diff --git "a/checkpoint-610560/trainer_state.json" "b/checkpoint-610560/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-610560/trainer_state.json"
@@ -0,0 +1,7241 @@
+{
+  "best_metric": 3.871347188949585,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/superlative-quantifier/transformer/2/checkpoints/checkpoint-610560",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 610560,
+  "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": 10.9302,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8188,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1906,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.984,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8202,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7093,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6114,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5357,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.458,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4169,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.367,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3165,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2771,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2283,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1883,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.156,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1163,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0957,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0587,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0439,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0241,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 4.9897,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9634,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9414,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9305,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.9035,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.88,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8601,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8631,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.827,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.8175,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.8139,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.794,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.782,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7801,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7588,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7544,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.72,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7175,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6971,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6912,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6867,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6796,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6601,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6631,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6458,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6482,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.647,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.6167,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.6073,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958085003370758e-05,
+      "loss": 4.6152,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957246408619706e-05,
+      "loss": 4.6147,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 4.5893,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5837,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 4.5817,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 4.5632,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 4.5713,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.545,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5539,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5452,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5421,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5183,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5349,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.5059,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.5163,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5142,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.5066,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4884,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4799,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.4762,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4883,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4829,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4666,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4765,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379669187473754e-05,
+      "loss": 4.4778,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 4.462,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 4.4572,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 4.4485,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 4.4496,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 4.4318,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 4.4348,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 4.4301,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.4426,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 4.4281,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.41,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.415,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.4113,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 4.4149,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926231505873767e-05,
+      "loss": 4.4099,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925394549003088e-05,
+      "loss": 4.3992,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924555954252036e-05,
+      "loss": 4.3962,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923717359500984e-05,
+      "loss": 4.3902,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922878764749932e-05,
+      "loss": 4.3788,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 4.3729,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.378,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920364618377149e-05,
+      "loss": 4.3855,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919526023626097e-05,
+      "loss": 4.3759,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918687428875045e-05,
+      "loss": 4.3633,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917848834123993e-05,
+      "loss": 4.3624,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917010239372941e-05,
+      "loss": 4.3671,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916173282502262e-05,
+      "loss": 4.3689,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91533468775121e-05,
+      "loss": 4.359,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914496093000158e-05,
+      "loss": 4.3543,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913657498249106e-05,
+      "loss": 4.3449,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912818903498054e-05,
+      "loss": 4.3549,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911980308747002e-05,
+      "loss": 4.3329,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911143351876323e-05,
+      "loss": 4.338,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.3412,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3321,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.329,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.3293,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3219,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3243,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052748264993326e-05,
+      "loss": 4.322,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90443623174828e-05,
+      "loss": 4.315,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.3238,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.3207,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019237232558704e-05,
+      "loss": 4.3142,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010851285048184e-05,
+      "loss": 4.3151,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900246533753767e-05,
+      "loss": 4.3051,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899407939002715e-05,
+      "loss": 4.3052,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985693442516624e-05,
+      "loss": 4.2969,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977307495006104e-05,
+      "loss": 4.2963,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968921547495584e-05,
+      "loss": 4.2954,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.2948,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2919,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2924,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2898,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.2784,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2876,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2866,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2683,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2593,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.268,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.2813,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886832293497681e-05,
+      "loss": 4.2819,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.2675,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.264,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884316509244525e-05,
+      "loss": 4.2692,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883477914493473e-05,
+      "loss": 4.2702,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.2605,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.2528,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.2576,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.2511,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.2492,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878447983867534e-05,
+      "loss": 4.258,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.2421,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876772432245803e-05,
+      "loss": 4.2531,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875933837494751e-05,
+      "loss": 4.2492,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8750952427436986e-05,
+      "loss": 4.2306,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.220055103302002,
+      "eval_runtime": 305.3236,
+      "eval_samples_per_second": 1249.792,
+      "eval_steps_per_second": 39.057,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8742566479926466e-05,
+      "loss": 4.2358,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.2273,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8725794584905426e-05,
+      "loss": 4.2397,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717425016198635e-05,
+      "loss": 4.2308,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709039068688115e-05,
+      "loss": 4.2353,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700653121177595e-05,
+      "loss": 4.2195,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692267173667075e-05,
+      "loss": 4.2165,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683881226156555e-05,
+      "loss": 4.2018,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867551165744977e-05,
+      "loss": 4.2335,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866712570993925e-05,
+      "loss": 4.2206,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865873976242873e-05,
+      "loss": 4.2147,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865035381491821e-05,
+      "loss": 4.2215,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864198424621142e-05,
+      "loss": 4.2081,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86335982987009e-05,
+      "loss": 4.204,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862521235119038e-05,
+      "loss": 4.2005,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.2002,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860844045616934e-05,
+      "loss": 4.1968,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860005450865882e-05,
+      "loss": 4.1914,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85916685611483e-05,
+      "loss": 4.2045,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858328261363778e-05,
+      "loss": 4.21,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.194,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.1849,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.1878,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549755202399435e-05,
+      "loss": 4.1923,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.1832,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532999686182124e-05,
+      "loss": 4.1737,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524613738671604e-05,
+      "loss": 4.1762,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8516227791161084e-05,
+      "loss": 4.1862,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.1631,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.1681,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.1754,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.1744,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847433081121595e-05,
+      "loss": 4.1785,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846594486370543e-05,
+      "loss": 4.1706,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845755891619491e-05,
+      "loss": 4.17,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844917296868439e-05,
+      "loss": 4.1746,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84408033999776e-05,
+      "loss": 4.152,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.1616,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842403150495656e-05,
+      "loss": 4.1466,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841564555744604e-05,
+      "loss": 4.1504,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840727598873925e-05,
+      "loss": 4.1541,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839889004122873e-05,
+      "loss": 4.1562,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839050409371821e-05,
