diff --git "a/checkpoint-2747520/trainer_state.json" "b/checkpoint-2747520/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2747520/trainer_state.json"
@@ -0,0 +1,32509 @@
+{
+  "best_metric": 3.863309860229492,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/superlative-quantifier/transformer/2/checkpoints/checkpoint-839520",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 2747520,
+  "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
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9999790351312236e-05,
+      "loss": 3.6801,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9991420782605445e-05,
+      "loss": 3.6841,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.998303483509493e-05,
+      "loss": 3.6876,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.997464888758441e-05,
+      "loss": 3.6863,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.996627931887762e-05,
+      "loss": 3.6977,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.99578933713671e-05,
+      "loss": 3.6876,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994950742385658e-05,
+      "loss": 3.6785,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994112147634606e-05,
+      "loss": 3.6762,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.993273552883554e-05,
+      "loss": 3.6866,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.992434958132502e-05,
+      "loss": 3.6914,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.99159636338145e-05,
+      "loss": 3.6838,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.990757768630398e-05,
+      "loss": 3.6853,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989922449640092e-05,
+      "loss": 3.6913,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98908385488904e-05,
+      "loss": 3.6798,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9882452601379886e-05,
+      "loss": 3.6791,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9874066653869366e-05,
+      "loss": 3.6756,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9865680706358846e-05,
+      "loss": 3.6813,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9857294758848326e-05,
+      "loss": 3.676,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9848908811337806e-05,
+      "loss": 3.6839,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9840522863827286e-05,
+      "loss": 3.6941,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9832136916316766e-05,
+      "loss": 3.6914,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9823750968806246e-05,
+      "loss": 3.6812,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9815365021295726e-05,
+      "loss": 3.6799,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98069790737852e-05,
+      "loss": 3.6812,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9798609505078415e-05,
+      "loss": 3.6774,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9790223557567895e-05,
+      "loss": 3.6809,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.978183761005737e-05,
+      "loss": 3.6693,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9773451662546854e-05,
+      "loss": 3.6889,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.976508209384007e-05,
+      "loss": 3.6793,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.975669614632955e-05,
+      "loss": 3.6667,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9748310198819023e-05,
+      "loss": 3.6739,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973994063011224e-05,
+      "loss": 3.6897,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973155468260172e-05,
+      "loss": 3.6878,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.97231687350912e-05,
+      "loss": 3.6842,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.971478278758067e-05,
+      "loss": 3.6775,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.970639684007015e-05,
+      "loss": 3.6778,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.969801089255963e-05,
+      "loss": 3.6902,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968962494504911e-05,
+      "loss": 3.6694,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968123899753859e-05,
+      "loss": 3.6655,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.967286942883181e-05,
+      "loss": 3.6686,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.966448348132129e-05,
+      "loss": 3.6719,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.965609753381077e-05,
+      "loss": 3.6825,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.964771158630025e-05,
+      "loss": 3.6729,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963934201759346e-05,
+      "loss": 3.6802,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963095607008294e-05,
+      "loss": 3.674,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.962257012257242e-05,
+      "loss": 3.6716,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96141841750619e-05,
+      "loss": 3.6865,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9605814606355106e-05,
+      "loss": 3.6732,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9597428658844586e-05,
+      "loss": 3.6513,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9589042711334066e-05,
+      "loss": 3.6851,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9580656763823546e-05,
+      "loss": 3.6775,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957228719511676e-05,
+      "loss": 3.6839,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.956390124760624e-05,
+      "loss": 3.6739,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.955551530009572e-05,
+      "loss": 3.6666,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.954714573138893e-05,
+      "loss": 3.6649,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953875978387841e-05,
+      "loss": 3.6718,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953037383636789e-05,
+      "loss": 3.6757,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952198788885737e-05,
+      "loss": 3.6608,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.951361832015058e-05,
+      "loss": 3.6803,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.950523237264006e-05,
+      "loss": 3.6724,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.949684642512954e-05,
+      "loss": 3.6611,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948846047761902e-05,
+      "loss": 3.6792,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948009090891223e-05,
+      "loss": 3.6627,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947170496140171e-05,
+      "loss": 3.6625,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9463319013891195e-05,
+      "loss": 3.6752,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9454933066380675e-05,
+      "loss": 3.6806,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9446563497673884e-05,
+      "loss": 3.6521,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9438177550163364e-05,
+      "loss": 3.6713,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9429791602652844e-05,
+      "loss": 3.6552,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9421422033946053e-05,
+      "loss": 3.6715,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9413036086435533e-05,
+      "loss": 3.6782,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.940465013892501e-05,
+      "loss": 3.6685,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.939626419141449e-05,
+      "loss": 3.6745,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.938787824390397e-05,
+      "loss": 3.6703,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937949229639345e-05,
+      "loss": 3.6903,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937110634888293e-05,
+      "loss": 3.6604,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.936272040137241e-05,
+      "loss": 3.6776,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.935435083266563e-05,
+      "loss": 3.6736,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.934596488515511e-05,
+      "loss": 3.6654,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.933757893764459e-05,
+      "loss": 3.6732,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932919299013407e-05,
+      "loss": 3.6709,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932080704262355e-05,
+      "loss": 3.6719,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.931242109511303e-05,
+      "loss": 3.6805,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.93040351476025e-05,
+      "loss": 3.6559,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.929564920009198e-05,
+      "loss": 3.6625,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.92872796313852e-05,
+      "loss": 3.6652,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.927891006267841e-05,
+      "loss": 3.6774,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.927052411516789e-05,
+      "loss": 3.6642,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926213816765737e-05,
+      "loss": 3.667,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.925375222014685e-05,
+      "loss": 3.672,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9245366272636333e-05,
+      "loss": 3.666,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.923698032512581e-05,
+      "loss": 3.6635,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.922861075641902e-05,
+      "loss": 3.65,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.92202248089085e-05,
+      "loss": 3.6645,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9211838861397976e-05,
+      "loss": 3.6694,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9203452913887456e-05,
+      "loss": 3.6765,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9195066966376936e-05,
+      "loss": 3.6592,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9186681018866416e-05,
+      "loss": 3.6599,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9178295071355896e-05,
+      "loss": 3.6632,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9169909123845375e-05,
+      "loss": 3.6712,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9161539555138585e-05,
+      "loss": 3.6731,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.915315360762807e-05,
+      "loss": 3.6667,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.914476766011755e-05,
+      "loss": 3.6688,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.913638171260703e-05,
+      "loss": 3.6632,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.912801214390024e-05,
+      "loss": 3.6607,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911964257519345e-05,
+      "loss": 3.6578,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911125662768293e-05,
+      "loss": 3.6598,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.910287068017241e-05,
+      "loss": 3.6766,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.909448473266189e-05,
+      "loss": 3.6544,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.908609878515137e-05,
+      "loss": 3.6636,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.907771283764085e-05,
+      "loss": 3.6565,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906932689013033e-05,
+      "loss": 3.6715,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906095732142354e-05,
+      "loss": 3.6638,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9052571373913025e-05,
+      "loss": 3.6654,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9044185426402505e-05,
+      "loss": 3.6633,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9035799478891985e-05,
+      "loss": 3.6755,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9027429910185194e-05,
+      "loss": 3.6661,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9019043962674674e-05,
+      "loss": 3.6724,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9010658015164154e-05,
+      "loss": 3.664,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9002272067653634e-05,
+      "loss": 3.6635,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8993886120143114e-05,
+      "loss": 3.6607,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8985500172632594e-05,
+      "loss": 3.6628,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8977114225122074e-05,
+      "loss": 3.6574,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8968728277611554e-05,
+      "loss": 3.664,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.896035870890476e-05,
+      "loss": 3.6605,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.895197276139424e-05,
+      "loss": 3.6675,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.894358681388372e-05,
+      "loss": 3.6649,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.893521724517694e-05,
+      "loss": 3.6652,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.892683129766642e-05,
+      "loss": 3.6617,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.89184453501559e-05,
+      "loss": 3.6701,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.891005940264538e-05,
+      "loss": 3.6546,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.890167345513486e-05,
+      "loss": 3.6646,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.889328750762434e-05,
+      "loss": 3.6383,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.888490156011381e-05,
+      "loss": 3.6706,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.887653199140703e-05,
+      "loss": 3.6649,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.886814604389651e-05,
+      "loss": 3.6738,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885976009638599e-05,
+      "loss": 3.6557,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885137414887546e-05,
+      "loss": 3.6694,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8843004580168676e-05,
+      "loss": 3.6636,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8834618632658156e-05,
+      "loss": 3.6749,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.882623268514764e-05,
+      "loss": 3.6545,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8817846737637116e-05,
+      "loss": 3.6599,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880947716893033e-05,
+      "loss": 3.6611,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880109122141981e-05,
+      "loss": 3.6705,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8792705273909285e-05,
+      "loss": 3.6603,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8784319326398765e-05,
+      "loss": 3.6608,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.877594975769198e-05,
+      "loss": 3.6687,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.876756381018146e-05,
+      "loss": 3.6577,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8759177862670934e-05,
+      "loss": 3.6681,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8674252033233643,
+      "eval_runtime": 305.8831,
+      "eval_samples_per_second": 1247.506,
+      "eval_steps_per_second": 38.985,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8750791915160414e-05,
+      "loss": 3.654,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8742405967649894e-05,
+      "loss": 3.6613,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.873402002013938e-05,
+      "loss": 3.6558,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.872563407262886e-05,
+      "loss": 3.6634,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.871726450392207e-05,
+      "loss": 3.6746,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870887855641155e-05,
+      "loss": 3.6581,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870049260890103e-05,
+      "loss": 3.6532,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869210666139051e-05,
+      "loss": 3.6507,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.868372071387999e-05,
+      "loss": 3.6624,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.867533476636947e-05,
+      "loss": 3.6645,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.866694881885895e-05,
+      "loss": 3.6541,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865856287134843e-05,
+      "loss": 3.6583,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865020968144537e-05,
+      "loss": 3.6663,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.864182373393485e-05,
+      "loss": 3.653,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8633437786424334e-05,
+      "loss": 3.6527,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8625051838913814e-05,
+      "loss": 3.647,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8616665891403294e-05,
+      "loss": 3.6572,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8608279943892774e-05,
+      "loss": 3.6506,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8599893996382254e-05,
+      "loss": 3.6515,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8591508048871734e-05,
+      "loss": 3.6676,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8583122101361214e-05,
+      "loss": 3.6659,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8574736153850694e-05,
+      "loss": 3.6564,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8566350206340174e-05,
+      "loss": 3.6539,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.855796425882965e-05,
+      "loss": 3.6547,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854959469012286e-05,
+      "loss": 3.6556,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854120874261234e-05,
+      "loss": 3.6537,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.853282279510182e-05,
+      "loss": 3.6417,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.85244368475913e-05,
+      "loss": 3.6602,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.851606727888452e-05,
+      "loss": 3.6568,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8507681331374e-05,
+      "loss": 3.6411,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849929538386347e-05,
+      "loss": 3.6493,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849092581515669e-05,
+      "loss": 3.6581,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.848253986764617e-05,
+      "loss": 3.6586,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.847415392013565e-05,
+      "loss": 3.6627,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.846576797262512e-05,
+      "loss": 3.6557,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84573820251146e-05,
+      "loss": 3.6486,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.844899607760408e-05,
+      "loss": 3.6669,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.844061013009356e-05,
+      "loss": 3.6424,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843222418258304e-05,
+      "loss": 3.6415,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842385461387626e-05,
+      "loss": 3.6428,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.841546866636574e-05,
+      "loss": 3.6462,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.840708271885522e-05,
+      "loss": 3.6544,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8398696771344697e-05,
+      "loss": 3.649,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8390327202637906e-05,
+      "loss": 3.6527,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8381941255127386e-05,
+      "loss": 3.6494,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8373555307616866e-05,
+      "loss": 3.6469,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8365169360106346e-05,
+      "loss": 3.6582,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8356799791399555e-05,
+      "loss": 3.6528,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8348413843889035e-05,
+      "loss": 3.6275,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8340027896378515e-05,
+      "loss": 3.6568,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8331641948867994e-05,
+      "loss": 3.6542,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.832327238016121e-05,
+      "loss": 3.6596,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.831488643265069e-05,
+      "loss": 3.6466,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.830650048514017e-05,
+      "loss": 3.6431,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.829811453762965e-05,
+      "loss": 3.6356,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828974496892286e-05,
+      "loss": 3.6474,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828135902141234e-05,
+      "loss": 3.649,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.827297307390182e-05,
+      "loss": 3.6388,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.826460350519503e-05,
+      "loss": 3.6514,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.825621755768451e-05,
+      "loss": 3.6482,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.824783161017399e-05,
+      "loss": 3.6393,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823944566266347e-05,
+      "loss": 3.6489,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823107609395668e-05,
+      "loss": 3.6388,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.822269014644616e-05,
+      "loss": 3.6397,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821432057773937e-05,
+      "loss": 3.6444,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.820593463022885e-05,
+      "loss": 3.6557,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.819754868271833e-05,
+      "loss": 3.633,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818916273520781e-05,
+      "loss": 3.6453,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818077678769729e-05,
+      "loss": 3.6253,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.81724072189905e-05,
+      "loss": 3.6477,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.816402127147998e-05,
+      "loss": 3.6519,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.815563532396946e-05,
+      "loss": 3.6476,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.814724937645894e-05,
+      "loss": 3.6457,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813886342894842e-05,
+      "loss": 3.6473,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.81304774814379e-05,
+      "loss": 3.6634,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812209153392738e-05,
+      "loss": 3.6382,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.811370558641686e-05,
+      "loss": 3.6541,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810533601771008e-05,
+      "loss": 3.6428,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.809695007019956e-05,
+      "loss": 3.6426,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.808856412268904e-05,
+      "loss": 3.6476,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.808017817517852e-05,
+      "loss": 3.6437,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8071792227668e-05,
+      "loss": 3.6494,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8063422658961207e-05,
+      "loss": 3.6577,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8055036711450686e-05,
+      "loss": 3.6299,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8046650763940166e-05,
+      "loss": 3.64,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8038264816429646e-05,
+      "loss": 3.6382,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8029895247722855e-05,
+      "loss": 3.6541,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8021509300212335e-05,
+      "loss": 3.6379,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8013123352701815e-05,
+      "loss": 3.6465,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8004737405191295e-05,
+      "loss": 3.6461,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799635145768078e-05,
+      "loss": 3.6428,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7987965510170255e-05,
+      "loss": 3.6375,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7979579562659735e-05,
+      "loss": 3.6287,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7971193615149215e-05,
+      "loss": 3.6374,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796282404644243e-05,
+      "loss": 3.647,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7954438098931904e-05,
+      "loss": 3.6542,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7946052151421384e-05,
+      "loss": 3.6313,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7937666203910864e-05,
+      "loss": 3.638,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792929663520407e-05,
+      "loss": 3.6392,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792091068769355e-05,
+      "loss": 3.6445,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.791252474018303e-05,
+      "loss": 3.6488,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.790413879267252e-05,
+      "loss": 3.6433,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789576922396573e-05,
+      "loss": 3.6437,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.788738327645521e-05,
+      "loss": 3.6401,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787899732894469e-05,
+      "loss": 3.6414,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7870627760237905e-05,
+      "loss": 3.6321,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786224181272738e-05,
+      "loss": 3.6374,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.785385586521686e-05,
+      "loss": 3.6485,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784546991770634e-05,
+      "loss": 3.6351,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.783708397019582e-05,
+      "loss": 3.6378,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78286980226853e-05,
+      "loss": 3.6336,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782031207517478e-05,
+      "loss": 3.6433,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781192612766426e-05,
+      "loss": 3.6372,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7803556558957474e-05,
+      "loss": 3.6418,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.779518699025068e-05,
+      "loss": 3.6411,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778680104274016e-05,
+      "loss": 3.653,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777841509522964e-05,
+      "loss": 3.6391,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777002914771912e-05,
+      "loss": 3.6449,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776165957901233e-05,
