302e-05,
+      "loss": 3.5938,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501262478191623e-05,
+      "loss": 3.5858,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.840613603591919,
+      "eval_runtime": 304.6263,
+      "eval_samples_per_second": 1252.653,
+      "eval_steps_per_second": 39.146,
+      "step": 915840
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 6.386433220345651e+17,
+  "trial_name": null,
+  "trial_params": null
+}