+      "loss": 4.1527,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382118146207687e-05,
+      "loss": 4.1476,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83737485775009e-05,
+      "loss": 4.1494,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836536262999038e-05,
+      "loss": 4.1511,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835697668247986e-05,
+      "loss": 4.1526,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834859073496934e-05,
+      "loss": 4.1382,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834022116626255e-05,
+      "loss": 4.1334,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833185159755576e-05,
+      "loss": 4.1443,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832346565004524e-05,
+      "loss": 4.1532,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831507970253472e-05,
+      "loss": 4.1341,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83066937550242e-05,
+      "loss": 4.1357,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829830780751368e-05,
+      "loss": 4.1331,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828992186000316e-05,
+      "loss": 4.1249,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828153591249264e-05,
+      "loss": 4.1337,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273166343785856e-05,
+      "loss": 4.12,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264780396275336e-05,
+      "loss": 4.1333,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256394448764816e-05,
+      "loss": 4.1292,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248008501254296e-05,
+      "loss": 4.1263,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239638932547505e-05,
+      "loss": 4.1108,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231252985036985e-05,
+      "loss": 4.1325,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222867037526465e-05,
+      "loss": 4.1058,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214481090015945e-05,
+      "loss": 4.1211,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206111521309154e-05,
+      "loss": 4.1242,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197725573798634e-05,
+      "loss": 4.1211,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189339626288114e-05,
+      "loss": 4.1063,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180953678777594e-05,
+      "loss": 4.1045,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817258411007081e-05,
+      "loss": 4.1031,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816419816256029e-05,
+      "loss": 4.1141,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815581221504977e-05,
+      "loss": 4.1159,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814742626753924e-05,
+      "loss": 4.1097,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813905669883246e-05,
+      "loss": 4.1183,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813067075132194e-05,
+      "loss": 4.1166,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812228480381141e-05,
+      "loss": 4.1133,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811389885630089e-05,
+      "loss": 4.1005,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810552928759411e-05,
+      "loss": 4.1105,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809714334008359e-05,
+      "loss": 4.1097,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808875739257306e-05,
+      "loss": 4.0883,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808037144506255e-05,
+      "loss": 4.1083,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807200187635576e-05,
+      "loss": 4.0932,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063615928845243e-05,
+      "loss": 4.1136,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055229981334717e-05,
+      "loss": 4.1024,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8046844033824197e-05,
+      "loss": 4.0809,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803847446511741e-05,
+      "loss": 4.0979,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8030088517606886e-05,
+      "loss": 4.0888,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8021702570096366e-05,
+      "loss": 4.1019,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8013316622585846e-05,
+      "loss": 4.0933,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800494705387906e-05,
+      "loss": 4.0885,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996561106368535e-05,
+      "loss": 4.09,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988175158858015e-05,
+      "loss": 4.0819,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979789211347494e-05,
+      "loss": 4.0761,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797140326383698e-05,
+      "loss": 4.0748,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796301731632646e-05,
+      "loss": 4.0811,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795464774761967e-05,
+      "loss": 4.091,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794626180010915e-05,
+      "loss": 4.0823,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793787585259863e-05,
+      "loss": 4.0729,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792948990508811e-05,
+      "loss": 4.0735,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792110395757759e-05,
+      "loss": 4.0813,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791271801006707e-05,
+      "loss": 4.0874,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790433206255655e-05,
+      "loss": 4.0773,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789594611504603e-05,
+      "loss": 4.0753,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788757654633924e-05,
+      "loss": 4.0701,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787919059882872e-05,
+      "loss": 4.0792,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78708046513182e-05,
+      "loss": 4.0576,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786241870380768e-05,
+      "loss": 4.0683,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854049135100895e-05,
+      "loss": 4.0726,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845663187590375e-05,
+      "loss": 4.0649,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837277240079855e-05,
+      "loss": 4.0598,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828891292569335e-05,
+      "loss": 4.0634,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820505345058815e-05,
+      "loss": 4.0637,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812119397548295e-05,
+      "loss": 4.0611,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803733450037775e-05,
+      "loss": 4.0655,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795363881330984e-05,
+      "loss": 4.0596,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786977933820464e-05,
+      "loss": 4.0716,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778591986309944e-05,
+      "loss": 4.0695,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777022241760315e-05,
+      "loss": 4.0642,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776183647009263e-05,
+      "loss": 4.0629,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775345052258212e-05,
+      "loss": 4.0601,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77450645750716e-05,
+      "loss": 4.0575,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773667862756108e-05,
+      "loss": 4.0515,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772829268005055e-05,
+      "loss": 4.0517,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771990673254003e-05,
+      "loss": 4.0565,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771152078502951e-05,
+      "loss": 4.0566,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770315121632272e-05,
+      "loss": 4.0532,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76947652688122e-05,
+      "loss": 4.0555,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768637932130168e-05,
+      "loss": 4.0495,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767799337379116e-05,
+      "loss": 4.0507,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766962380508437e-05,
+      "loss": 4.0499,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766123785757386e-05,
+      "loss": 4.0589,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765285191006334e-05,
+      "loss": 4.0414,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764446596255282e-05,
+      "loss": 4.031,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7636096393846026e-05,
+      "loss": 4.0382,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7627710446335506e-05,
+      "loss": 4.0548,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7619324498824986e-05,
+      "loss": 4.0538,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610938551314466e-05,
+      "loss": 4.0476,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7602568982607675e-05,
+      "loss": 4.0416,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7594183035097155e-05,
+      "loss": 4.0509,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7585797087586635e-05,
+      "loss": 4.048,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577411140076115e-05,
+      "loss": 4.0449,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569041571369324e-05,
+      "loss": 4.0379,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756067200266254e-05,
+      "loss": 4.0387,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755228605515202e-05,
+      "loss": 4.0399,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75439001076415e-05,
+      "loss": 4.0315,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753551416013098e-05,
+      "loss": 4.0499,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752712821262046e-05,
+      "loss": 4.0312,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751875864391367e-05,
+      "loss": 4.0413,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751037269640315e-05,
+      "loss": 4.0393,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750198674889263e-05,
+      "loss": 4.0253,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.047743320465088,
+      "eval_runtime": 308.0601,
+      "eval_samples_per_second": 1238.69,
+      "eval_steps_per_second": 38.71,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749360080138211e-05,
+      "loss": 4.0303,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748521485387159e-05,
+      "loss": 4.0236,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747682890636107e-05,
+      "loss": 4.033,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746844295885055e-05,
+      "loss": 4.0287,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746005701134003e-05,
+      "loss": 4.037,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745167106382951e-05,
+      "loss": 4.0209,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744328511631899e-05,
+      "loss": 4.0188,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434899168808475e-05,
+      "loss": 4.0065,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426513221297955e-05,
+      "loss": 4.039,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418127273787435e-05,
+      "loss": 4.0264,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740974132627691e-05,
+      "loss": 4.0146,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401371757570124e-05,
+      "loss": 4.0322,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739301856766706e-05,
+      "loss": 4.0135,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738463262015654e-05,
+      "loss": 4.0175,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737624667264602e-05,
+      "loss": 4.0076,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73678607251355e-05,
+      "loss": 4.0099,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735947477762498e-05,
+      "loss": 4.0145,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735108883011446e-05,
+      "loss": 4.0055,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734270288260394e-05,
+      "loss": 4.0217,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733431693509343e-05,
+      "loss": 4.0238,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732593098758291e-05,
+      "loss": 4.012,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731754504007238e-05,
+      "loss": 4.0072,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730915909256186e-05,
+      "loss": 4.0095,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730078952385508e-05,
+      "loss": 4.0136,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729240357634456e-05,
+      "loss": 4.0009,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728401762883403e-05,
+      "loss": 4.0015,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727563168132351e-05,
+      "loss": 3.9962,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726726211261673e-05,
+      "loss": 4.0139,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725887616510621e-05,
+      "loss": 3.9876,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725049021759568e-05,
+      "loss": 3.9968,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724210427008517e-05,
+      "loss": 4.0034,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723373470137838e-05,
+      "loss": 4.0051,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225348753867856e-05,
+      "loss": 4.0056,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216962806357336e-05,
+      "loss": 4.0027,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208576858846816e-05,
+      "loss": 3.998,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720020729014003e-05,
+      "loss": 4.0074,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191821342629505e-05,
+      "loss": 3.99,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183435395118985e-05,
+      "loss": 3.9936,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175049447608465e-05,
+      "loss": 3.9792,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716667987890168e-05,
+      "loss": 3.9877,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715831031019489e-05,
+      "loss": 3.9857,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714992436268437e-05,
+      "loss": 3.9968,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714153841517385e-05,