+      "loss": 3.6454,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.775327363150181e-05,
+      "loss": 3.6355,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.774488768399129e-05,
+      "loss": 3.6419,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.773650173648077e-05,
+      "loss": 3.6334,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.772811578897025e-05,
+      "loss": 3.6336,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771972984145973e-05,
+      "loss": 3.6409,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771134389394921e-05,
+      "loss": 3.6381,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.770295794643869e-05,
+      "loss": 3.6434,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.769458837773191e-05,
+      "loss": 3.6396,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7686218809025116e-05,
+      "loss": 3.6445,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7677832861514596e-05,
+      "loss": 3.6355,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7669446914004076e-05,
+      "loss": 3.6475,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7661060966493556e-05,
+      "loss": 3.6325,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7652675018983036e-05,
+      "loss": 3.6384,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7644289071472516e-05,
+      "loss": 3.6133,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7635903123961996e-05,
+      "loss": 3.6469,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7627517176451476e-05,
+      "loss": 3.6396,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7619163986548414e-05,
+      "loss": 3.652,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7610778039037894e-05,
+      "loss": 3.6322,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760239209152738e-05,
+      "loss": 3.6435,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.759400614401686e-05,
+      "loss": 3.6364,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758562019650634e-05,
+      "loss": 3.6552,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757725062779955e-05,
+      "loss": 3.6335,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756886468028903e-05,
+      "loss": 3.6347,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756047873277851e-05,
+      "loss": 3.639,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755209278526799e-05,
+      "loss": 3.6445,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754370683775747e-05,
+      "loss": 3.6362,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753532089024695e-05,
+      "loss": 3.6362,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.752695132154016e-05,
+      "loss": 3.6423,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751856537402964e-05,
+      "loss": 3.639,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751017942651912e-05,
+      "loss": 3.6431,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8649113178253174,
+      "eval_runtime": 311.1998,
+      "eval_samples_per_second": 1226.193,
+      "eval_steps_per_second": 38.319,
+      "step": 763200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.75017934790086e-05,
+      "loss": 3.6258,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.749340753149808e-05,
+      "loss": 3.6355,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.748502158398756e-05,
+      "loss": 3.6357,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7476652015280774e-05,
+      "loss": 3.637,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7468266067770254e-05,
+      "loss": 3.6495,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7459880120259734e-05,
+      "loss": 3.6328,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.745149417274921e-05,
+      "loss": 3.634,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.744310822523869e-05,
+      "loss": 3.6282,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.74347386565319e-05,
+      "loss": 3.6336,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.742635270902138e-05,
+      "loss": 3.6399,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7417966761510856e-05,
+      "loss": 3.636,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7409580814000336e-05,
+      "loss": 3.6352,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740122762409728e-05,
+      "loss": 3.6438,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.739284167658677e-05,
+      "loss": 3.6308,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.738445572907625e-05,
+      "loss": 3.6286,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.737606978156573e-05,
+      "loss": 3.6207,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.736768383405521e-05,
+      "loss": 3.6369,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735929788654468e-05,
+      "loss": 3.6288,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735091193903416e-05,
+      "loss": 3.6296,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.734252599152364e-05,
+      "loss": 3.6425,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.733414004401312e-05,
+      "loss": 3.6435,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.73257540965026e-05,
+      "loss": 3.6333,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.731736814899208e-05,
+      "loss": 3.6285,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730898220148156e-05,
+      "loss": 3.6324,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730061263277477e-05,
+      "loss": 3.6327,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729222668526425e-05,
+      "loss": 3.6319,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.728384073775374e-05,
+      "loss": 3.6179,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.727545479024322e-05,
+      "loss": 3.6367,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7267085221536426e-05,
+      "loss": 3.633,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7258699274025906e-05,
+      "loss": 3.6189,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7250313326515386e-05,
+      "loss": 3.6264,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7241943757808595e-05,
+      "loss": 3.6358,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7233557810298075e-05,
+      "loss": 3.6316,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7225171862787555e-05,
+      "loss": 3.643,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7216785915277035e-05,
+      "loss": 3.6321,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7208399967766515e-05,
+      "loss": 3.6247,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7200014020255995e-05,
+      "loss": 3.6475,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7191628072745475e-05,
+      "loss": 3.6152,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.718325850403869e-05,
+      "loss": 3.6189,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.717487255652817e-05,
+      "loss": 3.6229,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.716648660901765e-05,
+      "loss": 3.6233,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.715810066150713e-05,
+      "loss": 3.6274,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714971471399661e-05,
+      "loss": 3.6291,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714132876648609e-05,
+      "loss": 3.631,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.713294281897557e-05,
+      "loss": 3.6268,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.712455687146504e-05,
+      "loss": 3.6238,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.711618730275826e-05,
+      "loss": 3.6336,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.710780135524774e-05,
+      "loss": 3.6289,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709941540773722e-05,
+      "loss": 3.6084,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709104583903043e-05,
+      "loss": 3.6341,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.708265989151991e-05,
+      "loss": 3.6293,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.707427394400939e-05,
+      "loss": 3.6362,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.706588799649887e-05,
+      "loss": 3.6251,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.705750204898835e-05,
+      "loss": 3.6228,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704911610147783e-05,
+      "loss": 3.6133,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704073015396731e-05,
+      "loss": 3.6204,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703236058526052e-05,
+      "loss": 3.6291,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.702397463775e-05,
+      "loss": 3.6172,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.701558869023948e-05,
+      "loss": 3.6277,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.700720274272896e-05,
+      "loss": 3.6251,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6998833174022166e-05,
+      "loss": 3.6173,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6990447226511646e-05,
+      "loss": 3.6249,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6982061279001126e-05,
+      "loss": 3.6169,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6973675331490606e-05,
+      "loss": 3.6188,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.696528938398009e-05,
+      "loss": 3.6188,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69569198152733e-05,
+      "loss": 3.6323,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694853386776278e-05,
+      "loss": 3.6107,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694014792025226e-05,
+      "loss": 3.6243,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693176197274174e-05,
+      "loss": 3.6024,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.692337602523122e-05,
+      "loss": 3.6225,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.691500645652443e-05,
+      "loss": 3.627,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.690662050901391e-05,
+      "loss": 3.6287,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.689823456150339e-05,
+      "loss": 3.6217,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688984861399287e-05,
+      "loss": 3.6238,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688146266648235e-05,
+      "loss": 3.6413,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.687307671897183e-05,
+      "loss": 3.616,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.686469077146131e-05,
+      "loss": 3.6358,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6856321202754526e-05,
+      "loss": 3.6194,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6847935255244006e-05,
+      "loss": 3.6205,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6839549307733486e-05,
+      "loss": 3.6256,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6831163360222966e-05,
+      "loss": 3.6189,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6822777412712446e-05,
+      "loss": 3.6286,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6814407844005655e-05,
+      "loss": 3.6342,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6806021896495135e-05,
+      "loss": 3.6148,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6797635948984615e-05,
+      "loss": 3.6129,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6789250001474095e-05,
+      "loss": 3.6157,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6780880432767304e-05,
+      "loss": 3.6288,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6772494485256784e-05,
+      "loss": 3.6174,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6764108537746264e-05,
+      "loss": 3.6249,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6755722590235744e-05,
+      "loss": 3.6232,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.674733664272523e-05,
+      "loss": 3.6226,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6738950695214704e-05,
+      "loss": 3.6141,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.673058112650792e-05,
+      "loss": 3.609,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.67221951789974e-05,
+      "loss": 3.6126,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.671380923148688e-05,
+      "loss": 3.6269,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.670542328397635e-05,
+      "loss": 3.6291,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.669703733646583e-05,
+      "loss": 3.6117,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668865138895531e-05,
+      "loss": 3.6145,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668028182024853e-05,
+      "loss": 3.6169,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6671895872738e-05,
+      "loss": 3.6243,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.666350992522748e-05,
+      "loss": 3.6279,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.665512397771697e-05,
+      "loss": 3.618,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664673803020645e-05,
+      "loss": 3.6225,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663835208269593e-05,
+      "loss": 3.6192,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662998251398914e-05,
+      "loss": 3.6155,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662159656647862e-05,
+      "loss": 3.6127,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.66132106189681e-05,
+      "loss": 3.6183,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.660482467145758e-05,
+      "loss": 3.6221,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6596455102750786e-05,
+      "loss": 3.6137,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6588069155240266e-05,
+      "loss": 3.6154,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6579683207729746e-05,
+      "loss": 3.6114,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6571297260219226e-05,
+      "loss": 3.6184,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6562911312708706e-05,
+      "loss": 3.6183,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6554525365198186e-05,
+      "loss": 3.6182,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.65461557964914e-05,
+      "loss": 3.6179,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.653776984898088e-05,
+      "loss": 3.6321,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652938390147036e-05,
+      "loss": 3.6219,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652099795395984e-05,
+      "loss": 3.6201,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.651261200644932e-05,
+      "loss": 3.6231,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.650424243774253e-05,
+      "loss": 3.6165,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.649585649023201e-05,
+      "loss": 3.623,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.648747054272149e-05,
+      "loss": 3.6069,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.647908459521097e-05,
+      "loss": 3.6127,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.647071502650418e-05,
+      "loss": 3.6186,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646232907899366e-05,
+      "loss": 3.6198,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.645394313148314e-05,
+      "loss": 3.6169,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.644555718397262e-05,
+      "loss": 3.6203,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.64371712364621e-05,
+      "loss": 3.6201,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6428785288951587e-05,
+      "loss": 3.6146,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6420415720244796e-05,
+      "loss": 3.6204,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6412029772734276e-05,
+      "loss": 3.617,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6403643825223756e-05,
+      "loss": 3.6138,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6395257877713236e-05,
+      "loss": 3.5928,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6386888309006445e-05,
+      "loss": 3.6279,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6378502361495925e-05,
+      "loss": 3.6165,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6370116413985405e-05,
+      "loss": 3.6285,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6361730466474884e-05,
+      "loss": 3.6114,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.635334451896436e-05,
+      "loss": 3.619,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.634495857145384e-05,
+      "loss": 3.6155,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6336572623943324e-05,
+      "loss": 3.6371,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6328186676432804e-05,
+      "loss": 3.6064,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6319817107726013e-05,
+      "loss": 3.6153,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631143116021549e-05,
+      "loss": 3.6185,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.630306159150871e-05,
+      "loss": 3.6216,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.629467564399819e-05,
+      "loss": 3.6148,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.628628969648766e-05,
+      "loss": 3.6164,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.627790374897714e-05,
+      "loss": 3.6177,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626951780146662e-05,
+      "loss": 3.6228,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.62611318539561e-05,
+      "loss": 3.6175,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.863309860229492,
+      "eval_runtime": 317.3238,
+      "eval_samples_per_second": 1202.529,
+      "eval_steps_per_second": 37.58,
+      "step": 839520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.625276228524931e-05,
+      "loss": 3.6141,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.624437633773879e-05,
+      "loss": 3.6124,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.623599039022828e-05,
+      "loss": 3.6128,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.622760444271776e-05,
+      "loss": 3.6172,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621921849520724e-05,
+      "loss": 3.6222,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621084892650045e-05,
+      "loss": 3.6133,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.620246297898993e-05,
+      "loss": 3.6154,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.619407703147941e-05,
+      "loss": 3.6074,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.618569108396889e-05,
+      "loss": 3.6099,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6177321515262096e-05,
+      "loss": 3.6169,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6168935567751576e-05,
+      "loss": 3.6169,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6160549620241056e-05,
+      "loss": 3.613,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6152180051534265e-05,
+      "loss": 3.6275,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6143794104023745e-05,
+      "loss": 3.6038,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.613540815651323e-05,
+      "loss": 3.6087,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.612702220900271e-05,
+      "loss": 3.6009,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611863626149219e-05,
+      "loss": 3.6158,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611025031398167e-05,
+      "loss": 3.6045,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.610186436647115e-05,
+      "loss": 3.6103,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.609347841896063e-05,
+      "loss": 3.6182,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.608510885025384e-05,
+      "loss": 3.6225,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.607672290274332e-05,
+      "loss": 3.6134,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.60683369552328e-05,
+      "loss": 3.6061,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605995100772228e-05,
+      "loss": 3.6131,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605158143901549e-05,
+      "loss": 3.6139,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.604319549150497e-05,
+      "loss": 3.6091,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.603480954399445e-05,
+      "loss": 3.5974,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.602642359648393e-05,
+      "loss": 3.6131,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6018054027777145e-05,
+      "loss": 3.6118,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6009668080266625e-05,
+      "loss": 3.5951,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6001282132756105e-05,
+      "loss": 3.6076,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5992896185245585e-05,
+      "loss": 3.6156,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5984510237735065e-05,
+      "loss": 3.6096,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5976140669028274e-05,
+      "loss": 3.6208,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5967754721517754e-05,
+      "loss": 3.6109,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5959368774007234e-05,
+      "loss": 3.6061,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595099920530044e-05,
+      "loss": 3.6283,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.594261325778992e-05,
+      "loss": 3.5915,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59342273102794e-05,
+      "loss": 3.6002,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.592584136276888e-05,
+      "loss": 3.6039,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.591745541525837e-05,
+      "loss": 3.6011,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590906946774785e-05,
+      "loss": 3.6066,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590069989904106e-05,
+      "loss": 3.6123,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589231395153054e-05,
+      "loss": 3.6079,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.588392800402002e-05,
+      "loss": 3.6074,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58755420565095e-05,
+      "loss": 3.5997,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.586717248780271e-05,
+      "loss": 3.6103,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585878654029219e-05,
+      "loss": 3.611,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585040059278167e-05,
+      "loss": 3.587,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.584201464527114e-05,
+      "loss": 3.6104,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.583364507656436e-05,
+      "loss": 3.609,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.582525912905384e-05,
+      "loss": 3.6141,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.581687318154332e-05,
+      "loss": 3.6052,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58084872340328e-05,
+      "loss": 3.6026,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580011766532601e-05,
+      "loss": 3.5959,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579173171781549e-05,
+      "loss": 3.5989,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.578334577030497e-05,
+      "loss": 3.6104,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5774959822794446e-05,
+      "loss": 3.5938,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5766573875283926e-05,
+      "loss": 3.6077,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.575820430657714e-05,
+      "loss": 3.6023,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5749818359066615e-05,
+      "loss": 3.6001,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5741432411556095e-05,
+      "loss": 3.6006,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5733046464045575e-05,
+      "loss": 3.5994,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.572466051653506e-05,
+      "loss": 3.6007,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.571629094782827e-05,
+      "loss": 3.5923,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.570790500031775e-05,
+      "loss": 3.6154,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569951905280723e-05,
+      "loss": 3.5938,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569113310529671e-05,
+      "loss": 3.5945,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.568274715778619e-05,
+      "loss": 3.5869,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.567436121027567e-05,
+      "loss": 3.6014,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.566599164156888e-05,
+      "loss": 3.6062,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.565760569405836e-05,
+      "loss": 3.609,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564921974654784e-05,
+      "loss": 3.605,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564083379903732e-05,
+      "loss": 3.6034,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.56324478515268e-05,
+      "loss": 3.6176,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.562407828282001e-05,
+      "loss": 3.5978,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5615692335309495e-05,
+      "loss": 3.6161,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5607306387798975e-05,
+      "loss": 3.5984,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5598920440288455e-05,
+      "loss": 3.5989,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5590550871581664e-05,
+      "loss": 3.602,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5582164924071144e-05,
+      "loss": 3.5978,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5573778976560624e-05,
+      "loss": 3.6068,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5565393029050104e-05,
+      "loss": 3.6175,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5557007081539584e-05,
+      "loss": 3.5978,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5548621134029064e-05,
+      "loss": 3.5882,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554025156532227e-05,
+      "loss": 3.5956,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553186561781175e-05,
+      "loss": 3.607,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.552347967030123e-05,
+      "loss": 3.6007,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.551509372279071e-05,
+      "loss": 3.603,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.550670777528019e-05,
+      "loss": 3.6046,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.549833820657341e-05,
+      "loss": 3.6039,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548995225906289e-05,