+      "loss": 3.9899,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713315246766333e-05,
+      "loss": 3.9869,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712476652015281e-05,
+      "loss": 3.9902,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711638057264229e-05,
+      "loss": 3.9915,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7108011003935505e-05,
+      "loss": 3.9921,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709962505642498e-05,
+      "loss": 3.9798,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709123910891446e-05,
+      "loss": 3.9729,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708285316140394e-05,
+      "loss": 3.9937,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707446721389342e-05,
+      "loss": 3.9927,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70660812663829e-05,
+      "loss": 3.9812,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705769531887238e-05,
+      "loss": 3.9818,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704930937136186e-05,
+      "loss": 3.9754,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040939802655074e-05,
+      "loss": 3.9734,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032553855144554e-05,
+      "loss": 3.9832,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024167907634034e-05,
+      "loss": 3.9681,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015781960123514e-05,
+      "loss": 3.9813,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007396012612994e-05,
+      "loss": 3.9791,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6999010065102474e-05,
+      "loss": 3.9724,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990624117591954e-05,
+      "loss": 3.9627,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698225454888516e-05,
+      "loss": 3.9839,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697386860137464e-05,
+      "loss": 3.9611,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696548265386412e-05,
+      "loss": 3.9724,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69570967063536e-05,
+      "loss": 3.973,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694872713764681e-05,
+      "loss": 3.9775,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694034119013629e-05,
+      "loss": 3.9591,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693195524262577e-05,
+      "loss": 3.9627,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692356929511526e-05,
+      "loss": 3.9537,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691519972640847e-05,
+      "loss": 3.9724,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690681377889795e-05,
+      "loss": 3.9753,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689842783138743e-05,
+      "loss": 3.9689,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689004188387691e-05,
+      "loss": 3.9736,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688165593636639e-05,
+      "loss": 3.973,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687326998885587e-05,
+      "loss": 3.9779,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864900420149076e-05,
+      "loss": 3.9563,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856514472638556e-05,
+      "loss": 3.9703,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848128525128036e-05,
+      "loss": 3.9732,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683974257761751e-05,
+      "loss": 3.9526,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831373008910725e-05,
+      "loss": 3.9679,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822987061400205e-05,
+      "loss": 3.9562,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681460111388969e-05,
+      "loss": 3.9784,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806215166379165e-05,
+      "loss": 3.9658,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6797829218868645e-05,
+      "loss": 3.9463,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789443271358125e-05,
+      "loss": 3.9599,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6781057323847605e-05,
+      "loss": 3.9579,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772687755140814e-05,
+      "loss": 3.9656,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764301807630294e-05,
+      "loss": 3.96,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755915860119774e-05,
+      "loss": 3.956,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747529912609254e-05,
+      "loss": 3.9573,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739143965098734e-05,
+      "loss": 3.9513,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730758017588214e-05,
+      "loss": 3.9479,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722372070077694e-05,
+      "loss": 3.9392,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6713986122567174e-05,
+      "loss": 3.952,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670563293266412e-05,
+      "loss": 3.9613,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66972469851536e-05,
+      "loss": 3.955,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668886103764308e-05,
+      "loss": 3.9434,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668047509013256e-05,
+      "loss": 3.9473,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667208914262204e-05,
+      "loss": 3.9492,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666370319511152e-05,
+      "loss": 3.961,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655317247601e-05,
+      "loss": 3.9527,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664693130009048e-05,
+      "loss": 3.9501,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663856173138369e-05,
+      "loss": 3.9414,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66301921626769e-05,
+      "loss": 3.9555,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621806215166383e-05,
+      "loss": 3.9313,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613420267655863e-05,
+      "loss": 3.9461,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660505069894907e-05,
+      "loss": 3.9446,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659666475143855e-05,
+      "loss": 3.9443,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658827880392803e-05,
+      "loss": 3.9358,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657989285641751e-05,
+      "loss": 3.9413,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657150690890699e-05,
+      "loss": 3.9407,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656312096139647e-05,
+      "loss": 3.9415,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655473501388595e-05,
+      "loss": 3.9425,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654634906637543e-05,
+      "loss": 3.9399,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653797949766864e-05,
+      "loss": 3.9472,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652959355015812e-05,
+      "loss": 3.9489,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65212076026476e-05,
+      "loss": 3.9477,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651282165513708e-05,
+      "loss": 3.9431,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65044520864303e-05,
+      "loss": 3.9404,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649606613891978e-05,
+      "loss": 3.9344,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648768019140926e-05,
+      "loss": 3.9349,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647929424389874e-05,
+      "loss": 3.936,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647090829638822e-05,
+      "loss": 3.9333,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462538727681426e-05,
+      "loss": 3.9427,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454152780170906e-05,
+      "loss": 3.9347,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445766832660386e-05,
+      "loss": 3.9405,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437380885149866e-05,
+      "loss": 3.936,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429011316443075e-05,
+      "loss": 3.9335,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420625368932555e-05,
+      "loss": 3.9306,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412239421422035e-05,
+      "loss": 3.9469,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640385347391152e-05,
+      "loss": 3.9258,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395467526401e-05,
+      "loss": 3.9242,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387081578890475e-05,
+      "loss": 3.9195,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378695631379955e-05,
+      "loss": 3.9389,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637032606267317e-05,
+      "loss": 3.9397,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636194011516265e-05,
+      "loss": 3.9381,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353554167652124e-05,
+      "loss": 3.9278,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345168220141604e-05,
+      "loss": 3.9406,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336782272631084e-05,
+      "loss": 3.9333,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328396325120564e-05,
+      "loss": 3.9361,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6320010377610043e-05,
+      "loss": 3.9257,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631164080890326e-05,
+      "loss": 3.93,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630325486139274e-05,
+      "loss": 3.9297,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629486891388222e-05,
+      "loss": 3.9184,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62864829663717e-05,
+      "loss": 3.9387,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627811339766491e-05,
+      "loss": 3.9284,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626972745015439e-05,
+      "loss": 3.9272,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626134150264387e-05,
+      "loss": 3.9297,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625295555513335e-05,
+      "loss": 3.9163,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.972442626953125,
+      "eval_runtime": 310.6739,
+      "eval_samples_per_second": 1228.269,
+      "eval_steps_per_second": 38.384,
+      "step": 228960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.624456960762283e-05,
+      "loss": 3.9209,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.623618366011231e-05,
+      "loss": 3.9174,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.622779771260179e-05,
+      "loss": 3.924,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621941176509127e-05,
+      "loss": 3.9221,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621102581758075e-05,
+      "loss": 3.9342,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.620263987007023e-05,
+      "loss": 3.915,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.619425392255971e-05,
+      "loss": 3.9104,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.618586797504919e-05,
+      "loss": 3.9024,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6177498406342404e-05,
+      "loss": 3.9296,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6169112458831884e-05,
+      "loss": 3.9246,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6160726511321364e-05,
+      "loss": 3.9059,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.615234056381084e-05,
+      "loss": 3.928,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.614397099510405e-05,
+      "loss": 3.912,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.613558504759353e-05,
+      "loss": 3.9142,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6127199100083006e-05,
+      "loss": 3.9079,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6118813152572486e-05,
+      "loss": 3.9071,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6110427205061966e-05,
+      "loss": 3.9103,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6102041257551446e-05,
+      "loss": 3.9052,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6093655310040926e-05,
+      "loss": 3.9198,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6085269362530406e-05,
+      "loss": 3.9196,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.607688341501989e-05,
+      "loss": 3.9149,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.60685138463131e-05,
+      "loss": 3.9094,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606012789880258e-05,
+      "loss": 3.9054,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.605174195129206e-05,
+      "loss": 3.9164,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.604335600378154e-05,
+      "loss": 3.8995,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.603498643507475e-05,
+      "loss": 3.9041,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.602660048756423e-05,
+      "loss": 3.8986,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.601821454005371e-05,
+      "loss": 3.9125,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.600982859254319e-05,
+      "loss": 3.8906,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600144264503267e-05,
+      "loss": 3.8974,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599307307632588e-05,
+      "loss": 3.9088,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598468712881536e-05,
+      "loss": 3.904,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5976301181304846e-05,
+      "loss": 3.9083,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5967915233794326e-05,
+      "loss": 3.9078,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5959529286283806e-05,
+      "loss": 3.9006,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5951143338773286e-05,
+      "loss": 3.9093,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5942773770066495e-05,
+      "loss": 3.8979,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5934387822555975e-05,
+      "loss": 3.8913,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5926001875045455e-05,
+      "loss": 3.8866,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5917615927534935e-05,
+      "loss": 3.8929,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5909246358828144e-05,
+      "loss": 3.8911,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5900860411317624e-05,
+      "loss": 3.9055,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5892474463807104e-05,
+      "loss": 3.894,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5884088516296584e-05,
+      "loss": 3.8931,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5875702568786064e-05,
+      "loss": 3.8997,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5867316621275544e-05,
+      "loss": 3.8933,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585894705256876e-05,
+      "loss": 3.9027,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585056110505824e-05,
+      "loss": 3.8867,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584217515754772e-05,
+      "loss": 3.8743,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58337892100372e-05,
+      "loss": 3.9054,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.582540326252667e-05,
+      "loss": 3.8967,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.581703369381989e-05,
+      "loss": 3.8952,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580864774630937e-05,
+      "loss": 3.8865,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580026179879884e-05,
+      "loss": 3.8853,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.579187585128832e-05,
+      "loss": 3.8809,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.57834899037778e-05,
+      "loss": 3.8904,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.577512033507102e-05,
+      "loss": 3.879,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.57667343875605e-05,
+      "loss": 3.8857,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.575834844004998e-05,
+      "loss": 3.894,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.574996249253946e-05,
+      "loss": 3.881,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.574157654502894e-05,
+      "loss": 3.872,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.573319059751842e-05,
+      "loss": 3.8979,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.57248046500079e-05,
+      "loss": 3.872,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5716435081301106e-05,
+      "loss": 3.8805,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5708049133790586e-05,
+      "loss": 3.8883,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5699663186280066e-05,
+      "loss": 3.8911,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5691277238769546e-05,
+      "loss": 3.8715,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5682907670062755e-05,
+      "loss": 3.8768,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5674521722552235e-05,
+      "loss": 3.863,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5666135775041715e-05,
+      "loss": 3.8889,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.56577498275312e-05,
+      "loss": 3.8861,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564936388002068e-05,
+      "loss": 3.8827,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564099431131389e-05,
+      "loss": 3.8844,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.563260836380337e-05,
+      "loss": 3.8873,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.562422241629285e-05,
+      "loss": 3.8938,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561583646878233e-05,
+      "loss": 3.8702,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560746690007554e-05,
+      "loss": 3.8835,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559908095256502e-05,
+      "loss": 3.8881,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.55906950050545e-05,
+      "loss": 3.866,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.558230905754398e-05,
+      "loss": 3.8838,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.557392311003346e-05,
+      "loss": 3.8743,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.556555354132667e-05,
+      "loss": 3.8886,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5557167593816156e-05,
+      "loss": 3.8823,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5548781646305635e-05,
+      "loss": 3.8642,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5540395698795115e-05,
+      "loss": 3.8761,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5532026130088325e-05,
+      "loss": 3.8729,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5523640182577804e-05,
+      "loss": 3.8846,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5515254235067284e-05,
+      "loss": 3.8731,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5506868287556764e-05,
+      "loss": 3.8711,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5498498718849973e-05,
+      "loss": 3.8738,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5490112771339453e-05,
+      "loss": 3.8725,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5481726823828933e-05,
+      "loss": 3.8668,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.547334087631841e-05,
+      "loss": 3.8568,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546495492880789e-05,
+      "loss": 3.8702,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.545658536010111e-05,
+      "loss": 3.8796,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544819941259059e-05,
+      "loss": 3.8789,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543981346508007e-05,
+      "loss": 3.8565,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543142751756955e-05,
+      "loss": 3.8679,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.542305794886276e-05,
+      "loss": 3.8675,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.541467200135224e-05,
+      "loss": 3.88,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.540628605384172e-05,
+      "loss": 3.8702,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.53979001063312e-05,
+      "loss": 3.8701,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538951415882068e-05,
+      "loss": 3.8609,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538114459011389e-05,
+      "loss": 3.8759,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.537275864260337e-05,
+      "loss": 3.852,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.536437269509285e-05,
+      "loss": 3.8675,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.535600312638606e-05,
+      "loss": 3.8639,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.534761717887554e-05,
+      "loss": 3.866,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533923123136502e-05,
+      "loss": 3.8567,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.53308452838545e-05,
+      "loss": 3.8699,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5322459336343976e-05,
+      "loss": 3.8573,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5314073388833456e-05,
+      "loss": 3.8634,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5305687441322936e-05,
+      "loss": 3.8637,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5297301493812416e-05,
+      "loss": 3.8639,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5288931925105625e-05,
+      "loss": 3.8696,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5280545977595105e-05,
+      "loss": 3.8701,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5272160030084585e-05,
+      "loss": 3.8672,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5263774082574065e-05,
+      "loss": 3.869,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.525540451386728e-05,
+      "loss": 3.8641,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.524701856635676e-05,
+      "loss": 3.8573,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523863261884624e-05,
+      "loss": 3.8545,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523024667133572e-05,
+      "loss": 3.8572,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.52218607238252e-05,
+      "loss": 3.8604,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.521349115511841e-05,
+      "loss": 3.8641,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.520510520760789e-05,
+      "loss": 3.857,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.519671926009737e-05,
+      "loss": 3.8643,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.518833331258685e-05,
+      "loss": 3.8561,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517996374388006e-05,
+      "loss": 3.8575,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517157779636954e-05,
+      "loss": 3.8556,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.516319184885902e-05,
+      "loss": 3.8692,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51548059013485e-05,
+      "loss": 3.8543,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5146419953837985e-05,
+      "loss": 3.847,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5138050385131194e-05,
+      "loss": 3.8415,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5129664437620674e-05,
+      "loss": 3.8623,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5121278490110154e-05,
+      "loss": 3.864,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5112892542599634e-05,
+      "loss": 3.8663,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.510452297389284e-05,
+      "loss": 3.8554,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.509613702638232e-05,
+      "loss": 3.8657,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.50877510788718e-05,
+      "loss": 3.8581,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507936513136128e-05,
+      "loss": 3.8592,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507097918385076e-05,
+      "loss": 3.8529,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.506260961514397e-05,
+      "loss": 3.8554,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.505422366763345e-05,
+      "loss": 3.8581,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.504583772012293e-05,
+      "loss": 3.8431,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503745177261242e-05,
+      "loss": 3.8655,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502908220390563e-05,
+      "loss": 3.8586,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502069625639511e-05,
+      "loss": 3.8515,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.501231030888459e-05,
+      "loss": 3.8521,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.500392436137407e-05,
+      "loss": 3.8489,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.929913282394409,
+      "eval_runtime": 303.7543,
+      "eval_samples_per_second": 1256.249,
+      "eval_steps_per_second": 39.259,
+      "step": 305280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.499553841386355e-05,
+      "loss": 3.8502,
+      "step": 305664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.498716884515676e-05,
+      "loss": 3.8476,
+      "step": 306176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.497878289764624e-05,
+      "loss": 3.8522,
+      "step": 306688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.497039695013572e-05,
+      "loss": 3.8494,
+      "step": 307200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4962027381428926e-05,
+      "loss": 3.8617,
+      "step": 307712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4953641433918406e-05,
+      "loss": 3.8474,
+      "step": 308224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4945255486407886e-05,
+      "loss": 3.8369,
+      "step": 308736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.493686953889737e-05,
+      "loss": 3.837,
+      "step": 309248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492848359138685e-05,
+      "loss": 3.8549,
+      "step": 309760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492011402268006e-05,
+      "loss": 3.8529,
+      "step": 310272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491172807516954e-05,
+      "loss": 3.8388,
+      "step": 310784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.490334212765902e-05,
+      "loss": 3.8538,
+      "step": 311296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.489497255895223e-05,
+      "loss": 3.8434,
+      "step": 311808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.488658661144171e-05,
+      "loss": 3.8469,
+      "step": 312320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.487820066393119e-05,
+      "loss": 3.8371,
+      "step": 312832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486981471642067e-05,
+      "loss": 3.8388,
+      "step": 313344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486142876891015e-05,
+      "loss": 3.8364,
+      "step": 313856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.485304282139963e-05,
+      "loss": 3.8347,
+      "step": 314368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.484465687388911e-05,
+      "loss": 3.8508,
+      "step": 314880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.483627092637859e-05,
+      "loss": 3.8503,
+      "step": 315392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4827901357671806e-05,
+      "loss": 3.8474,
+      "step": 315904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4819515410161286e-05,
+      "loss": 3.8432,
+      "step": 316416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.481112946265076e-05,
+      "loss": 3.8354,
+      "step": 316928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.480274351514024e-05,
+      "loss": 3.8453,
+      "step": 317440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4794373946433455e-05,
+      "loss": 3.8348,
+      "step": 317952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4785987998922935e-05,
+      "loss": 3.8364,
+      "step": 318464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.477760205141241e-05,
+      "loss": 3.8296,
+      "step": 318976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476921610390189e-05,
+      "loss": 3.8407,
+      "step": 319488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4760846535195104e-05,
+      "loss": 3.8284,
+      "step": 320000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.475246058768458e-05,
+      "loss": 3.8253,
+      "step": 320512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4744074640174064e-05,
+      "loss": 3.8424,
+      "step": 321024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4735688692663544e-05,
+      "loss": 3.8335,
+      "step": 321536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4727302745153024e-05,