+      "loss": 3.5918,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548156631155237e-05,
+      "loss": 3.5869,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.547318036404185e-05,
+      "loss": 3.5961,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.546479441653133e-05,
+      "loss": 3.6014,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.545642484782454e-05,
+      "loss": 3.6152,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.544803890031402e-05,
+      "loss": 3.5874,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.54396529528035e-05,
+      "loss": 3.5963,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543126700529298e-05,
+      "loss": 3.5989,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.542288105778245e-05,
+      "loss": 3.5988,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5414511489075666e-05,
+      "loss": 3.606,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5406125541565146e-05,
+      "loss": 3.5998,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.539773959405463e-05,
+      "loss": 3.6041,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5389353646544106e-05,
+      "loss": 3.5959,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538098407783732e-05,
+      "loss": 3.5973,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53725981303268e-05,
+      "loss": 3.5935,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5364212182816275e-05,
+      "loss": 3.597,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5355826235305755e-05,
+      "loss": 3.6056,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.534745666659897e-05,
+      "loss": 3.5929,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.533907071908845e-05,
+      "loss": 3.5946,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5330684771577924e-05,
+      "loss": 3.5888,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5322298824067404e-05,
+      "loss": 3.5982,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5313912876556884e-05,
+      "loss": 3.5987,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.530552692904637e-05,
+      "loss": 3.6009,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.529715736033958e-05,
+      "loss": 3.5969,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528877141282906e-05,
+      "loss": 3.6047,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528038546531854e-05,
+      "loss": 3.6051,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527199951780802e-05,
+      "loss": 3.6043,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52636135702975e-05,
+      "loss": 3.6037,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.525524400159071e-05,
+      "loss": 3.597,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.524685805408019e-05,
+      "loss": 3.5993,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523847210656967e-05,
+      "loss": 3.5912,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523008615905915e-05,
+      "loss": 3.5927,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.522171659035236e-05,
+      "loss": 3.5998,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.521333064284184e-05,
+      "loss": 3.601,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5204944695331324e-05,
+      "loss": 3.5984,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5196558747820804e-05,
+      "loss": 3.5993,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5188189179114013e-05,
+      "loss": 3.6007,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5179803231603493e-05,
+      "loss": 3.5958,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5171417284092973e-05,
+      "loss": 3.6008,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.516303133658245e-05,
+      "loss": 3.5962,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.515466176787566e-05,
+      "loss": 3.5971,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.514627582036514e-05,
+      "loss": 3.5706,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.513788987285462e-05,
+      "loss": 3.6063,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.51295039253441e-05,
+      "loss": 3.5962,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512111797783358e-05,
+      "loss": 3.6114,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.511274840912679e-05,
+      "loss": 3.5962,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.510436246161628e-05,
+      "loss": 3.5972,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.509597651410576e-05,
+      "loss": 3.5991,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.508759056659524e-05,
+      "loss": 3.6131,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.507920461908472e-05,
+      "loss": 3.5867,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.507083505037793e-05,
+      "loss": 3.5989,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506244910286741e-05,
+      "loss": 3.5942,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.505406315535689e-05,
+      "loss": 3.5998,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.504567720784637e-05,
+      "loss": 3.5976,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5037307639139576e-05,
+      "loss": 3.5978,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5028921691629056e-05,
+      "loss": 3.5953,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5020535744118536e-05,
+      "loss": 3.6039,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5012149796608016e-05,
+      "loss": 3.598,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.864187717437744,
+      "eval_runtime": 306.8833,
+      "eval_samples_per_second": 1243.44,
+      "eval_steps_per_second": 38.858,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5003763849097496e-05,
+      "loss": 3.6003,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4995377901586976e-05,
+      "loss": 3.5928,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4986991954076456e-05,
+      "loss": 3.5912,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4978606006565936e-05,
+      "loss": 3.6002,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4970220059055416e-05,
+      "loss": 3.6014,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496185049034863e-05,
+      "loss": 3.5989,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495346454283811e-05,
+      "loss": 3.5959,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4945078595327585e-05,
+      "loss": 3.5857,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4936692647817065e-05,
+      "loss": 3.59,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492832307911028e-05,
+      "loss": 3.605,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491993713159976e-05,
+      "loss": 3.5936,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4911551184089234e-05,
+      "loss": 3.5912,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.490318161538245e-05,
+      "loss": 3.6106,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489479566787193e-05,
+      "loss": 3.5852,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.488640972036141e-05,
+      "loss": 3.5869,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487802377285089e-05,
+      "loss": 3.5833,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486963782534037e-05,
+      "loss": 3.5948,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486125187782985e-05,
+      "loss": 3.5855,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.485286593031933e-05,
+      "loss": 3.5933,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484447998280881e-05,
+      "loss": 3.6003,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483611041410202e-05,
+      "loss": 3.5998,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48277244665915e-05,
+      "loss": 3.5971,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481933851908098e-05,
+      "loss": 3.5857,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481095257157046e-05,
+      "loss": 3.5917,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480258300286367e-05,
+      "loss": 3.599,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479419705535315e-05,
+      "loss": 3.5887,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4785811107842634e-05,
+      "loss": 3.5786,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4777425160332114e-05,
+      "loss": 3.5907,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476905559162532e-05,
+      "loss": 3.5933,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.47606696441148e-05,
+      "loss": 3.5783,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.475228369660428e-05,
+      "loss": 3.5885,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.474389774909376e-05,
+      "loss": 3.5921,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.473551180158324e-05,
+      "loss": 3.5946,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.472714223287645e-05,
+      "loss": 3.598,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.471875628536593e-05,
+      "loss": 3.592,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.471037033785541e-05,
+      "loss": 3.5908,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470198439034489e-05,
+      "loss": 3.6045,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46936148216381e-05,
+      "loss": 3.5739,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468522887412759e-05,
+      "loss": 3.5808,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467684292661707e-05,
+      "loss": 3.5882,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466845697910655e-05,
+      "loss": 3.5787,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466008741039976e-05,
+      "loss": 3.5895,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4651701462889237e-05,
+      "loss": 3.5923,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4643315515378717e-05,
+      "loss": 3.589,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4634929567868196e-05,
+      "loss": 3.587,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4626543620357676e-05,
+      "loss": 3.5863,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4618174051650886e-05,
+      "loss": 3.5887,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4609788104140365e-05,
+      "loss": 3.5971,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4601402156629845e-05,
+      "loss": 3.5643,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4593016209119325e-05,
+      "loss": 3.5878,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4584630261608805e-05,
+      "loss": 3.5935,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.457626069290202e-05,
+      "loss": 3.5967,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45678747453915e-05,
+      "loss": 3.5866,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455948879788098e-05,
+      "loss": 3.5798,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455110285037046e-05,
+      "loss": 3.5793,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.454273328166367e-05,
+      "loss": 3.5775,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.453434733415315e-05,
+      "loss": 3.5893,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.452596138664263e-05,
+      "loss": 3.5773,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.451757543913211e-05,
+      "loss": 3.5905,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450918949162159e-05,
+      "loss": 3.5811,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45008199229148e-05,
+      "loss": 3.5829,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449243397540428e-05,
+      "loss": 3.5787,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448404802789376e-05,
+      "loss": 3.5839,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447566208038324e-05,
+      "loss": 3.58,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4467292511676455e-05,
+      "loss": 3.5747,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4458906564165935e-05,
+      "loss": 3.5935,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4450520616655415e-05,
+      "loss": 3.5825,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4442134669144895e-05,
+      "loss": 3.5759,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443374872163437e-05,
+      "loss": 3.5671,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4425379152927584e-05,
+      "loss": 3.5854,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4416993205417064e-05,
+      "loss": 3.5835,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440860725790654e-05,
+      "loss": 3.592,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440022131039602e-05,
+      "loss": 3.5817,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.43918353628855e-05,
+      "loss": 3.5887,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.438346579417871e-05,
+      "loss": 3.5969,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.437507984666819e-05,
+      "loss": 3.5841,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436669389915767e-05,
+      "loss": 3.5928,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435830795164715e-05,
+      "loss": 3.5822,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434993838294037e-05,
+      "loss": 3.5802,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434155243542984e-05,
+      "loss": 3.5861,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433316648791932e-05,
+      "loss": 3.5816,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.43247805404088e-05,
+      "loss": 3.5835,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.431641097170201e-05,
+      "loss": 3.5996,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.430802502419149e-05,
+      "loss": 3.5818,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429963907668097e-05,
+      "loss": 3.5704,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429125312917045e-05,
+      "loss": 3.5773,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4282883560463666e-05,
+      "loss": 3.583,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4274497612953146e-05,
+      "loss": 3.5864,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4266111665442626e-05,
+      "loss": 3.5816,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4257725717932106e-05,
+      "loss": 3.5861,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4249339770421586e-05,
+      "loss": 3.5849,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4240970201714795e-05,
+      "loss": 3.5739,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4232584254204275e-05,
+      "loss": 3.5674,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4224198306693755e-05,
+      "loss": 3.5788,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4215812359183235e-05,
+      "loss": 3.5824,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4207426411672715e-05,
+      "loss": 3.5969,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4199056842965924e-05,
+      "loss": 3.569,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4190670895455404e-05,
+      "loss": 3.5771,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4182284947944884e-05,
+      "loss": 3.5827,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4173899000434364e-05,
+      "loss": 3.5819,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.416552943172758e-05,
+      "loss": 3.5889,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.415714348421706e-05,
+      "loss": 3.5814,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414875753670654e-05,
+      "loss": 3.58,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414037158919602e-05,
+      "loss": 3.5819,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413200202048923e-05,
+      "loss": 3.5775,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.412361607297871e-05,
+      "loss": 3.5754,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.411523012546819e-05,
+      "loss": 3.5824,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.410684417795767e-05,
+      "loss": 3.5851,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409847460925088e-05,
+      "loss": 3.5765,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409008866174036e-05,
+      "loss": 3.5733,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408170271422984e-05,
+      "loss": 3.5728,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407331676671932e-05,
+      "loss": 3.5804,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4064947198012534e-05,
+      "loss": 3.58,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4056561250502014e-05,
+      "loss": 3.5816,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4048175302991494e-05,
+      "loss": 3.5786,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4039789355480973e-05,
+      "loss": 3.585,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403141978677418e-05,
+      "loss": 3.5905,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402303383926366e-05,
+      "loss": 3.5824,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401464789175314e-05,
+      "loss": 3.5852,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400626194424262e-05,
+      "loss": 3.5834,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399789237553583e-05,
+      "loss": 3.5762,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398950642802531e-05,
+      "loss": 3.5727,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398112048051479e-05,
+      "loss": 3.574,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397273453300427e-05,
+      "loss": 3.5804,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.396436496429749e-05,
+      "loss": 3.5815,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.395597901678697e-05,
+      "loss": 3.5808,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.394759306927645e-05,
+      "loss": 3.5812,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393920712176593e-05,
+      "loss": 3.5808,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393082117425541e-05,
+      "loss": 3.5763,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3922451605548616e-05,
+      "loss": 3.5847,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3914065658038096e-05,
+      "loss": 3.577,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3905679710527576e-05,
+      "loss": 3.5812,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3897293763017056e-05,
+      "loss": 3.5542,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3888924194310265e-05,
+      "loss": 3.5829,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3880538246799745e-05,
+      "loss": 3.5823,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3872152299289225e-05,
+      "loss": 3.5896,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.386376635177871e-05,
+      "loss": 3.5798,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.385538040426819e-05,
+      "loss": 3.5708,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38470108355614e-05,
+      "loss": 3.5824,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383862488805088e-05,
+      "loss": 3.5964,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383023894054036e-05,
+      "loss": 3.569,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382186937183357e-05,
+      "loss": 3.5827,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.381348342432305e-05,
+      "loss": 3.5771,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380509747681253e-05,
+      "loss": 3.5816,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379671152930201e-05,
+      "loss": 3.5803,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378832558179149e-05,
+      "loss": 3.5795,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.37799560130847e-05,
+      "loss": 3.5798,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377157006557418e-05,
+      "loss": 3.5838,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3763184118063665e-05,
+      "loss": 3.583,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.863978147506714,
+      "eval_runtime": 307.2769,
+      "eval_samples_per_second": 1241.847,
+      "eval_steps_per_second": 38.809,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3754798170553145e-05,
+      "loss": 3.5692,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.374641222304262e-05,
+      "loss": 3.5777,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3738042654335834e-05,
+      "loss": 3.5711,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3729656706825314e-05,
+      "loss": 3.5849,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3721270759314794e-05,
+      "loss": 3.584,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.371288481180427e-05,
+      "loss": 3.5774,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3704515243097483e-05,
+      "loss": 3.5775,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3696129295586963e-05,
+      "loss": 3.5662,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.368774334807644e-05,
+      "loss": 3.573,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3679357400565917e-05,
+      "loss": 3.585,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367098783185913e-05,
+      "loss": 3.5756,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366260188434861e-05,
+      "loss": 3.5773,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365423231564183e-05,
+      "loss": 3.5891,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.364584636813131e-05,
+      "loss": 3.5685,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363746042062079e-05,
+      "loss": 3.5677,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362907447311027e-05,
+      "loss": 3.5655,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362068852559974e-05,
+      "loss": 3.5789,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361230257808922e-05,
+      "loss": 3.5683,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.36039166305787e-05,
+      "loss": 3.5734,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.359553068306818e-05,
+      "loss": 3.5817,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.358716111436139e-05,
+      "loss": 3.5854,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357877516685087e-05,
+      "loss": 3.5791,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357038921934035e-05,
+      "loss": 3.5719,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356200327182984e-05,
+      "loss": 3.5731,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3553633703123046e-05,
+      "loss": 3.5832,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3545247755612526e-05,
+      "loss": 3.571,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3536861808102006e-05,
+      "loss": 3.5623,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3528475860591486e-05,
+      "loss": 3.5757,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3520106291884695e-05,
+      "loss": 3.5695,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3511720344374175e-05,
+      "loss": 3.5636,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3503334396863655e-05,
+      "loss": 3.5699,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3494948449353135e-05,
+      "loss": 3.5741,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3486578880646344e-05,
+      "loss": 3.5759,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3478192933135824e-05,
+      "loss": 3.5805,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3469806985625304e-05,
+      "loss": 3.5759,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346142103811479e-05,
+      "loss": 3.5742,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3453051469408e-05,
+      "loss": 3.5858,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.344466552189748e-05,
+      "loss": 3.5583,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.343627957438696e-05,
+      "loss": 3.5647,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.342789362687644e-05,
+      "loss": 3.5698,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341950767936592e-05,
+      "loss": 3.5565,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341113811065913e-05,
+      "loss": 3.5732,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.340275216314861e-05,
+      "loss": 3.5767,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339436621563809e-05,
+      "loss": 3.5739,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.338598026812757e-05,
+      "loss": 3.5676,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.337761069942078e-05,
+      "loss": 3.5711,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336922475191026e-05,
+      "loss": 3.5711,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3360838804399744e-05,
+      "loss": 3.5807,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3352452856889224e-05,
+      "loss": 3.5486,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334408328818243e-05,
+      "loss": 3.5661,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.333569734067191e-05,
+      "loss": 3.5807,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.332731139316139e-05,
+      "loss": 3.5795,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331892544565087e-05,
+      "loss": 3.5695,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331055587694408e-05,
+      "loss": 3.5587,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.330216992943356e-05,
+      "loss": 3.5636,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.329378398192304e-05,
+      "loss": 3.5601,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.328539803441252e-05,
+      "loss": 3.5737,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3277012086902e-05,
+      "loss": 3.5604,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.326864251819521e-05,
+      "loss": 3.5699,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32602565706847e-05,
+      "loss": 3.5656,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325187062317418e-05,
+      "loss": 3.5678,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.324348467566366e-05,
+      "loss": 3.5608,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.323509872815314e-05,
+      "loss": 3.5687,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.322672915944635e-05,
+      "loss": 3.5582,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.321834321193583e-05,