+      "loss": 3.8439,
+      "step": 322048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471893317644623e-05,
+      "loss": 3.8429,
+      "step": 322560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471054722893571e-05,
+      "loss": 3.8313,
+      "step": 323072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.470216128142519e-05,
+      "loss": 3.8432,
+      "step": 323584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.469377533391467e-05,
+      "loss": 3.8348,
+      "step": 324096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.468540576520788e-05,
+      "loss": 3.8247,
+      "step": 324608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.467701981769736e-05,
+      "loss": 3.8178,
+      "step": 325120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466863387018684e-05,
+      "loss": 3.8252,
+      "step": 325632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466024792267632e-05,
+      "loss": 3.8265,
+      "step": 326144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.465187835396953e-05,
+      "loss": 3.8403,
+      "step": 326656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.464349240645902e-05,
+      "loss": 3.8289,
+      "step": 327168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.46351064589485e-05,
+      "loss": 3.8271,
+      "step": 327680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.462672051143798e-05,
+      "loss": 3.8367,
+      "step": 328192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4618350942731186e-05,
+      "loss": 3.8302,
+      "step": 328704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4609964995220666e-05,
+      "loss": 3.8352,
+      "step": 329216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4601579047710146e-05,
+      "loss": 3.82,
+      "step": 329728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4593193100199626e-05,
+      "loss": 3.807,
+      "step": 330240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4584823531492835e-05,
+      "loss": 3.8394,
+      "step": 330752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4576437583982315e-05,
+      "loss": 3.836,
+      "step": 331264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4568051636471795e-05,
+      "loss": 3.8295,
+      "step": 331776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4559665688961275e-05,
+      "loss": 3.8229,
+      "step": 332288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4551296120254484e-05,
+      "loss": 3.8199,
+      "step": 332800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.454291017274397e-05,
+      "loss": 3.815,
+      "step": 333312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.453452422523345e-05,
+      "loss": 3.8246,
+      "step": 333824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.452613827772293e-05,
+      "loss": 3.8176,
+      "step": 334336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.451776870901614e-05,
+      "loss": 3.8202,
+      "step": 334848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450938276150562e-05,
+      "loss": 3.8297,
+      "step": 335360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.45009968139951e-05,
+      "loss": 3.8217,
+      "step": 335872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.449261086648458e-05,
+      "loss": 3.81,
+      "step": 336384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.448424129777779e-05,
+      "loss": 3.8346,
+      "step": 336896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.447585535026727e-05,
+      "loss": 3.8093,
+      "step": 337408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.446746940275675e-05,
+      "loss": 3.8114,
+      "step": 337920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445908345524623e-05,
+      "loss": 3.8272,
+      "step": 338432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445071388653944e-05,
+      "loss": 3.83,
+      "step": 338944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4442327939028925e-05,
+      "loss": 3.8069,
+      "step": 339456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4433941991518405e-05,
+      "loss": 3.8149,
+      "step": 339968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4425556044007885e-05,
+      "loss": 3.8016,
+      "step": 340480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4417186475301094e-05,
+      "loss": 3.8235,
+      "step": 340992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4408800527790574e-05,
+      "loss": 3.8268,
+      "step": 341504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4400414580280054e-05,
+      "loss": 3.8152,
+      "step": 342016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4392028632769534e-05,
+      "loss": 3.8224,
+      "step": 342528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.438365906406274e-05,
+      "loss": 3.8243,
+      "step": 343040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.437527311655222e-05,
+      "loss": 3.8372,
+      "step": 343552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.43668871690417e-05,
+      "loss": 3.8086,
+      "step": 344064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.435850122153118e-05,
+      "loss": 3.8229,
+      "step": 344576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.435013165282439e-05,
+      "loss": 3.8257,
+      "step": 345088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434174570531388e-05,
+      "loss": 3.8046,
+      "step": 345600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.433335975780336e-05,
+      "loss": 3.8239,
+      "step": 346112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.432497381029284e-05,
+      "loss": 3.8152,
+      "step": 346624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.431660424158605e-05,
+      "loss": 3.8258,
+      "step": 347136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.430821829407553e-05,
+      "loss": 3.8217,
+      "step": 347648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429983234656501e-05,
+      "loss": 3.8052,
+      "step": 348160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429144639905449e-05,
+      "loss": 3.8101,
+      "step": 348672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4283076830347696e-05,
+      "loss": 3.8109,
+      "step": 349184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4274690882837176e-05,
+      "loss": 3.8278,
+      "step": 349696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4266304935326656e-05,
+      "loss": 3.812,
+      "step": 350208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4257918987816136e-05,
+      "loss": 3.8097,
+      "step": 350720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4249549419109345e-05,
+      "loss": 3.8116,
+      "step": 351232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424116347159883e-05,
+      "loss": 3.8159,
+      "step": 351744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.423277752408831e-05,
+      "loss": 3.8077,
+      "step": 352256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.422439157657779e-05,
+      "loss": 3.7953,
+      "step": 352768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.421600562906727e-05,
+      "loss": 3.8121,
+      "step": 353280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.420763606036048e-05,
+      "loss": 3.8187,
+      "step": 353792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.419925011284996e-05,
+      "loss": 3.8185,
+      "step": 354304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.419086416533944e-05,
+      "loss": 3.7958,
+      "step": 354816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.418247821782892e-05,
+      "loss": 3.8078,
+      "step": 355328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.417410864912213e-05,
+      "loss": 3.8077,
+      "step": 355840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.416572270161161e-05,
+      "loss": 3.817,
+      "step": 356352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.415733675410109e-05,
+      "loss": 3.8152,
+      "step": 356864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.414895080659057e-05,
+      "loss": 3.8104,
+      "step": 357376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.414058123788378e-05,
+      "loss": 3.804,
+      "step": 357888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4132195290373266e-05,
+      "loss": 3.8152,
+      "step": 358400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4123809342862746e-05,
+      "loss": 3.7943,
+      "step": 358912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4115423395352226e-05,
+      "loss": 3.8096,
+      "step": 359424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4107053826645435e-05,
+      "loss": 3.8031,
+      "step": 359936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4098667879134915e-05,
+      "loss": 3.8114,
+      "step": 360448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4090281931624395e-05,
+      "loss": 3.7944,
+      "step": 360960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4081895984113875e-05,
+      "loss": 3.8125,
+      "step": 361472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4073526415407084e-05,
+      "loss": 3.8016,
+      "step": 361984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4065140467896564e-05,
+      "loss": 3.8057,
+      "step": 362496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4056754520386044e-05,
+      "loss": 3.8053,
+      "step": 363008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4048368572875524e-05,
+      "loss": 3.8053,
+      "step": 363520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403999900416873e-05,
+      "loss": 3.811,
+      "step": 364032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403161305665822e-05,
+      "loss": 3.8154,
+      "step": 364544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.40232271091477e-05,
+      "loss": 3.8115,
+      "step": 365056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.401484116163718e-05,
+      "loss": 3.8104,
+      "step": 365568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.400647159293039e-05,
+      "loss": 3.8074,
+      "step": 366080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.399808564541987e-05,
+      "loss": 3.802,
+      "step": 366592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398969969790935e-05,
+      "loss": 3.7933,
+      "step": 367104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398131375039883e-05,
+      "loss": 3.8071,
+      "step": 367616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.397294418169204e-05,
+      "loss": 3.7999,
+      "step": 368128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.396455823418152e-05,
+      "loss": 3.8062,
+      "step": 368640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3956172286671e-05,
+      "loss": 3.8017,
+      "step": 369152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.394778633916048e-05,
+      "loss": 3.8063,
+      "step": 369664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3939416770453686e-05,
+      "loss": 3.8017,
+      "step": 370176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.393103082294317e-05,
+      "loss": 3.8013,
+      "step": 370688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.392264487543265e-05,
+      "loss": 3.7985,
+      "step": 371200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.391425892792213e-05,
+      "loss": 3.8112,
+      "step": 371712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.390588935921534e-05,
+      "loss": 3.8017,
+      "step": 372224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.389750341170482e-05,
+      "loss": 3.7933,
+      "step": 372736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38891174641943e-05,
+      "loss": 3.7833,
+      "step": 373248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388073151668378e-05,
+      "loss": 3.8063,
+      "step": 373760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.387236194797699e-05,
+      "loss": 3.8064,
+      "step": 374272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.386397600046647e-05,
+      "loss": 3.8082,
+      "step": 374784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.385559005295595e-05,
+      "loss": 3.8015,
+      "step": 375296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3847204105445424e-05,
+      "loss": 3.8084,
+      "step": 375808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383883453673864e-05,
+      "loss": 3.8016,
+      "step": 376320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383044858922813e-05,
+      "loss": 3.8076,
+      "step": 376832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.382206264171761e-05,
+      "loss": 3.7973,
+      "step": 377344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.381367669420708e-05,
+      "loss": 3.8002,
+      "step": 377856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3805307125500296e-05,
+      "loss": 3.7995,
+      "step": 378368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3796921177989776e-05,
+      "loss": 3.7933,
+      "step": 378880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3788535230479256e-05,
+      "loss": 3.8081,
+      "step": 379392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.378014928296873e-05,
+      "loss": 3.8047,
+      "step": 379904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3771779714261945e-05,
+      "loss": 3.7992,
+      "step": 380416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3763393766751425e-05,
+      "loss": 3.7965,
+      "step": 380928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.37550078192409e-05,
+      "loss": 3.7989,
+      "step": 381440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.903867244720459,
+      "eval_runtime": 305.6576,
+      "eval_samples_per_second": 1248.426,
+      "eval_steps_per_second": 39.014,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374662187173038e-05,
+      "loss": 3.795,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3738235924219865e-05,
+      "loss": 3.7905,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3729849976709345e-05,
+      "loss": 3.7989,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3721464029198824e-05,
+      "loss": 3.7929,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3713078081688304e-05,
+      "loss": 3.8068,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3704692134177784e-05,
+      "loss": 3.7937,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3696306186667264e-05,
+      "loss": 3.7848,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3687920239156744e-05,
+      "loss": 3.7809,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3679550670449953e-05,
+      "loss": 3.8005,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367116472293943e-05,
+      "loss": 3.7983,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366277877542891e-05,
+      "loss": 3.7878,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365439282791839e-05,
+      "loss": 3.7971,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364605601681907e-05,
+      "loss": 3.7935,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363767006930855e-05,
+      "loss": 3.7913,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362928412179803e-05,
+      "loss": 3.7855,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362089817428751e-05,
+      "loss": 3.7814,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361251222677699e-05,
+      "loss": 3.7861,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360412627926647e-05,
+      "loss": 3.7799,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359574033175595e-05,
+      "loss": 3.7984,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358735438424543e-05,
+      "loss": 3.7999,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357896843673491e-05,
+      "loss": 3.7974,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357058248922439e-05,
+      "loss": 3.7889,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356219654171387e-05,
+      "loss": 3.7858,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355381059420335e-05,
+      "loss": 3.7901,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3545441025496556e-05,
+      "loss": 3.7816,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3537055077986036e-05,
+      "loss": 3.7856,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3528669130475516e-05,
+      "loss": 3.7736,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3520283182964996e-05,
+      "loss": 3.7915,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351191361425821e-05,
+      "loss": 3.7803,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350352766674769e-05,
+      "loss": 3.7739,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349514171923717e-05,
+      "loss": 3.7864,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348675577172665e-05,
+      "loss": 3.7888,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347838620301986e-05,
+      "loss": 3.7939,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347000025550934e-05,
+      "loss": 3.7883,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346161430799882e-05,
+      "loss": 3.7797,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34532283604883e-05,
+      "loss": 3.7911,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344485879178151e-05,
+      "loss": 3.7876,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343647284427099e-05,
+      "loss": 3.7707,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342808689676047e-05,
+      "loss": 3.7707,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341970094924995e-05,
+      "loss": 3.7719,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3411331380543165e-05,
+      "loss": 3.7784,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402945433032645e-05,
+      "loss": 3.7884,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3394559485522125e-05,
+      "loss": 3.781,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3386173538011605e-05,
+      "loss": 3.7745,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377803969304814e-05,
+      "loss": 3.7858,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3369418021794294e-05,
+      "loss": 3.7772,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3361032074283774e-05,
+      "loss": 3.7896,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3352646126773254e-05,
+      "loss": 3.7697,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334427655806646e-05,
+      "loss": 3.7544,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333589061055594e-05,
+      "loss": 3.7915,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332750466304542e-05,
+      "loss": 3.7825,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33191187155349e-05,
+      "loss": 3.7818,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331074914682812e-05,
+      "loss": 3.7765,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33023631993176e-05,
+      "loss": 3.773,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329397725180708e-05,
+      "loss": 3.7643,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328559130429656e-05,
+      "loss": 3.7751,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327722173558977e-05,
+      "loss": 3.7675,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326883578807925e-05,
+      "loss": 3.773,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326044984056873e-05,
+      "loss": 3.7794,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325208027186194e-05,
+      "loss": 3.772,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324369432435142e-05,
+      "loss": 3.7625,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32353083768409e-05,
+      "loss": 3.7836,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322692242933038e-05,
+      "loss": 3.7625,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321855286062359e-05,
+      "loss": 3.763,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321016691311307e-05,
+      "loss": 3.7796,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320178096560255e-05,
+      "loss": 3.78,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319339501809203e-05,
+      "loss": 3.7566,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318502544938524e-05,
+      "loss": 3.7678,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317663950187472e-05,
+      "loss": 3.7541,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31682535543642e-05,
+      "loss": 3.7735,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315986760685368e-05,
+      "loss": 3.7765,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315149803814689e-05,
+      "loss": 3.7732,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314311209063637e-05,
+      "loss": 3.7721,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313472614312585e-05,
+      "loss": 3.7768,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312634019561533e-05,
+      "loss": 3.7855,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3117970626908546e-05,
+      "loss": 3.7621,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109584679398026e-05,
+      "loss": 3.779,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3101198731887506e-05,
+      "loss": 3.7774,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309281278437698e-05,
+      "loss": 3.7571,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3084443215670195e-05,
+      "loss": 3.7772,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3076057268159675e-05,
+      "loss": 3.766,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067671320649155e-05,
+      "loss": 3.7785,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305928537313863e-05,
+      "loss": 3.7749,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3050915804431844e-05,
+      "loss": 3.7602,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042529856921324e-05,
+      "loss": 3.7586,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3034143909410804e-05,
+      "loss": 3.7637,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025757961900284e-05,
+      "loss": 3.7802,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30173883931935e-05,
+      "loss": 3.7646,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300900244568298e-05,
+      "loss": 3.7648,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300061649817245e-05,
+      "loss": 3.7633,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299223055066193e-05,
+      "loss": 3.7696,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298386098195515e-05,
+      "loss": 3.7612,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297547503444463e-05,
+      "loss": 3.7476,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29670890869341e-05,
+      "loss": 3.7648,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295870313942358e-05,
+      "loss": 3.7717,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29503335707168e-05,
+      "loss": 3.7734,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294194762320628e-05,
+      "loss": 3.7541,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293356167569576e-05,
+      "loss": 3.7574,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292517572818524e-05,
+      "loss": 3.7631,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2916806159478454e-05,
+      "loss": 3.7671,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290842021196793e-05,
+      "loss": 3.7738,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290003426445741e-05,
+      "loss": 3.7594,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289164831694689e-05,
+      "loss": 3.7623,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28832787482401e-05,
+      "loss": 3.7671,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2874892800729576e-05,
+      "loss": 3.7505,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2866506853219056e-05,
+      "loss": 3.7609,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2858120905708536e-05,
+      "loss": 3.755,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284975133700175e-05,
+      "loss": 3.7676,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2841365389491225e-05,
+      "loss": 3.7502,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283297944198071e-05,
+      "loss": 3.7673,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282459349447019e-05,
+      "loss": 3.7554,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28162239257634e-05,
+      "loss": 3.7612,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280783797825288e-05,
+      "loss": 3.7587,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279945203074236e-05,
+      "loss": 3.7624,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2791082462035576e-05,
+      "loss": 3.7603,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278269651452505e-05,
+      "loss": 3.7729,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277431056701453e-05,
+      "loss": 3.7653,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276592461950401e-05,
+      "loss": 3.7673,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275753867199349e-05,
+      "loss": 3.7593,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274915272448297e-05,
+      "loss": 3.7621,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274076677697245e-05,
+      "loss": 3.7491,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2732397208265665e-05,
+      "loss": 3.7567,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2724011260755145e-05,
+      "loss": 3.7569,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715625313244625e-05,
+      "loss": 3.7617,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2707239365734105e-05,
+      "loss": 3.7614,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2698869797027314e-05,
+      "loss": 3.7593,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2690483849516794e-05,
+      "loss": 3.7577,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2682097902006274e-05,
+      "loss": 3.7591,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2673711954495754e-05,
+      "loss": 3.756,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266534238578896e-05,
+      "loss": 3.7633,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265695643827844e-05,
+      "loss": 3.7568,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264857049076792e-05,
+      "loss": 3.7509,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26401845432574e-05,
+      "loss": 3.7373,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263181497455062e-05,
+      "loss": 3.765,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26234290270401e-05,
+      "loss": 3.7602,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261504307952958e-05,
+      "loss": 3.7644,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260665713201906e-05,
+      "loss": 3.7573,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.259828756331227e-05,
+      "loss": 3.7621,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258990161580175e-05,
+      "loss": 3.7545,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258151566829123e-05,
+      "loss": 3.7686,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257312972078071e-05,
+      "loss": 3.7497,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256476015207392e-05,
+      "loss": 3.756,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25563742045634e-05,
+      "loss": 3.756,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254798825705288e-05,
+      "loss": 3.7499,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253960230954236e-05,
+      "loss": 3.766,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253123274083557e-05,
+      "loss": 3.7573,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252284679332505e-05,
+      "loss": 3.762,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251446084581453e-05,
+      "loss": 3.7486,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250607489830401e-05,
+      "loss": 3.7564,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8884012699127197,
+      "eval_runtime": 306.1308,
+      "eval_samples_per_second": 1246.497,
+      "eval_steps_per_second": 38.954,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.249770532959722e-05,
+      "loss": 3.7517,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.24893193820867e-05,
+      "loss": 3.7488,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248093343457618e-05,
+      "loss": 3.7541,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.247254748706566e-05,