+      "loss": 3.555,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320995726442531e-05,
+      "loss": 3.5744,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320157131691479e-05,
+      "loss": 3.5677,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3193201748207996e-05,
+      "loss": 3.5577,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3184815800697476e-05,
+      "loss": 3.5522,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3176429853186956e-05,
+      "loss": 3.5661,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3168043905676436e-05,
+      "loss": 3.5675,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315967433696965e-05,
+      "loss": 3.5783,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315128838945913e-05,
+      "loss": 3.5631,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.314290244194861e-05,
+      "loss": 3.5711,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.313451649443809e-05,
+      "loss": 3.5793,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.31261469257313e-05,
+      "loss": 3.5615,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.311776097822078e-05,
+      "loss": 3.5701,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310937503071026e-05,
+      "loss": 3.5694,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.310098908319974e-05,
+      "loss": 3.5669,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309261951449295e-05,
+      "loss": 3.5645,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.308423356698243e-05,
+      "loss": 3.5649,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.307584761947191e-05,
+      "loss": 3.5688,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.306746167196139e-05,
+      "loss": 3.581,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.305907572445087e-05,
+      "loss": 3.5634,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3050706155744085e-05,
+      "loss": 3.5509,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3042320208233565e-05,
+      "loss": 3.56,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3033934260723045e-05,
+      "loss": 3.5669,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3025548313212525e-05,
+      "loss": 3.5685,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3017178744505734e-05,
+      "loss": 3.5674,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3008792796995214e-05,
+      "loss": 3.5669,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3000406849484694e-05,
+      "loss": 3.5709,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299202090197417e-05,
+      "loss": 3.5548,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.298365133326738e-05,
+      "loss": 3.5531,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.297526538575686e-05,
+      "loss": 3.5604,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.296687943824634e-05,
+      "loss": 3.5605,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295849349073582e-05,
+      "loss": 3.5749,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295012392202904e-05,
+      "loss": 3.5602,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294173797451852e-05,
+      "loss": 3.5577,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2933352027008e-05,
+      "loss": 3.5675,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.292496607949747e-05,
+      "loss": 3.5633,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.291659651079069e-05,
+      "loss": 3.5708,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.290821056328017e-05,
+      "loss": 3.5694,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289982461576964e-05,
+      "loss": 3.5622,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289143866825912e-05,
+      "loss": 3.5635,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.288306909955234e-05,
+      "loss": 3.5673,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.287468315204182e-05,
+      "loss": 3.5559,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.28662972045313e-05,
+      "loss": 3.5628,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.285791125702078e-05,
+      "loss": 3.5687,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284954168831399e-05,
+      "loss": 3.5602,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284115574080347e-05,
+      "loss": 3.5558,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2832769793292946e-05,
+      "loss": 3.5568,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2824383845782426e-05,
+      "loss": 3.5615,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.281601427707564e-05,
+      "loss": 3.5653,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2807628329565115e-05,
+      "loss": 3.5637,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2799242382054595e-05,
+      "loss": 3.5583,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2790856434544075e-05,
+      "loss": 3.5703,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278248686583729e-05,
+      "loss": 3.5718,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2774100918326764e-05,
+      "loss": 3.5682,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.276571497081625e-05,
+      "loss": 3.5701,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.275732902330573e-05,
+      "loss": 3.5654,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2748959454598946e-05,
+      "loss": 3.5594,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.274057350708842e-05,
+      "loss": 3.5548,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.27321875595779e-05,
+      "loss": 3.5607,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.272380161206738e-05,
+      "loss": 3.5569,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.271543204336059e-05,
+      "loss": 3.5665,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.270704609585007e-05,
+      "loss": 3.5677,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269866014833955e-05,
+      "loss": 3.5673,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269027420082903e-05,
+      "loss": 3.5615,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268190463212224e-05,
+      "loss": 3.5574,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.267351868461172e-05,
+      "loss": 3.5677,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.26651327371012e-05,
+      "loss": 3.5619,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2656746789590684e-05,
+      "loss": 3.5643,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2648360842080164e-05,
+      "loss": 3.5376,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263999127337337e-05,
+      "loss": 3.564,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263160532586285e-05,
+      "loss": 3.5664,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.262321937835233e-05,
+      "loss": 3.5723,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.261483343084181e-05,
+      "loss": 3.5629,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.260644748333129e-05,
+      "loss": 3.5561,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25980779146245e-05,
+      "loss": 3.5627,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258969196711398e-05,
+      "loss": 3.5822,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258130601960346e-05,
+      "loss": 3.5546,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.257292007209294e-05,
+      "loss": 3.5661,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.256455050338615e-05,
+      "loss": 3.5625,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.255616455587564e-05,
+      "loss": 3.5638,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.254777860836512e-05,
+      "loss": 3.5642,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25393926608546e-05,
+      "loss": 3.5655,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253102309214781e-05,
+      "loss": 3.5579,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2522637144637287e-05,
+      "loss": 3.5711,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2514251197126767e-05,
+      "loss": 3.5621,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8648340702056885,
+      "eval_runtime": 307.9383,
+      "eval_samples_per_second": 1239.18,
+      "eval_steps_per_second": 38.725,
+      "step": 1068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2505865249616247e-05,
+      "loss": 3.5563,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2497479302105726e-05,
+      "loss": 3.562,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2489093354595206e-05,
+      "loss": 3.5544,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2480707407084686e-05,
+      "loss": 3.5686,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2472337838377895e-05,
+      "loss": 3.5663,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2463951890867375e-05,
+      "loss": 3.5659,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2455565943356855e-05,
+      "loss": 3.5569,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2447179995846335e-05,
+      "loss": 3.551,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.243881042713955e-05,
+      "loss": 3.5568,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.243042447962903e-05,
+      "loss": 3.5711,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242203853211851e-05,
+      "loss": 3.5592,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.241365258460799e-05,
+      "loss": 3.56,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.240529939470493e-05,
+      "loss": 3.5736,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.239691344719441e-05,
+      "loss": 3.5552,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238852749968389e-05,
+      "loss": 3.5514,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238014155217337e-05,
+      "loss": 3.5475,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.237175560466285e-05,
+      "loss": 3.56,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.236336965715233e-05,
+      "loss": 3.5539,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.235498370964181e-05,
+      "loss": 3.5567,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.234659776213129e-05,
+      "loss": 3.5633,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2338228193424505e-05,
+      "loss": 3.5742,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2329842245913985e-05,
+      "loss": 3.5605,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2321456298403465e-05,
+      "loss": 3.5547,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2313070350892945e-05,
+      "loss": 3.5582,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2304700782186154e-05,
+      "loss": 3.56,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2296314834675634e-05,
+      "loss": 3.5587,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2287928887165114e-05,
+      "loss": 3.5459,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2279542939654594e-05,
+      "loss": 3.5575,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.22711733709478e-05,
+      "loss": 3.554,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.226278742343728e-05,
+      "loss": 3.5485,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.225440147592676e-05,
+      "loss": 3.5508,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.224601552841624e-05,
+      "loss": 3.5619,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223762958090572e-05,
+      "loss": 3.5555,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222926001219894e-05,
+      "loss": 3.5666,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222087406468842e-05,
+      "loss": 3.5587,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.22124881171779e-05,
+      "loss": 3.5562,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220410216966737e-05,
+      "loss": 3.5652,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.219573260096059e-05,
+      "loss": 3.5448,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.218734665345007e-05,
+      "loss": 3.5537,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.217896070593955e-05,
+      "loss": 3.5498,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.217057475842902e-05,
+      "loss": 3.5417,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2162205189722236e-05,
+      "loss": 3.5569,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2153819242211716e-05,
+      "loss": 3.5597,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2145433294701196e-05,
+      "loss": 3.5524,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2137047347190676e-05,
+      "loss": 3.5532,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212867777848389e-05,
+      "loss": 3.5537,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212029183097337e-05,
+      "loss": 3.558,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2111905883462845e-05,
+      "loss": 3.5591,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2103519935952325e-05,
+      "loss": 3.5374,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.209515036724554e-05,
+      "loss": 3.5507,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.208676441973502e-05,
+      "loss": 3.5622,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2078378472224494e-05,
+      "loss": 3.5616,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2069992524713974e-05,
+      "loss": 3.5531,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206162295600719e-05,
+      "loss": 3.5432,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.205323700849667e-05,
+      "loss": 3.5485,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.204485106098615e-05,
+      "loss": 3.5464,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.203646511347563e-05,
+      "loss": 3.5523,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2028095544768846e-05,
+      "loss": 3.5452,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201970959725832e-05,
+      "loss": 3.5559,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20113236497478e-05,
+      "loss": 3.5538,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.200293770223728e-05,
+      "loss": 3.547,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1994568133530495e-05,
+      "loss": 3.5413,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.198618218601997e-05,
+      "loss": 3.5544,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.197779623850945e-05,
+      "loss": 3.5409,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.196941029099893e-05,
+      "loss": 3.5415,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1961040722292144e-05,
+      "loss": 3.5564,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1952654774781624e-05,
+      "loss": 3.5522,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1944268827271104e-05,
+      "loss": 3.5428,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1935882879760584e-05,
+      "loss": 3.5346,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.192751331105379e-05,
+      "loss": 3.5518,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191912736354327e-05,
+      "loss": 3.551,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191074141603275e-05,
+      "loss": 3.559,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190235546852223e-05,
+      "loss": 3.549,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.189398589981544e-05,
+      "loss": 3.5557,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.188559995230492e-05,
+      "loss": 3.5583,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.18772140047944e-05,
+      "loss": 3.5522,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.186882805728388e-05,
+      "loss": 3.5531,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.186045848857709e-05,
+      "loss": 3.5506,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185207254106658e-05,
+      "loss": 3.5557,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.184368659355606e-05,
+      "loss": 3.5462,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.183530064604554e-05,
+      "loss": 3.5493,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1826931077338746e-05,
+      "loss": 3.551,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1818545129828226e-05,
+      "loss": 3.565,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1810159182317706e-05,
+      "loss": 3.5481,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1801773234807186e-05,
+      "loss": 3.5359,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1793403666100395e-05,
+      "loss": 3.5467,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1785017718589875e-05,
+      "loss": 3.5514,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1776631771079355e-05,
+      "loss": 3.553,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1768245823568835e-05,
+      "loss": 3.5536,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1759876254862044e-05,
+      "loss": 3.5503,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.175149030735153e-05,
+      "loss": 3.5525,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.174310435984101e-05,
+      "loss": 3.5376,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.173471841233049e-05,
+      "loss": 3.5382,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.17263488436237e-05,
+      "loss": 3.5411,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.171796289611318e-05,
+      "loss": 3.5464,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170957694860266e-05,
+      "loss": 3.559,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170119100109214e-05,
+      "loss": 3.5452,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169282143238535e-05,
+      "loss": 3.5384,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.168443548487483e-05,
+      "loss": 3.5552,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.167604953736431e-05,
+      "loss": 3.5472,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.166766358985379e-05,
+      "loss": 3.5576,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1659294021147e-05,
+      "loss": 3.5516,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1650908073636485e-05,
+      "loss": 3.5422,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1642522126125965e-05,
+      "loss": 3.5493,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1634136178615445e-05,
+      "loss": 3.549,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1625750231104925e-05,
+      "loss": 3.5365,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1617380662398134e-05,
+      "loss": 3.5507,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1608994714887614e-05,
+      "loss": 3.5502,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1600608767377094e-05,
+      "loss": 3.546,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1592222819866574e-05,
+      "loss": 3.5399,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.158385325115978e-05,
+      "loss": 3.5439,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157546730364926e-05,
+      "loss": 3.5419,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.156708135613874e-05,
+      "loss": 3.5495,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.155869540862822e-05,
+      "loss": 3.5511,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.155032583992144e-05,
+      "loss": 3.5431,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154193989241092e-05,
+      "loss": 3.5535,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.15335539449004e-05,
+      "loss": 3.5575,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.152516799738988e-05,
+      "loss": 3.5479,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.151679842868309e-05,
+      "loss": 3.5514,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150841248117257e-05,
+      "loss": 3.5519,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150002653366205e-05,
+      "loss": 3.5462,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.149164058615153e-05,
+      "loss": 3.5368,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1483271017444736e-05,
+      "loss": 3.5469,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1474885069934216e-05,
+      "loss": 3.5439,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1466499122423696e-05,
+      "loss": 3.5494,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1458113174913176e-05,
+      "loss": 3.5498,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144974360620639e-05,
+      "loss": 3.5537,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144135765869587e-05,
+      "loss": 3.5437,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.143297171118535e-05,
+      "loss": 3.5415,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.142458576367483e-05,
+      "loss": 3.5546,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.141621619496804e-05,
+      "loss": 3.5479,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.140783024745752e-05,
+      "loss": 3.5466,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1399444299947e-05,
+      "loss": 3.5263,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.139105835243648e-05,
+      "loss": 3.5466,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138268878372969e-05,
+      "loss": 3.5512,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.137430283621917e-05,
+      "loss": 3.5588,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.136591688870865e-05,
+      "loss": 3.5483,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.135753094119813e-05,
+      "loss": 3.5379,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1349161372491346e-05,
+      "loss": 3.5486,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1340775424980826e-05,
+      "loss": 3.5626,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1332389477470306e-05,
+      "loss": 3.5434,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1324003529959786e-05,
+      "loss": 3.5491,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1315617582449266e-05,
+      "loss": 3.5482,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1307248013742475e-05,
+      "loss": 3.5505,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1298862066231955e-05,
+      "loss": 3.5464,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1290476118721435e-05,
+      "loss": 3.5479,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1282090171210915e-05,
+      "loss": 3.5396,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1273720602504124e-05,
+      "loss": 3.5596,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1265334654993604e-05,
+      "loss": 3.547,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8658666610717773,
+      "eval_runtime": 307.8951,
+      "eval_samples_per_second": 1239.354,
+      "eval_steps_per_second": 38.731,
+      "step": 1144800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1256948707483084e-05,
+      "loss": 3.5398,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1248562759972563e-05,
+      "loss": 3.5455,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.124019319126578e-05,
+      "loss": 3.5396,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123180724375526e-05,
+      "loss": 3.5538,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.122342129624474e-05,
+      "loss": 3.5532,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.121503534873422e-05,
+      "loss": 3.552,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.120666578002743e-05,
+      "loss": 3.5403,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.119827983251691e-05,
+      "loss": 3.535,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118989388500639e-05,
+      "loss": 3.5374,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118150793749587e-05,
+      "loss": 3.5623,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.117313836878908e-05,
+      "loss": 3.5401,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.116475242127856e-05,
+      "loss": 3.5444,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.115636647376804e-05,
+      "loss": 3.5535,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.114799690506125e-05,
+      "loss": 3.5481,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.113961095755073e-05,
+      "loss": 3.5339,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.113122501004021e-05,
+      "loss": 3.538,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.112283906252969e-05,
+      "loss": 3.5371,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1114453115019166e-05,
+      "loss": 3.5386,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1106067167508646e-05,
+      "loss": 3.5424,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1097681219998126e-05,
+      "loss": 3.5484,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.108931165129134e-05,
+      "loss": 3.5604,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1080925703780815e-05,
+      "loss": 3.5426,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1072539756270295e-05,
+      "loss": 3.5434,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1064153808759775e-05,
+      "loss": 3.5382,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.105578424005299e-05,
+      "loss": 3.5482,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.104739829254247e-05,
+      "loss": 3.5384,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.103901234503195e-05,
+      "loss": 3.5344,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.103062639752143e-05,
+      "loss": 3.541,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102225682881464e-05,
+      "loss": 3.5424,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.101387088130412e-05,
+      "loss": 3.5332,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.10054849337936e-05,
+      "loss": 3.5325,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.099709898628308e-05,
+      "loss": 3.549,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.098872941757629e-05,
+      "loss": 3.5404,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.098034347006577e-05,
+      "loss": 3.5458,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097195752255525e-05,
+      "loss": 3.5491,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.096357157504473e-05,
+      "loss": 3.5413,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.095520200633794e-05,
+      "loss": 3.5471,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0946816058827424e-05,
+      "loss": 3.532,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0938430111316904e-05,
+      "loss": 3.5368,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0930044163806384e-05,
+      "loss": 3.5335,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0921674595099593e-05,
+      "loss": 3.5323,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0913288647589073e-05,
+      "loss": 3.5363,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.090490270007855e-05,
+      "loss": 3.5483,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.089651675256803e-05,
+      "loss": 3.534,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.088813080505751e-05,
+      "loss": 3.5422,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087976123635072e-05,
+      "loss": 3.5367,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.08713752888402e-05,
+      "loss": 3.5428,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.086298934132968e-05,
+      "loss": 3.5443,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.085460339381916e-05,
+      "loss": 3.5229,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.084623382511238e-05,
+      "loss": 3.5323,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.083784787760186e-05,
+      "loss": 3.5446,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.082946193009134e-05,
+      "loss": 3.5477,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.082107598258082e-05,
+      "loss": 3.5351,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.081270641387403e-05,
+      "loss": 3.5323,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.080432046636351e-05,
+      "loss": 3.5355,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.079593451885299e-05,
+      "loss": 3.5281,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.078754857134247e-05,
+      "loss": 3.5366,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0779179002635676e-05,
+      "loss": 3.5305,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0770793055125156e-05,
+      "loss": 3.5416,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0762407107614636e-05,
+      "loss": 3.5358,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0754021160104116e-05,
+      "loss": 3.5337,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.074565159139733e-05,
+      "loss": 3.5249,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.073726564388681e-05,
+      "loss": 3.5412,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.072887969637629e-05,
+      "loss": 3.5242,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.072049374886577e-05,
+      "loss": 3.531,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071212418015898e-05,
+      "loss": 3.5407,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.070373823264846e-05,
+      "loss": 3.5403,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.069535228513794e-05,
+      "loss": 3.5234,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.068696633762742e-05,
+      "loss": 3.5217,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.067859676892063e-05,
+      "loss": 3.535,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.067021082141011e-05,
+      "loss": 3.5351,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.066182487389959e-05,
+      "loss": 3.5443,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.065343892638907e-05,
+      "loss": 3.53,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0645069357682285e-05,
+      "loss": 3.5453,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0636683410171765e-05,
+      "loss": 3.5396,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0628297462661245e-05,
+      "loss": 3.5397,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0619911515150725e-05,
+      "loss": 3.5392,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0611541946443934e-05,
+      "loss": 3.5377,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0603155998933414e-05,
+      "loss": 3.5404,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0594770051422894e-05,
+      "loss": 3.5287,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0586384103912374e-05,
+      "loss": 3.5351,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0578014535205583e-05,
+      "loss": 3.5361,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056962858769506e-05,
+      "loss": 3.5512,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056124264018454e-05,
+      "loss": 3.5323,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055285669267402e-05,
+      "loss": 3.5288,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.054448712396724e-05,
+      "loss": 3.5251,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.053610117645672e-05,
+      "loss": 3.5399,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.05277152289462e-05,
+      "loss": 3.5345,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.051932928143568e-05,
+      "loss": 3.5413,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0510943333925156e-05,
+      "loss": 3.5327,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0502573765218368e-05,
+      "loss": 3.5378,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0494187817707848e-05,
+      "loss": 3.5235,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0485801870197328e-05,
+      "loss": 3.5266,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0477415922686808e-05,
+      "loss": 3.5251,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0469046353980017e-05,
+      "loss": 3.5299,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.04606604064695e-05,
+      "loss": 3.5425,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.045227445895898e-05,
+      "loss": 3.5361,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.044388851144846e-05,
+      "loss": 3.5236,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.043551894274167e-05,
+      "loss": 3.5359,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.042713299523115e-05,
+      "loss": 3.5355,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.041874704772063e-05,
+      "loss": 3.5409,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.041036110021011e-05,
+      "loss": 3.5362,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0401991531503322e-05,
+      "loss": 3.53,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.03936055839928e-05,
+      "loss": 3.5318,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.038521963648228e-05,
+      "loss": 3.5384,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.037683368897176e-05,
+      "loss": 3.521,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.036846412026497e-05,
+      "loss": 3.5333,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0360078172754454e-05,
+      "loss": 3.5377,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0351692225243934e-05,
+      "loss": 3.5274,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0343306277733414e-05,
+      "loss": 3.5262,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0334936709026623e-05,
+      "loss": 3.5324,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0326550761516103e-05,
+      "loss": 3.5263,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0318164814005583e-05,
+      "loss": 3.5357,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0309778866495063e-05,
+      "loss": 3.5338,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0301409297788275e-05,
+      "loss": 3.5262,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0293023350277755e-05,
+      "loss": 3.5426,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0284637402767235e-05,
+      "loss": 3.5425,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0276251455256715e-05,
+      "loss": 3.5338,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0267881886549924e-05,
+      "loss": 3.5412,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0259495939039408e-05,
+      "loss": 3.534,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0251109991528888e-05,
+      "loss": 3.5307,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0242724044018368e-05,
+      "loss": 3.5259,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0234354475311577e-05,
+      "loss": 3.5275,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0225968527801057e-05,
+      "loss": 3.5289,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0217582580290537e-05,
+      "loss": 3.5324,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0209196632780013e-05,
+      "loss": 3.5316,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.020082706407323e-05,
+      "loss": 3.5395,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.019244111656271e-05,
+      "loss": 3.5306,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.018405516905219e-05,
+      "loss": 3.5271,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0175669221541662e-05,
+      "loss": 3.5384,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0167299652834878e-05,
+      "loss": 3.5356,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.015891370532436e-05,
+      "loss": 3.5318,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.015052775781384e-05,
+      "loss": 3.5148,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0142141810303314e-05,
+      "loss": 3.5287,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.013377224159653e-05,
+      "loss": 3.5404,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.012538629408601e-05,
+      "loss": 3.5394,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0117000346575487e-05,
+      "loss": 3.5324,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0108614399064967e-05,
+      "loss": 3.5225,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0100244830358183e-05,
+      "loss": 3.534,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0091858882847663e-05,
+      "loss": 3.546,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0083472935337136e-05,
+      "loss": 3.5296,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0075086987826616e-05,
+      "loss": 3.5344,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.006671741911983e-05,
+      "loss": 3.5306,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0058331471609315e-05,
+      "loss": 3.5337,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0049945524098788e-05,
+      "loss": 3.529,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0041559576588268e-05,
+      "loss": 3.541,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0033190007881484e-05,
+      "loss": 3.5233,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.002480406037096e-05,
+      "loss": 3.5417,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.001641811286044e-05,
+      "loss": 3.5315,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.000803216534992e-05,
+      "loss": 3.5256,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.866699457168579,
+      "eval_runtime": 338.1606,
+      "eval_samples_per_second": 1128.431,
+      "eval_steps_per_second": 35.264,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.99996462178394e-05,
+      "loss": 3.532,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.999126027032888e-05,
+      "loss": 3.5262,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.998287432281836e-05,
+      "loss": 3.5379,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.997450475411157e-05,
+      "loss": 3.5337,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9966118806601053e-05,
+      "loss": 3.538,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9957732859090533e-05,
+      "loss": 3.5264,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9949346911580013e-05,
+      "loss": 3.5209,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9940960964069493e-05,
+      "loss": 3.5217,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9932591395362702e-05,
+      "loss": 3.5471,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9924205447852182e-05,
+      "loss": 3.53,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9915819500341662e-05,
+      "loss": 3.5257,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.990743355283114e-05,
+      "loss": 3.5357,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9899063984124354e-05,
+      "loss": 3.5349,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9890678036613834e-05,
+      "loss": 3.5224,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9882292089103314e-05,
+      "loss": 3.5226,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9873906141592794e-05,
+      "loss": 3.521,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9865536572886006e-05,
+      "loss": 3.5259,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9857150625375486e-05,
+      "loss": 3.5236,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9848764677864966e-05,
+      "loss": 3.5378,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9840378730354446e-05,
+      "loss": 3.5471,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9832009161647655e-05,
+      "loss": 3.5292,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9823623214137135e-05,
+      "loss": 3.5224,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9815237266626615e-05,
+      "loss": 3.5276,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9806867697919828e-05,
+      "loss": 3.5346,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9798481750409308e-05,
+      "loss": 3.521,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9790095802898788e-05,
+      "loss": 3.5217,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9781709855388268e-05,
+      "loss": 3.5241,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9773340286681477e-05,
+      "loss": 3.534,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9764954339170957e-05,
+      "loss": 3.5155,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.975656839166044e-05,
+      "loss": 3.5168,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.974818244414992e-05,
+      "loss": 3.5345,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.973981287544313e-05,
+      "loss": 3.5291,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.973142692793261e-05,
+      "loss": 3.5294,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.972304098042209e-05,
+      "loss": 3.5359,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.971465503291157e-05,
+      "loss": 3.5247,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.970626908540105e-05,
+      "loss": 3.5317,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.969789951669426e-05,
+      "loss": 3.5187,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.968951356918374e-05,
+      "loss": 3.5247,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.968112762167322e-05,
+      "loss": 3.5157,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.96727416741627e-05,
+      "loss": 3.516,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.966437210545591e-05,
+      "loss": 3.5232,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9655986157945394e-05,
+      "loss": 3.5329,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9647600210434874e-05,
+      "loss": 3.52,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9639214262924354e-05,
+      "loss": 3.5265,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9630844694217563e-05,
+      "loss": 3.5242,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9622458746707043e-05,
+      "loss": 3.5269,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9614072799196523e-05,
+      "loss": 3.5305,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9605686851686003e-05,
+      "loss": 3.5133,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9597317282979215e-05,
+      "loss": 3.5144,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9588931335468695e-05,
+      "loss": 3.5304,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9580545387958175e-05,
+      "loss": 3.535,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9572159440447655e-05,
+      "loss": 3.5188,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9563789871740864e-05,
+      "loss": 3.5197,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9555403924230347e-05,
+      "loss": 3.5221,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9547017976719827e-05,
+      "loss": 3.5121,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9538632029209307e-05,
+      "loss": 3.5209,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9530262460502516e-05,
+      "loss": 3.5156,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9521876512991996e-05,
+      "loss": 3.5261,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9513490565481476e-05,
+      "loss": 3.5234,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9505104617970956e-05,
+      "loss": 3.5228,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.949673504926417e-05,
+      "loss": 3.5098,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.948834910175365e-05,
+      "loss": 3.5283,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.947996315424313e-05,
+      "loss": 3.5091,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.947157720673261e-05,
+      "loss": 3.5176,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9463207638025818e-05,
+      "loss": 3.5254,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.94548216905153e-05,
+      "loss": 3.5278,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.944643574300478e-05,
+      "loss": 3.509,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.943804979549426e-05,
+      "loss": 3.5097,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.942968022678747e-05,
+      "loss": 3.5164,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.942129427927695e-05,
+      "loss": 3.5257,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.941290833176643e-05,
+      "loss": 3.5235,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.940452238425591e-05,
+      "loss": 3.5186,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9396152815549122e-05,
+      "loss": 3.5284,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9387766868038602e-05,
+      "loss": 3.5294,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9379380920528082e-05,
+      "loss": 3.5225,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9370994973017562e-05,
+      "loss": 3.5234,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9362641783114504e-05,
+      "loss": 3.5231,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9354255835603984e-05,
+      "loss": 3.5293,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9345869888093464e-05,
+      "loss": 3.5104,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9337483940582944e-05,
+      "loss": 3.5231,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9329097993072424e-05,
+      "loss": 3.5192,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9320712045561904e-05,
+      "loss": 3.5357,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9312326098051384e-05,
+      "loss": 3.5208,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.930394015054086e-05,
+      "loss": 3.5093,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9295570581834076e-05,
+      "loss": 3.5129,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9287184634323556e-05,
+      "loss": 3.5205,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9278798686813036e-05,
+      "loss": 3.5262,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.927041273930251e-05,
+      "loss": 3.5244,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9262026791791993e-05,
+      "loss": 3.5174,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.925365722308521e-05,
+      "loss": 3.52,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.924527127557469e-05,
+      "loss": 3.5127,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.923688532806416e-05,
+      "loss": 3.5094,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.922849938055364e-05,
+      "loss": 3.509,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9220129811846857e-05,
+      "loss": 3.5222,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9211743864336334e-05,
+      "loss": 3.5261,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9203357916825814e-05,
+      "loss": 3.5199,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9194971969315294e-05,
+      "loss": 3.5123,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.918660240060851e-05,
+      "loss": 3.5188,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9178216453097983e-05,
+      "loss": 3.5213,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9169830505587463e-05,
+      "loss": 3.5268,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9161444558076946e-05,
+      "loss": 3.5223,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9153074989370162e-05,
+      "loss": 3.5182,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9144689041859635e-05,
+      "loss": 3.5152,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9136303094349115e-05,
+      "loss": 3.5269,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9127917146838595e-05,
+      "loss": 3.509,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9119547578131808e-05,
+      "loss": 3.5165,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9111161630621288e-05,
+      "loss": 3.5247,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9102775683110768e-05,
+      "loss": 3.5155,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9094389735600248e-05,
+      "loss": 3.51,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9086020166893457e-05,
+      "loss": 3.519,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9077634219382937e-05,
+      "loss": 3.5167,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9069248271872417e-05,
+      "loss": 3.5179,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.90608623243619e-05,
+      "loss": 3.5162,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.905249275565511e-05,
+      "loss": 3.5156,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.904410680814459e-05,
+      "loss": 3.5252,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.903572086063407e-05,
+      "loss": 3.5269,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.902733491312355e-05,
+      "loss": 3.5182,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.901896534441676e-05,
+      "loss": 3.5296,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.901057939690624e-05,
+      "loss": 3.5223,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.900219344939572e-05,
+      "loss": 3.515,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.89938075018852e-05,
+      "loss": 3.5089,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.898543793317841e-05,
+      "loss": 3.5153,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.897705198566789e-05,
+      "loss": 3.5176,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.896866603815737e-05,
+      "loss": 3.5171,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.896028009064685e-05,
+      "loss": 3.5208,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8951910521940063e-05,
+      "loss": 3.5221,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8943524574429543e-05,
+      "loss": 3.5221,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8935138626919023e-05,
+      "loss": 3.5124,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8926752679408503e-05,
+      "loss": 3.5226,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8918383110701715e-05,
+      "loss": 3.5241,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8909997163191195e-05,
+      "loss": 3.5142,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8901611215680675e-05,
+      "loss": 3.5037,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8893225268170155e-05,
+      "loss": 3.5131,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8884855699463364e-05,
+      "loss": 3.5281,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8876469751952844e-05,
+      "loss": 3.5276,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8868083804442324e-05,
+      "loss": 3.5107,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8859697856931804e-05,
+      "loss": 3.5111,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8851328288225016e-05,
+      "loss": 3.5275,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8842942340714496e-05,
+      "loss": 3.5271,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8834556393203976e-05,
+      "loss": 3.5199,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8826170445693456e-05,
+      "loss": 3.5174,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8817800876986665e-05,
+      "loss": 3.5199,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.880941492947615e-05,
+      "loss": 3.5187,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.880102898196563e-05,
+      "loss": 3.5133,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.879264303445511e-05,
+      "loss": 3.5291,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8784273465748318e-05,
+      "loss": 3.5149,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8775887518237798e-05,
+      "loss": 3.528,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8767501570727278e-05,
+      "loss": 3.5199,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8759115623216758e-05,
+      "loss": 3.5086,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8688018321990967,
+      "eval_runtime": 308.3479,
+      "eval_samples_per_second": 1237.534,
+      "eval_steps_per_second": 38.674,
+      "step": 1297440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.875074605450997e-05,
+      "loss": 3.5154,
+      "step": 1297920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.874236010699945e-05,
+      "loss": 3.514,
+      "step": 1298432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.873397415948893e-05,
+      "loss": 3.5233,
+      "step": 1298944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.872558821197841e-05,
+      "loss": 3.5208,
+      "step": 1299456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.871720226446789e-05,
+      "loss": 3.5255,
+      "step": 1299968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8708832695761102e-05,
+      "loss": 3.5123,