+      "loss": 3.7512,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.246417791835887e-05,
+      "loss": 3.7645,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245579197084835e-05,
+      "loss": 3.7511,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244740602333783e-05,
+      "loss": 3.7446,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243902007582731e-05,
+      "loss": 3.7368,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243065050712052e-05,
+      "loss": 3.7578,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2422264559610006e-05,
+      "loss": 3.7543,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2413878612099486e-05,
+      "loss": 3.7469,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2405492664588966e-05,
+      "loss": 3.7544,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2397139474685904e-05,
+      "loss": 3.7511,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2388753527175384e-05,
+      "loss": 3.7494,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2380367579664864e-05,
+      "loss": 3.7413,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2371981632154344e-05,
+      "loss": 3.7411,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2363595684643824e-05,
+      "loss": 3.7434,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2355209737133304e-05,
+      "loss": 3.7409,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2346823789622784e-05,
+      "loss": 3.7521,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2338437842112264e-05,
+      "loss": 3.757,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2330051894601744e-05,
+      "loss": 3.7547,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2321665947091224e-05,
+      "loss": 3.748,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2313279999580704e-05,
+      "loss": 3.7442,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2304894052070184e-05,
+      "loss": 3.7464,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.22965244833634e-05,
+      "loss": 3.7434,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.228813853585288e-05,
+      "loss": 3.7461,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227975258834236e-05,
+      "loss": 3.7293,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227136664083183e-05,
+      "loss": 3.7522,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.226299707212505e-05,
+      "loss": 3.741,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.225461112461453e-05,
+      "loss": 3.7325,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2246225177104e-05,
+      "loss": 3.7428,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.223783922959348e-05,
+      "loss": 3.7511,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.22294696608867e-05,
+      "loss": 3.7475,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222108371337618e-05,
+      "loss": 3.7491,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.221269776586566e-05,
+      "loss": 3.7386,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.220431181835514e-05,
+      "loss": 3.7519,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2195942249648353e-05,
+      "loss": 3.7441,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2187556302137833e-05,
+      "loss": 3.731,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217917035462731e-05,
+      "loss": 3.732,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217080078592052e-05,
+      "loss": 3.7294,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.216241483841e-05,
+      "loss": 3.739,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2154028890899476e-05,
+      "loss": 3.7488,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2145642943388956e-05,
+      "loss": 3.7374,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.213727337468217e-05,
+      "loss": 3.7364,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212888742717165e-05,
+      "loss": 3.7425,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212050147966113e-05,
+      "loss": 3.7355,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.211211553215061e-05,
+      "loss": 3.7487,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.210374596344383e-05,
+      "loss": 3.7323,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.209536001593331e-05,
+      "loss": 3.7092,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.208697406842278e-05,
+      "loss": 3.7513,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.207858812091226e-05,
+      "loss": 3.7434,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2070218552205476e-05,
+      "loss": 3.742,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206183260469495e-05,
+      "loss": 3.7349,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.205344665718443e-05,
+      "loss": 3.7338,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.204506070967391e-05,
+      "loss": 3.7243,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2036691140967125e-05,
+      "loss": 3.7374,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2028305193456605e-05,
+      "loss": 3.7304,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2019919245946085e-05,
+      "loss": 3.7308,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2011533298435565e-05,
+      "loss": 3.7395,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.200316372972878e-05,
+      "loss": 3.734,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1994777782218254e-05,
+      "loss": 3.7215,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1986391834707734e-05,
+      "loss": 3.7427,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1978005887197214e-05,
+      "loss": 3.7226,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196963631849042e-05,
+      "loss": 3.7285,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19612503709799e-05,
+      "loss": 3.7395,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195286442346938e-05,
+      "loss": 3.7392,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.194447847595886e-05,
+      "loss": 3.7205,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.193610890725207e-05,
+      "loss": 3.7258,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.192772295974156e-05,
+      "loss": 3.7154,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191933701223104e-05,
+      "loss": 3.7356,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191096744352425e-05,
+      "loss": 3.7365,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.190258149601373e-05,
+      "loss": 3.7345,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.189419554850321e-05,
+      "loss": 3.7336,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.188580960099269e-05,
+      "loss": 3.7348,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.187742365348217e-05,
+      "loss": 3.7477,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186903770597165e-05,
+      "loss": 3.7258,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186065175846113e-05,
+      "loss": 3.7368,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.185226581095061e-05,
+      "loss": 3.7377,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1843896242243817e-05,
+      "loss": 3.7202,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1835510294733297e-05,
+      "loss": 3.7352,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1827124347222776e-05,
+      "loss": 3.7338,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1818738399712256e-05,
+      "loss": 3.7332,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181036883100547e-05,
+      "loss": 3.7374,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180198288349495e-05,
+      "loss": 3.722,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.179359693598443e-05,
+      "loss": 3.7215,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.178521098847391e-05,
+      "loss": 3.7263,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.177684141976712e-05,
+      "loss": 3.7366,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.17684554722566e-05,
+      "loss": 3.7287,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.176006952474608e-05,
+      "loss": 3.7247,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.175168357723556e-05,
+      "loss": 3.7277,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.174331400852877e-05,
+      "loss": 3.7332,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.173492806101825e-05,
+      "loss": 3.7233,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.172654211350773e-05,
+      "loss": 3.7088,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.171815616599721e-05,
+      "loss": 3.7271,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1709786597290426e-05,
+      "loss": 3.7303,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1701400649779906e-05,
+      "loss": 3.7356,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1693014702269386e-05,
+      "loss": 3.7168,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1684628754758866e-05,
+      "loss": 3.7208,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1676259186052075e-05,
+      "loss": 3.7255,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1667873238541555e-05,
+      "loss": 3.7277,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1659487291031035e-05,
+      "loss": 3.7375,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1651101343520515e-05,
+      "loss": 3.7234,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1642731774813724e-05,
+      "loss": 3.7255,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1634345827303204e-05,
+      "loss": 3.7295,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1625959879792684e-05,
+      "loss": 3.7127,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1617573932282164e-05,
+      "loss": 3.7246,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160920436357538e-05,
+      "loss": 3.7147,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160081841606486e-05,
+      "loss": 3.7343,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.159243246855434e-05,
+      "loss": 3.7127,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.158404652104382e-05,
+      "loss": 3.7285,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.157567695233703e-05,
+      "loss": 3.7198,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.156729100482651e-05,
+      "loss": 3.7215,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155890505731599e-05,
+      "loss": 3.7226,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155051910980547e-05,
+      "loss": 3.7261,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154214954109868e-05,
+      "loss": 3.7233,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.153376359358816e-05,
+      "loss": 3.7338,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.152537764607764e-05,
+      "loss": 3.7224,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151700807737085e-05,
+      "loss": 3.7363,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150862212986033e-05,
+      "loss": 3.7227,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150023618234981e-05,
+      "loss": 3.7215,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.149185023483929e-05,
+      "loss": 3.7188,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14834806661325e-05,
+      "loss": 3.7182,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.147509471862198e-05,
+      "loss": 3.7175,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.146670877111146e-05,
+      "loss": 3.7258,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.145832282360094e-05,
+      "loss": 3.7207,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144995325489415e-05,
+      "loss": 3.7257,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144156730738363e-05,
+      "loss": 3.7206,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.143318135987311e-05,
+      "loss": 3.7234,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.142479541236259e-05,
+      "loss": 3.7227,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.141642584365581e-05,
+      "loss": 3.7214,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.140803989614529e-05,
+      "loss": 3.7176,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139965394863477e-05,
+      "loss": 3.7185,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139126800112425e-05,
+      "loss": 3.7003,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1382898432417456e-05,
+      "loss": 3.7281,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1374512484906936e-05,
+      "loss": 3.724,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1366126537396416e-05,
+      "loss": 3.7312,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1357740589885896e-05,
+      "loss": 3.7169,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1349371021179105e-05,
+      "loss": 3.7278,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1340985073668585e-05,
+      "loss": 3.7233,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1332599126158065e-05,
+      "loss": 3.7323,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1324213178647545e-05,
+      "loss": 3.7127,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.131584360994076e-05,
+      "loss": 3.716,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.130745766243024e-05,
+      "loss": 3.7222,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129907171491972e-05,
+      "loss": 3.7176,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.12906857674092e-05,
+      "loss": 3.7251,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128231619870241e-05,
+      "loss": 3.7233,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.127393025119189e-05,
+      "loss": 3.7247,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.126554430368137e-05,
+      "loss": 3.7125,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.125715835617085e-05,
+      "loss": 3.7234,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.878814697265625,
+      "eval_runtime": 306.2043,
+      "eval_samples_per_second": 1246.198,
+      "eval_steps_per_second": 38.945,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124877240866032e-05,
+      "loss": 3.7192,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.12403864611498e-05,
+      "loss": 3.7182,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123200051363928e-05,