+      "step": 1300480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8700446748250582e-05,
+      "loss": 3.5094,
+      "step": 1300992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8692060800740062e-05,
+      "loss": 3.5012,
+      "step": 1301504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8683674853229542e-05,
+      "loss": 3.533,
+      "step": 1302016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.867530528452275e-05,
+      "loss": 3.519,
+      "step": 1302528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.866691933701223e-05,
+      "loss": 3.5105,
+      "step": 1303040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.865853338950171e-05,
+      "loss": 3.5239,
+      "step": 1303552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8650163820794924e-05,
+      "loss": 3.5153,
+      "step": 1304064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8641777873284404e-05,
+      "loss": 3.5109,
+      "step": 1304576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8633391925773884e-05,
+      "loss": 3.5079,
+      "step": 1305088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8625005978263364e-05,
+      "loss": 3.5113,
+      "step": 1305600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8616620030752843e-05,
+      "loss": 3.5088,
+      "step": 1306112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8608234083242323e-05,
+      "loss": 3.5108,
+      "step": 1306624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8599848135731803e-05,
+      "loss": 3.5268,
+      "step": 1307136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8591478567025016e-05,
+      "loss": 3.5318,
+      "step": 1307648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8583092619514496e-05,
+      "loss": 3.5154,
+      "step": 1308160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8574706672003976e-05,
+      "loss": 3.5096,
+      "step": 1308672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8566320724493456e-05,
+      "loss": 3.5142,
+      "step": 1309184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8557934776982936e-05,
+      "loss": 3.5193,
+      "step": 1309696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8549565208276148e-05,
+      "loss": 3.5095,
+      "step": 1310208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8541179260765628e-05,
+      "loss": 3.5065,
+      "step": 1310720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8532793313255108e-05,
+      "loss": 3.5104,
+      "step": 1311232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8524423744548317e-05,
+      "loss": 3.5216,
+      "step": 1311744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8516037797037797e-05,
+      "loss": 3.4975,
+      "step": 1312256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8507651849527277e-05,
+      "loss": 3.505,
+      "step": 1312768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8499265902016757e-05,
+      "loss": 3.5184,
+      "step": 1313280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.849089633330997e-05,
+      "loss": 3.517,
+      "step": 1313792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.848251038579945e-05,
+      "loss": 3.5226,
+      "step": 1314304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.847412443828893e-05,
+      "loss": 3.517,
+      "step": 1314816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.846573849077841e-05,
+      "loss": 3.5119,
+      "step": 1315328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8457352543267886e-05,
+      "loss": 3.5199,
+      "step": 1315840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8448982974561102e-05,
+      "loss": 3.503,
+      "step": 1316352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8440597027050582e-05,
+      "loss": 3.5127,
+      "step": 1316864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8432211079540062e-05,
+      "loss": 3.5005,
+      "step": 1317376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8423825132029535e-05,
+      "loss": 3.5047,
+      "step": 1317888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.841545556332275e-05,
+      "loss": 3.5098,
+      "step": 1318400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.840706961581223e-05,
+      "loss": 3.5164,
+      "step": 1318912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.839868366830171e-05,
+      "loss": 3.5135,
+      "step": 1319424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8390297720791187e-05,
+      "loss": 3.5086,
+      "step": 1319936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8381928152084403e-05,
+      "loss": 3.5128,
+      "step": 1320448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8373542204573883e-05,
+      "loss": 3.5143,
+      "step": 1320960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8365156257063356e-05,
+      "loss": 3.5139,
+      "step": 1321472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.835677030955284e-05,
+      "loss": 3.5031,
+      "step": 1321984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8348400740846056e-05,
+      "loss": 3.5006,
+      "step": 1322496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8340014793335535e-05,
+      "loss": 3.514,
+      "step": 1323008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.833162884582501e-05,
+      "loss": 3.5206,
+      "step": 1323520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.832324289831449e-05,
+      "loss": 3.5083,
+      "step": 1324032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8314873329607704e-05,
+      "loss": 3.5065,
+      "step": 1324544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8306487382097184e-05,
+      "loss": 3.5037,
+      "step": 1325056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.829810143458666e-05,
+      "loss": 3.502,
+      "step": 1325568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.828971548707614e-05,
+      "loss": 3.5072,
+      "step": 1326080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8281345918369357e-05,
+      "loss": 3.5014,
+      "step": 1326592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.827295997085883e-05,
+      "loss": 3.5121,
+      "step": 1327104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.826457402334831e-05,
+      "loss": 3.5107,
+      "step": 1327616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8256188075837793e-05,
+      "loss": 3.5058,
+      "step": 1328128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.824781850713101e-05,
+      "loss": 3.4951,
+      "step": 1328640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8239432559620482e-05,
+      "loss": 3.516,
+      "step": 1329152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8231046612109962e-05,
+      "loss": 3.4954,
+      "step": 1329664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8222660664599442e-05,
+      "loss": 3.5052,
+      "step": 1330176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8214291095892658e-05,
+      "loss": 3.5159,
+      "step": 1330688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8205905148382135e-05,
+      "loss": 3.5134,
+      "step": 1331200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8197519200871615e-05,
+      "loss": 3.4977,
+      "step": 1331712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8189133253361095e-05,
+      "loss": 3.4982,
+      "step": 1332224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8180763684654304e-05,
+      "loss": 3.5025,
+      "step": 1332736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8172377737143784e-05,
+      "loss": 3.5083,
+      "step": 1333248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8163991789633264e-05,
+      "loss": 3.5103,
+      "step": 1333760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8155605842122747e-05,
+      "loss": 3.5075,
+      "step": 1334272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8147236273415956e-05,
+      "loss": 3.5155,
+      "step": 1334784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8138850325905436e-05,
+      "loss": 3.5138,
+      "step": 1335296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8130464378394916e-05,
+      "loss": 3.5137,
+      "step": 1335808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8122078430884396e-05,
+      "loss": 3.5021,
+      "step": 1336320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.811370886217761e-05,
+      "loss": 3.5146,
+      "step": 1336832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.810533929347082e-05,
+      "loss": 3.5154,
+      "step": 1337344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8096953345960304e-05,
+      "loss": 3.4942,
+      "step": 1337856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8088567398449777e-05,
+      "loss": 3.5155,
+      "step": 1338368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8080181450939257e-05,
+      "loss": 3.5021,
+      "step": 1338880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8071795503428737e-05,
+      "loss": 3.5266,
+      "step": 1339392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8063409555918217e-05,
+      "loss": 3.5085,
+      "step": 1339904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8055023608407697e-05,
+      "loss": 3.4918,
+      "step": 1340416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.804663766089718e-05,
+      "loss": 3.506,
+      "step": 1340928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.803826809219039e-05,
+      "loss": 3.5005,
+      "step": 1341440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.802988214467987e-05,
+      "loss": 3.5166,
+      "step": 1341952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.802149619716935e-05,
+      "loss": 3.5088,
+      "step": 1342464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.801311024965883e-05,
+      "loss": 3.5061,
+      "step": 1342976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8004740680952042e-05,
+      "loss": 3.5089,
+      "step": 1343488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7996354733441522e-05,
+      "loss": 3.4985,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7987968785931002e-05,
+      "loss": 3.498,
+      "step": 1344512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7979582838420482e-05,
+      "loss": 3.4948,
+      "step": 1345024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.797121326971369e-05,
+      "loss": 3.5082,
+      "step": 1345536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.796282732220317e-05,
+      "loss": 3.5123,
+      "step": 1346048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.795444137469265e-05,
+      "loss": 3.5076,
+      "step": 1346560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7946055427182134e-05,
+      "loss": 3.4974,
+      "step": 1347072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7937685858475343e-05,
+      "loss": 3.5057,
+      "step": 1347584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7929299910964823e-05,
+      "loss": 3.5057,
+      "step": 1348096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7920913963454303e-05,
+      "loss": 3.5132,
+      "step": 1348608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7912528015943783e-05,
+      "loss": 3.507,
+      "step": 1349120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7904158447236996e-05,
+      "loss": 3.5044,
+      "step": 1349632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7895772499726476e-05,
+      "loss": 3.5043,
+      "step": 1350144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7887386552215956e-05,
+      "loss": 3.5122,
+      "step": 1350656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7879000604705436e-05,
+      "loss": 3.4935,
+      "step": 1351168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7870631035998645e-05,
+      "loss": 3.5081,
+      "step": 1351680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7862245088488125e-05,
+      "loss": 3.5097,
+      "step": 1352192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7853859140977605e-05,
+      "loss": 3.5023,
+      "step": 1352704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7845473193467088e-05,
+      "loss": 3.496,
+      "step": 1353216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7837103624760297e-05,
+      "loss": 3.5047,
+      "step": 1353728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7828717677249777e-05,
+      "loss": 3.5026,
+      "step": 1354240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7820331729739257e-05,
+      "loss": 3.5038,
+      "step": 1354752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7811945782228737e-05,
+      "loss": 3.5036,
+      "step": 1355264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.780357621352195e-05,
+      "loss": 3.4997,
+      "step": 1355776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.779519026601143e-05,
+      "loss": 3.5137,
+      "step": 1356288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.778680431850091e-05,
+      "loss": 3.515,
+      "step": 1356800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.777841837099039e-05,
+      "loss": 3.5069,
+      "step": 1357312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.77700488022836e-05,
+      "loss": 3.5102,
+      "step": 1357824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.776166285477308e-05,
+      "loss": 3.5102,
+      "step": 1358336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7753276907262558e-05,
+      "loss": 3.5028,
+      "step": 1358848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.774489095975204e-05,
+      "loss": 3.4962,
+      "step": 1359360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.773652139104525e-05,
+      "loss": 3.4976,
+      "step": 1359872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.772813544353473e-05,
+      "loss": 3.505,
+      "step": 1360384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.771974949602421e-05,
+      "loss": 3.5072,
+      "step": 1360896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.771136354851369e-05,
+      "loss": 3.5074,
+      "step": 1361408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7702993979806903e-05,
+      "loss": 3.5078,
+      "step": 1361920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7694608032296383e-05,
+      "loss": 3.5028,
+      "step": 1362432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7686222084785863e-05,
+      "loss": 3.5057,
+      "step": 1362944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7677836137275343e-05,
+      "loss": 3.5049,
+      "step": 1363456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7669466568568552e-05,
+      "loss": 3.5124,
+      "step": 1363968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7661080621058032e-05,
+      "loss": 3.5037,
+      "step": 1364480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7652694673547512e-05,
+      "loss": 3.4895,
+      "step": 1364992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7644308726036995e-05,
+      "loss": 3.4958,
+      "step": 1365504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7635939157330204e-05,
+      "loss": 3.5132,
+      "step": 1366016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7627553209819684e-05,
+      "loss": 3.5127,
+      "step": 1366528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7619167262309164e-05,
+      "loss": 3.503,
+      "step": 1367040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7610781314798644e-05,
+      "loss": 3.4961,
+      "step": 1367552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7602411746091857e-05,
+      "loss": 3.5164,
+      "step": 1368064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7594025798581337e-05,
+      "loss": 3.5102,
+      "step": 1368576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7585639851070817e-05,
+      "loss": 3.5091,
+      "step": 1369088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7577253903560297e-05,
+      "loss": 3.5064,
+      "step": 1369600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7568884334853506e-05,
+      "loss": 3.5042,
+      "step": 1370112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7560498387342986e-05,
+      "loss": 3.5063,
+      "step": 1370624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7552112439832466e-05,
+      "loss": 3.4975,
+      "step": 1371136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.754372649232195e-05,
+      "loss": 3.5171,
+      "step": 1371648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7535356923615158e-05,
+      "loss": 3.502,
+      "step": 1372160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7526970976104638e-05,
+      "loss": 3.5138,
+      "step": 1372672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7518585028594118e-05,
+      "loss": 3.5059,
+      "step": 1373184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7510199081083598e-05,
+      "loss": 3.495,
+      "step": 1373696
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.870008945465088,
+      "eval_runtime": 308.3151,
+      "eval_samples_per_second": 1237.666,
+      "eval_steps_per_second": 38.678,
+      "step": 1373760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7501813133573078e-05,
+      "loss": 3.5063,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7493427186062558e-05,
+      "loss": 3.5015,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7485041238552034e-05,
+      "loss": 3.5066,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7476655291041514e-05,
+      "loss": 3.5063,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.746828572233473e-05,
+      "loss": 3.5143,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.745989977482421e-05,
+      "loss": 3.5007,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7451513827313687e-05,
+      "loss": 3.4965,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7443127879803167e-05,
+      "loss": 3.4887,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7434758311096383e-05,
+      "loss": 3.5209,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7426372363585856e-05,
+      "loss": 3.5056,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7417986416075336e-05,
+      "loss": 3.4949,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7409600468564816e-05,
+      "loss": 3.514,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7401247278661764e-05,
+      "loss": 3.5015,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7392861331151244e-05,
+      "loss": 3.5001,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7384475383640724e-05,
+      "loss": 3.4929,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7376089436130204e-05,
+      "loss": 3.4966,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7367703488619677e-05,
+      "loss": 3.4982,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7359317541109157e-05,
+      "loss": 3.4929,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.735093159359864e-05,
+      "loss": 3.5133,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.734254564608812e-05,
+      "loss": 3.5157,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.733417607738133e-05,
+      "loss": 3.5047,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.732579012987081e-05,
+      "loss": 3.4979,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.731740418236029e-05,
+      "loss": 3.5017,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7309034613653505e-05,
+      "loss": 3.5067,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7300648666142982e-05,
+      "loss": 3.4948,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7292262718632462e-05,
+      "loss": 3.4968,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7283876771121942e-05,
+      "loss": 3.4953,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.727550720241515e-05,
+      "loss": 3.5094,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.726712125490463e-05,
+      "loss": 3.4835,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.725873530739411e-05,
+      "loss": 3.4937,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7250349359883594e-05,
+      "loss": 3.5052,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7241979791176803e-05,
+      "loss": 3.5042,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7233593843666283e-05,
+      "loss": 3.5064,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7225207896155763e-05,
+      "loss": 3.5044,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7216821948645243e-05,
+      "loss": 3.4956,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7208452379938456e-05,
+      "loss": 3.5098,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7200066432427935e-05,
+      "loss": 3.4923,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7191680484917415e-05,
+      "loss": 3.4995,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7183294537406895e-05,
+      "loss": 3.4832,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7174924968700104e-05,
+      "loss": 3.4943,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7166539021189584e-05,
+      "loss": 3.4937,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7158153073679064e-05,
+      "loss": 3.5061,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7149767126168544e-05,
+      "loss": 3.4985,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7141397557461757e-05,
+      "loss": 3.4951,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7133011609951237e-05,
+      "loss": 3.5025,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7124625662440717e-05,
+      "loss": 3.5004,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7116239714930197e-05,
+      "loss": 3.4996,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7107870146223406e-05,
+      "loss": 3.4913,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.709948419871289e-05,
+      "loss": 3.4825,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.709109825120237e-05,
+      "loss": 3.5072,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.708271230369185e-05,
+      "loss": 3.503,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7074342734985058e-05,
+      "loss": 3.4969,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7065956787474538e-05,
+      "loss": 3.4951,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7057570839964018e-05,
+      "loss": 3.489,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7049184892453498e-05,
+      "loss": 3.4885,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.704081532374671e-05,
+      "loss": 3.4964,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.703242937623619e-05,
+      "loss": 3.4885,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.702404342872567e-05,
+      "loss": 3.4974,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.701565748121515e-05,
+      "loss": 3.4979,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.700728791250836e-05,
+      "loss": 3.4899,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6998901964997843e-05,
+      "loss": 3.4828,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6990516017487323e-05,
+      "loss": 3.5055,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6982130069976803e-05,
+      "loss": 3.4844,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6973760501270012e-05,
+      "loss": 3.4927,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6965374553759492e-05,
+      "loss": 3.4978,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6956988606248972e-05,
+      "loss": 3.5022,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.694860265873845e-05,
+      "loss": 3.4857,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6940233090031664e-05,
+      "loss": 3.4886,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6931847142521144e-05,
+      "loss": 3.4836,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6923461195010624e-05,
+      "loss": 3.4976,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6915075247500104e-05,
+      "loss": 3.4993,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6906705678793313e-05,
+      "loss": 3.4946,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6898319731282796e-05,
+      "loss": 3.5005,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6889933783772276e-05,
+      "loss": 3.4983,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6881547836261756e-05,
+      "loss": 3.5065,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6873178267554965e-05,
+      "loss": 3.4873,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6864792320044445e-05,
+      "loss": 3.4995,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6856422751337658e-05,
+      "loss": 3.5054,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6848036803827138e-05,
+      "loss": 3.4827,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6839650856316618e-05,
+      "loss": 3.4994,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6831264908806098e-05,
+      "loss": 3.4884,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6822878961295578e-05,
+      "loss": 3.5122,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6814493013785058e-05,
+      "loss": 3.4947,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6806107066274538e-05,
+      "loss": 3.4804,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.679773749756775e-05,
+      "loss": 3.4905,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.678935155005723e-05,
+      "loss": 3.4921,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.678096560254671e-05,
+      "loss": 3.5029,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.677257965503619e-05,
+      "loss": 3.4954,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.67642100863294e-05,
+      "loss": 3.4921,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.675582413881888e-05,
+      "loss": 3.4942,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.674743819130836e-05,
+      "loss": 3.4877,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6739052243797842e-05,
+      "loss": 3.4865,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.673068267509105e-05,
+      "loss": 3.4781,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.672229672758053e-05,
+      "loss": 3.4951,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.671391078007001e-05,
+      "loss": 3.4992,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.670552483255949e-05,
+      "loss": 3.4958,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6697155263852704e-05,
+      "loss": 3.4873,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6688769316342184e-05,
+      "loss": 3.4938,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6680383368831664e-05,
+      "loss": 3.4894,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6671997421321144e-05,
+      "loss": 3.5035,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6663627852614353e-05,
+      "loss": 3.4936,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6655241905103833e-05,
+      "loss": 3.4938,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6646855957593313e-05,
+      "loss": 3.4903,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6638470010082796e-05,
+      "loss": 3.5017,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6630100441376005e-05,
+      "loss": 3.4778,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6621714493865485e-05,
+      "loss": 3.4969,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6613328546354965e-05,
+      "loss": 3.4904,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6604942598844445e-05,
+      "loss": 3.4951,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6596573030137657e-05,
+      "loss": 3.4825,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6588187082627137e-05,
+      "loss": 3.4935,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6579801135116617e-05,
+      "loss": 3.4883,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6571415187606097e-05,
+      "loss": 3.4915,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6563045618899306e-05,
+      "loss": 3.4917,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6554659671388786e-05,
+      "loss": 3.4908,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6546273723878266e-05,
+      "loss": 3.4989,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.653788777636775e-05,
+      "loss": 3.5021,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.652951820766096e-05,
+      "loss": 3.4962,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.652113226015044e-05,
+      "loss": 3.4998,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.651274631263992e-05,
+      "loss": 3.4975,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.65043603651294e-05,
+      "loss": 3.4871,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.649599079642261e-05,
+      "loss": 3.4866,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.648760484891209e-05,
+      "loss": 3.4897,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.647921890140157e-05,
+      "loss": 3.488,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.647083295389105e-05,
+      "loss": 3.4969,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.646246338518426e-05,
+      "loss": 3.4914,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.645407743767374e-05,
+      "loss": 3.4977,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.644569149016322e-05,
+      "loss": 3.4914,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.64373055426527e-05,
+      "loss": 3.4904,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6428935973945912e-05,
+      "loss": 3.4905,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6420550026435392e-05,
+      "loss": 3.5052,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6412164078924872e-05,
+      "loss": 3.4865,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6403778131414352e-05,
+      "loss": 3.4844,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.639540856270756e-05,
+      "loss": 3.4793,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6387022615197045e-05,
+      "loss": 3.4985,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6378636667686525e-05,
+      "loss": 3.4984,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6370250720176005e-05,
+      "loss": 3.4949,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6361881151469214e-05,
+      "loss": 3.4851,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6353495203958694e-05,
+      "loss": 3.5037,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6345109256448174e-05,
+      "loss": 3.4941,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.633672330893765e-05,
+      "loss": 3.4984,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6328353740230866e-05,
+      "loss": 3.4916,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6319967792720346e-05,
+      "loss": 3.4938,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6311581845209826e-05,
+      "loss": 3.493,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6303195897699303e-05,
+      "loss": 3.4836,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6294826328992515e-05,
+      "loss": 3.5021,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6286440381482e-05,
+      "loss": 3.4932,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.627805443397148e-05,
+      "loss": 3.4977,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.626966848646095e-05,
+      "loss": 3.4931,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6261298917754167e-05,
+      "loss": 3.4808,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8721539974212646,
+      "eval_runtime": 306.9529,
+      "eval_samples_per_second": 1243.158,
+      "eval_steps_per_second": 38.85,
+      "step": 1450080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6252912970243647e-05,
+      "loss": 3.4939,
+      "step": 1450496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6244527022733124e-05,
+      "loss": 3.4916,
+      "step": 1451008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6236141075222604e-05,
+      "loss": 3.4925,
+      "step": 1451520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.622777150651582e-05,
+      "loss": 3.4942,
+      "step": 1452032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.62193855590053e-05,
+      "loss": 3.5052,
+      "step": 1452544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6210999611494773e-05,
+      "loss": 3.4904,
+      "step": 1453056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6202613663984253e-05,
+      "loss": 3.482,
+      "step": 1453568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.619424409527747e-05,
+      "loss": 3.4761,
+      "step": 1454080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6185858147766952e-05,
+      "loss": 3.5029,
+      "step": 1454592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6177472200256425e-05,
+      "loss": 3.4973,
+      "step": 1455104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6169086252745905e-05,
+      "loss": 3.4763,
+      "step": 1455616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.616071668403912e-05,
+      "loss": 3.5011,
+      "step": 1456128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.615234711533233e-05,
+      "loss": 3.4896,
+      "step": 1456640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6143961167821814e-05,
+      "loss": 3.488,
+      "step": 1457152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6135575220311294e-05,
+      "loss": 3.4841,
+      "step": 1457664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6127189272800773e-05,
+      "loss": 3.4843,
+      "step": 1458176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6118803325290247e-05,
+      "loss": 3.4847,
+      "step": 1458688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6110417377779727e-05,
+      "loss": 3.4805,
+      "step": 1459200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6102031430269207e-05,
+      "loss": 3.5003,
+      "step": 1459712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.609364548275869e-05,
+      "loss": 3.5016,
+      "step": 1460224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.60852759140519e-05,
+      "loss": 3.4939,
+      "step": 1460736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.607688996654138e-05,
+      "loss": 3.4875,
+      "step": 1461248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.606850401903086e-05,
+      "loss": 3.4872,
+      "step": 1461760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6060134450324068e-05,
+      "loss": 3.4958,
+      "step": 1462272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.605174850281355e-05,
+      "loss": 3.4822,
+      "step": 1462784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.604336255530303e-05,
+      "loss": 3.4862,
+      "step": 1463296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.603497660779251e-05,
+      "loss": 3.4805,
+      "step": 1463808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.602660703908572e-05,
+      "loss": 3.496,
+      "step": 1464320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.60182210915752e-05,
+      "loss": 3.4743,
+      "step": 1464832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.600983514406468e-05,
+      "loss": 3.4828,
+      "step": 1465344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.600144919655416e-05,
+      "loss": 3.4931,
+      "step": 1465856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5993079627847373e-05,
+      "loss": 3.4893,
+      "step": 1466368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5984693680336853e-05,
+      "loss": 3.4935,
+      "step": 1466880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5976307732826333e-05,
+      "loss": 3.4929,
+      "step": 1467392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5967921785315813e-05,
+      "loss": 3.485,
+      "step": 1467904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.595955221660902e-05,
+      "loss": 3.4955,
+      "step": 1468416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5951166269098505e-05,
+      "loss": 3.4825,
+      "step": 1468928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5942780321587985e-05,
+      "loss": 3.4786,
+      "step": 1469440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5934394374077465e-05,
+      "loss": 3.4746,
+      "step": 1469952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5926024805370674e-05,
+      "loss": 3.4805,
+      "step": 1470464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5917638857860154e-05,
+      "loss": 3.4785,
+      "step": 1470976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5909252910349634e-05,
+      "loss": 3.4967,
+      "step": 1471488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5900866962839114e-05,
+      "loss": 3.4847,
+      "step": 1472000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5892497394132326e-05,
+      "loss": 3.4832,
+      "step": 1472512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5884111446621806e-05,
+      "loss": 3.4935,
+      "step": 1473024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5875725499111286e-05,
+      "loss": 3.484,
+      "step": 1473536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5867339551600766e-05,
+      "loss": 3.4932,
+      "step": 1474048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5858969982893975e-05,
+      "loss": 3.4801,
+      "step": 1474560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.585058403538346e-05,
+      "loss": 3.4665,
+      "step": 1475072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.584219808787294e-05,
+      "loss": 3.497,
+      "step": 1475584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.583381214036242e-05,
+      "loss": 3.49,
+      "step": 1476096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5825442571655628e-05,
+      "loss": 3.4905,
+      "step": 1476608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5817056624145108e-05,
+      "loss": 3.4795,
+      "step": 1477120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5808670676634588e-05,
+      "loss": 3.4769,
+      "step": 1477632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5800284729124068e-05,
+      "loss": 3.4769,
+      "step": 1478144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.579191516041728e-05,
+      "loss": 3.4841,
+      "step": 1478656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.578352921290676e-05,
+      "loss": 3.4777,
+      "step": 1479168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.577514326539624e-05,
+      "loss": 3.4796,
+      "step": 1479680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.576675731788572e-05,
+      "loss": 3.4886,
+      "step": 1480192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.575838774917893e-05,
+      "loss": 3.4788,
+      "step": 1480704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5750001801668412e-05,
+      "loss": 3.4697,
+      "step": 1481216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5741615854157892e-05,
+      "loss": 3.4938,
+      "step": 1481728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5733229906647372e-05,
+      "loss": 3.4712,
+      "step": 1482240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.572486033794058e-05,
+      "loss": 3.4762,
+      "step": 1482752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.571647439043006e-05,
+      "loss": 3.4888,
+      "step": 1483264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.570808844291954e-05,
+      "loss": 3.4915,
+      "step": 1483776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.569970249540902e-05,
+      "loss": 3.4706,
+      "step": 1484288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5691332926702234e-05,
+      "loss": 3.4792,
+      "step": 1484800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5682946979191714e-05,
+      "loss": 3.467,
+      "step": 1485312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5674561031681194e-05,
+      "loss": 3.4891,
+      "step": 1485824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5666175084170674e-05,
+      "loss": 3.485,
+      "step": 1486336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5657805515463883e-05,
+      "loss": 3.483,
+      "step": 1486848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5649419567953366e-05,
+      "loss": 3.487,
+      "step": 1487360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5641033620442846e-05,
+      "loss": 3.4869,
+      "step": 1487872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5632647672932326e-05,
+      "loss": 3.4952,
+      "step": 1488384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5624278104225535e-05,
+      "loss": 3.4736,
+      "step": 1488896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5615892156715015e-05,
+      "loss": 3.4861,
+      "step": 1489408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5607522588008227e-05,
+      "loss": 3.4939,
+      "step": 1489920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5599136640497707e-05,
+      "loss": 3.4684,
+      "step": 1490432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5590750692987187e-05,
+      "loss": 3.486,
+      "step": 1490944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5582364745476667e-05,
+      "loss": 3.4792,
+      "step": 1491456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5573978797966147e-05,
+      "loss": 3.494,
+      "step": 1491968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5565592850455627e-05,
+      "loss": 3.487,
+      "step": 1492480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5557206902945107e-05,
+      "loss": 3.468,
+      "step": 1492992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5548820955434587e-05,
+      "loss": 3.4788,
+      "step": 1493504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.55404513867278e-05,
+      "loss": 3.478,
+      "step": 1494016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.553206543921728e-05,
+      "loss": 3.4907,
+      "step": 1494528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.552367949170676e-05,
+      "loss": 3.4828,
+      "step": 1495040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.551530992299997e-05,
+      "loss": 3.4791,
+      "step": 1495552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.550692397548945e-05,
+      "loss": 3.4791,
+      "step": 1496064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.549853802797893e-05,
+      "loss": 3.4792,
+      "step": 1496576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.549015208046841e-05,
+      "loss": 3.4745,
+      "step": 1497088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.548178251176162e-05,
+      "loss": 3.4653,
+      "step": 1497600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.54733965642511e-05,
+      "loss": 3.4814,
+      "step": 1498112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.546501061674058e-05,
+      "loss": 3.4893,
+      "step": 1498624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.545662466923006e-05,
+      "loss": 3.4876,
+      "step": 1499136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5448255100523273e-05,
+      "loss": 3.4691,
+      "step": 1499648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5439869153012753e-05,
+      "loss": 3.482,
+      "step": 1500160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5431483205502233e-05,
+      "loss": 3.4759,
+      "step": 1500672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5423097257991713e-05,
+      "loss": 3.4898,
+      "step": 1501184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5414727689284922e-05,
+      "loss": 3.4811,
+      "step": 1501696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5406341741774402e-05,
+      "loss": 3.4821,
+      "step": 1502208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5397955794263882e-05,
+      "loss": 3.4768,
+      "step": 1502720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5389569846753362e-05,
+      "loss": 3.4882,
+      "step": 1503232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5381200278046575e-05,
+      "loss": 3.4662,
+      "step": 1503744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5372814330536055e-05,
+      "loss": 3.4839,
+      "step": 1504256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5364428383025535e-05,
+      "loss": 3.4745,
+      "step": 1504768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5356042435515015e-05,
+      "loss": 3.4819,
+      "step": 1505280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5347672866808224e-05,
+      "loss": 3.4712,
+      "step": 1505792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5339286919297707e-05,
+      "loss": 3.4861,
+      "step": 1506304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5330900971787187e-05,
+      "loss": 3.4734,
+      "step": 1506816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5322515024276667e-05,
+      "loss": 3.4773,
+      "step": 1507328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5314145455569876e-05,
+      "loss": 3.48,
+      "step": 1507840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5305759508059356e-05,
+      "loss": 3.4775,
+      "step": 1508352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5297373560548836e-05,
+      "loss": 3.4845,
+      "step": 1508864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5288987613038316e-05,
+      "loss": 3.49,
+      "step": 1509376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.528061804433153e-05,
+      "loss": 3.4823,
+      "step": 1509888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5272232096821008e-05,
+      "loss": 3.4879,
+      "step": 1510400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5263846149310488e-05,
+      "loss": 3.4828,
+      "step": 1510912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5255460201799968e-05,
+      "loss": 3.4753,
+      "step": 1511424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5247090633093177e-05,
+      "loss": 3.4722,
+      "step": 1511936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.523870468558266e-05,
+      "loss": 3.4779,
+      "step": 1512448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.523031873807214e-05,
+      "loss": 3.4789,
+      "step": 1512960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.522193279056162e-05,
+      "loss": 3.4834,
+      "step": 1513472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.521356322185483e-05,
+      "loss": 3.4792,
+      "step": 1513984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.520517727434431e-05,
+      "loss": 3.4857,
+      "step": 1514496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.519679132683379e-05,
+      "loss": 3.4762,
+      "step": 1515008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.518840537932327e-05,
+      "loss": 3.4768,
+      "step": 1515520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5180035810616482e-05,
+      "loss": 3.4757,
+      "step": 1516032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5171649863105962e-05,
+      "loss": 3.4909,
+      "step": 1516544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5163263915595442e-05,
+      "loss": 3.4761,
+      "step": 1517056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5154877968084915e-05,
+      "loss": 3.4706,
+      "step": 1517568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.514650839937813e-05,
+      "loss": 3.4649,
+      "step": 1518080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5138122451867614e-05,
+      "loss": 3.4846,
+      "step": 1518592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5129736504357094e-05,
+      "loss": 3.4866,
+      "step": 1519104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5121350556846567e-05,
+      "loss": 3.4849,
+      "step": 1519616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5112980988139783e-05,
+      "loss": 3.4754,
+      "step": 1520128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5104595040629263e-05,
+      "loss": 3.4911,
+      "step": 1520640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5096209093118743e-05,
+      "loss": 3.4808,
+      "step": 1521152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.508782314560822e-05,
+      "loss": 3.4837,
+      "step": 1521664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5079453576901436e-05,
+      "loss": 3.4795,
+      "step": 1522176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5071067629390916e-05,
+      "loss": 3.4803,
+      "step": 1522688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.506268168188039e-05,
+      "loss": 3.4838,
+      "step": 1523200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.505429573436987e-05,
+      "loss": 3.4695,
+      "step": 1523712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5045926165663085e-05,
+      "loss": 3.489,
+      "step": 1524224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5037540218152568e-05,
+      "loss": 3.4857,
+      "step": 1524736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.502915427064204e-05,
+      "loss": 3.4831,
+      "step": 1525248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.502076832313152e-05,
+      "loss": 3.4787,
+      "step": 1525760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5012398754424737e-05,
+      "loss": 3.4732,
+      "step": 1526272
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.872343063354492,
+      "eval_runtime": 306.6327,
+      "eval_samples_per_second": 1244.456,
+      "eval_steps_per_second": 38.89,
+      "step": 1526400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5004012806914217e-05,
+      "loss": 3.4778,
+      "step": 1526784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4995626859403694e-05,
+      "loss": 3.4799,
+      "step": 1527296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4987240911893177e-05,
+      "loss": 3.4806,
+      "step": 1527808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4978854964382657e-05,
+      "loss": 3.4829,
+      "step": 1528320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4970485395675866e-05,
+      "loss": 3.491,
+      "step": 1528832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4962099448165346e-05,
+      "loss": 3.4798,
+      "step": 1529344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4953713500654826e-05,
+      "loss": 3.4679,
+      "step": 1529856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4945327553144306e-05,
+      "loss": 3.4692,
+      "step": 1530368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4936957984437518e-05,
+      "loss": 3.4887,
+      "step": 1530880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4928572036926998e-05,
+      "loss": 3.4834,
+      "step": 1531392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4920186089416478e-05,
+      "loss": 3.4693,
+      "step": 1531904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4911800141905955e-05,
+      "loss": 3.4864,
+      "step": 1532416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.49034469520029e-05,
+      "loss": 3.478,
+      "step": 1532928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.489506100449238e-05,
+      "loss": 3.4775,
+      "step": 1533440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.488667505698186e-05,
+      "loss": 3.4719,
+      "step": 1533952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.487828910947134e-05,
+      "loss": 3.4738,
+      "step": 1534464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.486990316196082e-05,
+      "loss": 3.469,
+      "step": 1534976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.48615172144503e-05,
+      "loss": 3.4685,
+      "step": 1535488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.485313126693978e-05,
+      "loss": 3.4883,
+      "step": 1536000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.484474531942926e-05,
+      "loss": 3.4868,
+      "step": 1536512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4836375750722472e-05,
+      "loss": 3.4845,
+      "step": 1537024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4827989803211952e-05,
+      "loss": 3.4763,
+      "step": 1537536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.481960385570143e-05,
+      "loss": 3.4736,
+      "step": 1538048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.481121790819091e-05,
+      "loss": 3.4827,
+      "step": 1538560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.480284833948412e-05,
+      "loss": 3.4733,
+      "step": 1539072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.47944623919736e-05,