+      "loss": 3.7139,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.122361456612876e-05,
+      "loss": 3.716,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.121524499742198e-05,
+      "loss": 3.7274,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.120685904991146e-05,
+      "loss": 3.7176,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119847310240094e-05,
+      "loss": 3.7099,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119008715489042e-05,
+      "loss": 3.7064,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11817012073799e-05,
+      "loss": 3.7172,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117331525986938e-05,
+      "loss": 3.7212,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.116492931235886e-05,
+      "loss": 3.7085,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115654336484834e-05,
+      "loss": 3.7167,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1148190174945276e-05,
+      "loss": 3.7215,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1139804227434756e-05,
+      "loss": 3.7107,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1131418279924236e-05,
+      "loss": 3.7094,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1123032332413716e-05,
+      "loss": 3.7061,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1114646384903196e-05,
+      "loss": 3.7088,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110626043739268e-05,
+      "loss": 3.7055,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109787448988216e-05,
+      "loss": 3.7136,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108948854237164e-05,
+      "loss": 3.7256,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108110259486112e-05,
+      "loss": 3.7208,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10727166473506e-05,
+      "loss": 3.7145,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106433069984008e-05,
+      "loss": 3.7082,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.105594475232956e-05,
+      "loss": 3.7096,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.104757518362277e-05,
+      "loss": 3.709,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103918923611225e-05,
+      "loss": 3.7102,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103080328860173e-05,
+      "loss": 3.6971,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.102241734109121e-05,
+      "loss": 3.7181,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101404777238442e-05,
+      "loss": 3.7074,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10056618248739e-05,
+      "loss": 3.7004,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099727587736338e-05,
+      "loss": 3.7023,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0988906308656597e-05,
+      "loss": 3.7181,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0980520361146076e-05,
+      "loss": 3.7137,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0972134413635556e-05,
+      "loss": 3.7148,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0963748466125036e-05,
+      "loss": 3.7046,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0955362518614516e-05,
+      "loss": 3.7127,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.094697657110399e-05,
+      "loss": 3.7153,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0938607002397205e-05,
+      "loss": 3.696,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0930221054886685e-05,
+      "loss": 3.6967,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092183510737616e-05,
+      "loss": 3.6946,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.091344915986564e-05,
+      "loss": 3.7053,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.090506321235512e-05,
+      "loss": 3.7114,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08966772648446e-05,
+      "loss": 3.7081,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0888291317334085e-05,
+      "loss": 3.7008,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0879905369823565e-05,
+      "loss": 3.7086,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0871535801116774e-05,
+      "loss": 3.7026,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0863149853606254e-05,
+      "loss": 3.7113,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0854763906095734e-05,
+      "loss": 3.704,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0846377958585214e-05,
+      "loss": 3.6763,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083800838987842e-05,
+      "loss": 3.7138,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08296224423679e-05,
+      "loss": 3.7078,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082123649485738e-05,
+      "loss": 3.7117,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081285054734686e-05,
+      "loss": 3.7015,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080448097864007e-05,
+      "loss": 3.6965,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079609503112955e-05,
+      "loss": 3.6948,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078770908361904e-05,
+      "loss": 3.7029,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077932313610852e-05,
+      "loss": 3.7003,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077095356740173e-05,
+      "loss": 3.6929,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076256761989121e-05,
+      "loss": 3.7048,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075418167238069e-05,
+      "loss": 3.7037,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.074579572487017e-05,
+      "loss": 3.6854,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.073742615616338e-05,
+      "loss": 3.7091,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072904020865286e-05,
+      "loss": 3.6941,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072065426114234e-05,
+      "loss": 3.6905,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071226831363182e-05,
+      "loss": 3.7052,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0703898744925026e-05,
+      "loss": 3.7074,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0695512797414506e-05,
+      "loss": 3.6868,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068712684990399e-05,
+      "loss": 3.6942,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.06787572811972e-05,
+      "loss": 3.6816,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067037133368668e-05,
+      "loss": 3.7005,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066200176497989e-05,
+      "loss": 3.7041,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065361581746937e-05,
+      "loss": 3.7027,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064522986995885e-05,
+      "loss": 3.6983,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063684392244833e-05,
+      "loss": 3.7022,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062845797493781e-05,
+      "loss": 3.7165,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062007202742729e-05,
+      "loss": 3.6895,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061168607991677e-05,
+      "loss": 3.7059,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.060330013240625e-05,
+      "loss": 3.7055,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.059493056369946e-05,
+      "loss": 3.6877,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0586544616188946e-05,
+      "loss": 3.7028,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0578158668678426e-05,
+      "loss": 3.7018,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0569772721167906e-05,
+      "loss": 3.7006,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0561403152461115e-05,
+      "loss": 3.7083,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0553017204950595e-05,
+      "loss": 3.6853,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0544631257440075e-05,
+      "loss": 3.6906,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0536245309929555e-05,
+      "loss": 3.6932,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0527875741222764e-05,
+      "loss": 3.7045,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0519489793712244e-05,
+      "loss": 3.6938,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0511103846201724e-05,
+      "loss": 3.6966,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0502717898691204e-05,
+      "loss": 3.6988,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049434832998441e-05,
+      "loss": 3.694,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04859623824739e-05,
+      "loss": 3.6933,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.047757643496338e-05,
+      "loss": 3.6777,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046919048745286e-05,
+      "loss": 3.6927,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046082091874607e-05,
+      "loss": 3.6986,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.045243497123555e-05,
+      "loss": 3.705,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044404902372503e-05,
+      "loss": 3.6852,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.043566307621451e-05,
+      "loss": 3.6866,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.042729350750772e-05,
+      "loss": 3.6938,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04189075599972e-05,
+      "loss": 3.6952,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041052161248668e-05,
+      "loss": 3.7002,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040213566497616e-05,
+      "loss": 3.6967,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039376609626937e-05,
+      "loss": 3.6933,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0385380148758853e-05,
+      "loss": 3.6938,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0376994201248333e-05,
+      "loss": 3.6872,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036862463254154e-05,
+      "loss": 3.6891,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036023868503102e-05,
+      "loss": 3.684,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.03518527375205e-05,
+      "loss": 3.7033,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034346679000998e-05,
+      "loss": 3.6823,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033508084249946e-05,
+      "loss": 3.6947,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032669489498894e-05,
+      "loss": 3.688,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0318308947478415e-05,
+      "loss": 3.6941,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0309922999967895e-05,
+      "loss": 3.6889,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030155343126111e-05,
+      "loss": 3.6942,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.029318386255432e-05,
+      "loss": 3.6899,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028479791504381e-05,
+      "loss": 3.7053,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027641196753329e-05,
+      "loss": 3.6907,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026802602002277e-05,
+      "loss": 3.7044,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0259656451315976e-05,
+      "loss": 3.6898,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0251270503805456e-05,
+      "loss": 3.691,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0242884556294936e-05,
+      "loss": 3.6872,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0234498608784416e-05,
+      "loss": 3.6914,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.022611266127389e-05,
+      "loss": 3.6839,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.021772671376337e-05,
+      "loss": 3.6945,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020934076625285e-05,
+      "loss": 3.685,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020095481874233e-05,
+      "loss": 3.693,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0192585250035545e-05,
+      "loss": 3.6931,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0184199302525025e-05,
+      "loss": 3.6877,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0175813355014505e-05,
+      "loss": 3.6923,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0167427407503985e-05,
+      "loss": 3.6957,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0159057838797194e-05,
+      "loss": 3.683,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0150671891286674e-05,
+      "loss": 3.6885,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0142285943776154e-05,
+      "loss": 3.6681,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0133899996265634e-05,
+      "loss": 3.6978,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.012553042755884e-05,
+      "loss": 3.6929,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011714448004832e-05,
+      "loss": 3.6969,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01087585325378e-05,
+      "loss": 3.6876,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010037258502728e-05,
+      "loss": 3.694,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009200301632049e-05,
+      "loss": 3.6922,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008361706880998e-05,
+      "loss": 3.6993,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007523112129946e-05,
+      "loss": 3.682,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006684517378894e-05,
+      "loss": 3.6888,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005847560508215e-05,
+      "loss": 3.69,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005008965757163e-05,
+      "loss": 3.6923,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004170371006111e-05,
+      "loss": 3.6921,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.003331776255059e-05,
+      "loss": 3.6892,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0024948193843797e-05,
+      "loss": 3.6961,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0016562246333276e-05,
+      "loss": 3.6802,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0008176298822756e-05,
+      "loss": 3.6968,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.871347188949585,
+      "eval_runtime": 305.8402,
+      "eval_samples_per_second": 1247.681,
+      "eval_steps_per_second": 38.991,
+      "step": 610560
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 4.078348705585889e+17,
+  "trial_name": null,
+  "trial_params": null
+}