+      "loss": 3.4749,
+      "step": 1539584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.478607644446308e-05,
+      "loss": 3.467,
+      "step": 1540096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.477769049695256e-05,
+      "loss": 3.4792,
+      "step": 1540608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4769320928245773e-05,
+      "loss": 3.4666,
+      "step": 1541120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.476093498073525e-05,
+      "loss": 3.4654,
+      "step": 1541632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.475254903322473e-05,
+      "loss": 3.4817,
+      "step": 1542144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4744179464517946e-05,
+      "loss": 3.4745,
+      "step": 1542656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4735793517007426e-05,
+      "loss": 3.485,
+      "step": 1543168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4727407569496902e-05,
+      "loss": 3.4825,
+      "step": 1543680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4719021621986382e-05,
+      "loss": 3.47,
+      "step": 1544192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4710652053279595e-05,
+      "loss": 3.4845,
+      "step": 1544704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4702266105769075e-05,
+      "loss": 3.4744,
+      "step": 1545216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4693880158258555e-05,
+      "loss": 3.4657,
+      "step": 1545728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4685494210748034e-05,
+      "loss": 3.4625,
+      "step": 1546240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4677124642041247e-05,
+      "loss": 3.467,
+      "step": 1546752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4668738694530724e-05,
+      "loss": 3.4698,
+      "step": 1547264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4660352747020203e-05,
+      "loss": 3.4846,
+      "step": 1547776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4651966799509683e-05,
+      "loss": 3.4727,
+      "step": 1548288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.46435972308029e-05,
+      "loss": 3.4719,
+      "step": 1548800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4635211283292376e-05,
+      "loss": 3.4814,
+      "step": 1549312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4626825335781856e-05,
+      "loss": 3.4734,
+      "step": 1549824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4618439388271336e-05,
+      "loss": 3.4785,
+      "step": 1550336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4610069819564548e-05,
+      "loss": 3.4657,
+      "step": 1550848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4601683872054028e-05,
+      "loss": 3.451,
+      "step": 1551360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4593297924543508e-05,
+      "loss": 3.4859,
+      "step": 1551872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4584911977032988e-05,
+      "loss": 3.4811,
+      "step": 1552384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4576542408326197e-05,
+      "loss": 3.4764,
+      "step": 1552896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4568156460815677e-05,
+      "loss": 3.4686,
+      "step": 1553408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4559770513305157e-05,
+      "loss": 3.4652,
+      "step": 1553920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4551384565794637e-05,
+      "loss": 3.4596,
+      "step": 1554432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4543031375891582e-05,
+      "loss": 3.4715,
+      "step": 1554944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4534645428381062e-05,
+      "loss": 3.4693,
+      "step": 1555456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4526259480870542e-05,
+      "loss": 3.4679,
+      "step": 1555968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4517873533360022e-05,
+      "loss": 3.4773,
+      "step": 1556480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.45094875858495e-05,
+      "loss": 3.4706,
+      "step": 1556992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4501101638338982e-05,
+      "loss": 3.458,
+      "step": 1557504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4492715690828462e-05,
+      "loss": 3.4815,
+      "step": 1558016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4484329743317942e-05,
+      "loss": 3.4582,
+      "step": 1558528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.447596017461115e-05,
+      "loss": 3.4597,
+      "step": 1559040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.446757422710063e-05,
+      "loss": 3.4803,
+      "step": 1559552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.445918827959011e-05,
+      "loss": 3.4804,
+      "step": 1560064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.445080233207959e-05,
+      "loss": 3.458,
+      "step": 1560576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4442432763372803e-05,
+      "loss": 3.4662,
+      "step": 1561088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4434046815862283e-05,
+      "loss": 3.4537,
+      "step": 1561600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4425660868351763e-05,
+      "loss": 3.4747,
+      "step": 1562112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4417274920841243e-05,
+      "loss": 3.4765,
+      "step": 1562624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4408905352134452e-05,
+      "loss": 3.4676,
+      "step": 1563136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4400519404623932e-05,
+      "loss": 3.4751,
+      "step": 1563648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4392133457113416e-05,
+      "loss": 3.4738,
+      "step": 1564160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4383747509602895e-05,
+      "loss": 3.49,
+      "step": 1564672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4375377940896105e-05,
+      "loss": 3.461,
+      "step": 1565184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4366991993385585e-05,
+      "loss": 3.4738,
+      "step": 1565696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4358606045875064e-05,
+      "loss": 3.4807,
+      "step": 1566208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4350220098364544e-05,
+      "loss": 3.4586,
+      "step": 1566720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4341850529657757e-05,
+      "loss": 3.476,
+      "step": 1567232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4333464582147237e-05,
+      "loss": 3.4696,
+      "step": 1567744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4325078634636717e-05,
+      "loss": 3.4809,
+      "step": 1568256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4316692687126197e-05,
+      "loss": 3.4749,
+      "step": 1568768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4308323118419406e-05,
+      "loss": 3.4613,
+      "step": 1569280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4299937170908886e-05,
+      "loss": 3.4621,
+      "step": 1569792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.429155122339837e-05,
+      "loss": 3.4647,
+      "step": 1570304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.428316527588785e-05,
+      "loss": 3.4828,
+      "step": 1570816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4274795707181058e-05,
+      "loss": 3.4688,
+      "step": 1571328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4266409759670538e-05,
+      "loss": 3.4648,
+      "step": 1571840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4258023812160018e-05,
+      "loss": 3.4675,
+      "step": 1572352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4249637864649498e-05,
+      "loss": 3.4711,
+      "step": 1572864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.424126829594271e-05,
+      "loss": 3.4622,
+      "step": 1573376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.423288234843219e-05,
+      "loss": 3.453,
+      "step": 1573888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.422449640092167e-05,
+      "loss": 3.4733,
+      "step": 1574400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.421612683221488e-05,
+      "loss": 3.4741,
+      "step": 1574912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.420774088470436e-05,
+      "loss": 3.4775,
+      "step": 1575424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.419935493719384e-05,
+      "loss": 3.4551,
+      "step": 1575936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4190968989683323e-05,
+      "loss": 3.4686,
+      "step": 1576448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4182599420976532e-05,
+      "loss": 3.464,
+      "step": 1576960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4174213473466012e-05,
+      "loss": 3.474,
+      "step": 1577472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4165827525955492e-05,
+      "loss": 3.4726,
+      "step": 1577984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4157441578444972e-05,
+      "loss": 3.4717,
+      "step": 1578496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4149072009738184e-05,
+      "loss": 3.4669,
+      "step": 1579008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4140686062227664e-05,
+      "loss": 3.4759,
+      "step": 1579520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4132316493520873e-05,
+      "loss": 3.4548,
+      "step": 1580032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4123930546010353e-05,
+      "loss": 3.4705,
+      "step": 1580544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4115544598499833e-05,
+      "loss": 3.4618,
+      "step": 1581056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4107158650989313e-05,
+      "loss": 3.4735,
+      "step": 1581568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4098772703478793e-05,
+      "loss": 3.4539,
+      "step": 1582080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4090386755968273e-05,
+      "loss": 3.4763,
+      "step": 1582592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4082000808457753e-05,
+      "loss": 3.4631,
+      "step": 1583104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4073631239750966e-05,
+      "loss": 3.4671,
+      "step": 1583616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4065245292240446e-05,
+      "loss": 3.4662,
+      "step": 1584128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4056859344729922e-05,
+      "loss": 3.4684,
+      "step": 1584640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4048473397219402e-05,
+      "loss": 3.4721,
+      "step": 1585152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4040103828512618e-05,
+      "loss": 3.4798,
+      "step": 1585664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4031717881002094e-05,
+      "loss": 3.4736,
+      "step": 1586176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4023331933491574e-05,
+      "loss": 3.4739,
+      "step": 1586688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4014945985981054e-05,
+      "loss": 3.4715,
+      "step": 1587200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4006576417274267e-05,
+      "loss": 3.4647,
+      "step": 1587712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3998190469763747e-05,
+      "loss": 3.4563,
+      "step": 1588224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3989804522253227e-05,
+      "loss": 3.4689,
+      "step": 1588736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3981418574742707e-05,
+      "loss": 3.4674,
+      "step": 1589248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.397304900603592e-05,
+      "loss": 3.4693,
+      "step": 1589760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3964663058525396e-05,
+      "loss": 3.4685,
+      "step": 1590272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3956277111014876e-05,
+      "loss": 3.4727,
+      "step": 1590784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3947891163504356e-05,
+      "loss": 3.4676,
+      "step": 1591296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3939521594797568e-05,
+      "loss": 3.467,
+      "step": 1591808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.393115202609078e-05,
+      "loss": 3.4628,
+      "step": 1592320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.392276607858026e-05,
+      "loss": 3.4766,
+      "step": 1592832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.391438013106974e-05,
+      "loss": 3.4695,
+      "step": 1593344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3905994183559217e-05,
+      "loss": 3.4622,
+      "step": 1593856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.38976082360487e-05,
+      "loss": 3.4524,
+      "step": 1594368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.388922228853818e-05,
+      "loss": 3.4733,
+      "step": 1594880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3880852719831393e-05,
+      "loss": 3.4733,
+      "step": 1595392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.387246677232087e-05,
+      "loss": 3.4726,
+      "step": 1595904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.386408082481035e-05,
+      "loss": 3.465,
+      "step": 1596416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.385569487729983e-05,
+      "loss": 3.4769,
+      "step": 1596928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.384730892978931e-05,
+      "loss": 3.4687,
+      "step": 1597440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3838922982278793e-05,
+      "loss": 3.4743,
+      "step": 1597952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3830537034768273e-05,
+      "loss": 3.4664,
+      "step": 1598464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.382215108725775e-05,
+      "loss": 3.4708,
+      "step": 1598976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3813781518550962e-05,
+      "loss": 3.4685,
+      "step": 1599488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.380541194984417e-05,
+      "loss": 3.4609,
+      "step": 1600000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3797026002333654e-05,
+      "loss": 3.475,
+      "step": 1600512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3788640054823134e-05,
+      "loss": 3.475,
+      "step": 1601024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3780254107312614e-05,
+      "loss": 3.4708,
+      "step": 1601536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3771868159802094e-05,
+      "loss": 3.4648,
+      "step": 1602048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3763482212291574e-05,
+      "loss": 3.4675,
+      "step": 1602560
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8748133182525635,
+      "eval_runtime": 305.0332,
+      "eval_samples_per_second": 1250.982,
+      "eval_steps_per_second": 39.094,
+      "step": 1602720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.375509626478105e-05,
+      "loss": 3.4685,
+      "step": 1603072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.374671031727053e-05,
+      "loss": 3.4673,
+      "step": 1603584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3738324369760014e-05,
+      "loss": 3.469,
+      "step": 1604096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3729938422249494e-05,
+      "loss": 3.4668,
+      "step": 1604608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3721568853542703e-05,
+      "loss": 3.4784,
+      "step": 1605120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3713182906032183e-05,
+      "loss": 3.4688,
+      "step": 1605632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3704796958521663e-05,
+      "loss": 3.4589,
+      "step": 1606144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3696411011011143e-05,
+      "loss": 3.4559,
+      "step": 1606656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3688041442304355e-05,
+      "loss": 3.4743,
+      "step": 1607168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3679655494793835e-05,
+      "loss": 3.4703,
+      "step": 1607680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3671269547283315e-05,
+      "loss": 3.4619,
+      "step": 1608192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3662883599772795e-05,
+      "loss": 3.4721,
+      "step": 1608704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3654530409869737e-05,
+      "loss": 3.4711,
+      "step": 1609216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3646144462359217e-05,
+      "loss": 3.4635,
+      "step": 1609728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3637758514848697e-05,
+      "loss": 3.4619,
+      "step": 1610240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3629372567338177e-05,
+      "loss": 3.4575,
+      "step": 1610752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3620986619827657e-05,
+      "loss": 3.4603,
+      "step": 1611264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3612600672317137e-05,
+      "loss": 3.4562,
+      "step": 1611776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3604214724806616e-05,
+      "loss": 3.4745,
+      "step": 1612288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3595828777296096e-05,
+      "loss": 3.4781,
+      "step": 1612800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.358745920858931e-05,
+      "loss": 3.4753,
+      "step": 1613312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.357907326107879e-05,
+      "loss": 3.4659,
+      "step": 1613824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.357068731356827e-05,
+      "loss": 3.463,
+      "step": 1614336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.356230136605775e-05,
+      "loss": 3.4669,
+      "step": 1614848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.355394817615469e-05,
+      "loss": 3.4585,
+      "step": 1615360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.354556222864417e-05,
+      "loss": 3.4655,
+      "step": 1615872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.353717628113365e-05,
+      "loss": 3.4528,
+      "step": 1616384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.352879033362313e-05,
+      "loss": 3.4706,
+      "step": 1616896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.352040438611261e-05,
+      "loss": 3.4569,
+      "step": 1617408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.351201843860209e-05,
+      "loss": 3.4533,
+      "step": 1617920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.350363249109157e-05,
+      "loss": 3.4638,
+      "step": 1618432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.349524654358105e-05,
+      "loss": 3.4688,
+      "step": 1618944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3486876974874263e-05,
+      "loss": 3.4748,
+      "step": 1619456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3478491027363743e-05,
+      "loss": 3.468,
+      "step": 1619968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3470105079853223e-05,
+      "loss": 3.4573,
+      "step": 1620480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.34617191323427e-05,
+      "loss": 3.4726,
+      "step": 1620992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.345334956363591e-05,
+      "loss": 3.467,
+      "step": 1621504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.344496361612539e-05,
+      "loss": 3.4516,
+      "step": 1622016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.343657766861487e-05,
+      "loss": 3.4546,
+      "step": 1622528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.342819172110435e-05,
+      "loss": 3.4513,
+      "step": 1623040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3419822152397564e-05,
+      "loss": 3.4605,
+      "step": 1623552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3411436204887044e-05,
+      "loss": 3.4704,
+      "step": 1624064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3403050257376524e-05,
+      "loss": 3.4655,
+      "step": 1624576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3394664309866e-05,
+      "loss": 3.4555,
+      "step": 1625088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3386294741159216e-05,
+      "loss": 3.468,
+      "step": 1625600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3377908793648696e-05,
+      "loss": 3.4594,
+      "step": 1626112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3369522846138173e-05,
+      "loss": 3.4716,
+      "step": 1626624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3361153277431385e-05,
+      "loss": 3.4541,
+      "step": 1627136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3352767329920865e-05,
+      "loss": 3.4365,
+      "step": 1627648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3344381382410345e-05,
+      "loss": 3.4756,
+      "step": 1628160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3335995434899825e-05,
+      "loss": 3.4639,
+      "step": 1628672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3327625866193038e-05,
+      "loss": 3.4668,
+      "step": 1629184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3319239918682518e-05,
+      "loss": 3.459,
+      "step": 1629696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3310853971171998e-05,
+      "loss": 3.4547,
+      "step": 1630208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3302468023661474e-05,
+      "loss": 3.4487,
+      "step": 1630720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3294098454954687e-05,
+      "loss": 3.4588,
+      "step": 1631232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.328571250744417e-05,
+      "loss": 3.4577,
+      "step": 1631744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3277326559933646e-05,
+      "loss": 3.457,
+      "step": 1632256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3268940612423126e-05,
+      "loss": 3.4665,
+      "step": 1632768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3260554664912606e-05,
+      "loss": 3.4575,
+      "step": 1633280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3252168717402086e-05,
+      "loss": 3.4492,
+      "step": 1633792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3243782769891566e-05,
+      "loss": 3.4679,
+      "step": 1634304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3235396822381046e-05,
+      "loss": 3.4505,
+      "step": 1634816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.322702725367426e-05,
+      "loss": 3.4488,
+      "step": 1635328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.321864130616374e-05,
+      "loss": 3.467,
+      "step": 1635840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.321025535865322e-05,
+      "loss": 3.4689,
+      "step": 1636352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.32018694111427e-05,
+      "loss": 3.446,
+      "step": 1636864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3193499842435908e-05,
+      "loss": 3.4565,
+      "step": 1637376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3185113894925388e-05,
+      "loss": 3.4432,
+      "step": 1637888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.317672794741487e-05,
+      "loss": 3.4596,
+      "step": 1638400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.316834199990435e-05,
+      "loss": 3.4637,
+      "step": 1638912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.315997243119756e-05,
+      "loss": 3.4595,
+      "step": 1639424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.315160286249077e-05,
+      "loss": 3.462,
+      "step": 1639936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.314321691498025e-05,
+      "loss": 3.4618,
+      "step": 1640448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3134830967469732e-05,
+      "loss": 3.4743,
+      "step": 1640960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3126445019959212e-05,
+      "loss": 3.4518,
+      "step": 1641472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3118059072448692e-05,
+      "loss": 3.4656,
+      "step": 1641984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3109673124938172e-05,
+      "loss": 3.4679,
+      "step": 1642496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.310130355623138e-05,
+      "loss": 3.4459,
+      "step": 1643008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.309291760872086e-05,
+      "loss": 3.4651,
+      "step": 1643520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.308453166121034e-05,
+      "loss": 3.4566,
+      "step": 1644032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3076145713699825e-05,
+      "loss": 3.4721,
+      "step": 1644544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.30677597661893e-05,
+      "loss": 3.4645,
+      "step": 1645056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.305937381867878e-05,
+      "loss": 3.4523,
+      "step": 1645568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.305098787116826e-05,
+      "loss": 3.4482,
+      "step": 1646080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.304260192365774e-05,
+      "loss": 3.4541,
+      "step": 1646592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3034232354950954e-05,
+      "loss": 3.472,
+      "step": 1647104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3025846407440434e-05,
+      "loss": 3.4572,
+      "step": 1647616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3017460459929914e-05,
+      "loss": 3.4551,
+      "step": 1648128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3009074512419393e-05,
+      "loss": 3.4571,
+      "step": 1648640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3000704943712603e-05,
+      "loss": 3.4591,
+      "step": 1649152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2992318996202083e-05,
+      "loss": 3.4509,
+      "step": 1649664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2983933048691562e-05,
+      "loss": 3.4407,
+      "step": 1650176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2975547101181046e-05,
+      "loss": 3.4589,
+      "step": 1650688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2967161153670526e-05,
+      "loss": 3.4643,
+      "step": 1651200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2958775206160006e-05,
+      "loss": 3.465,
+      "step": 1651712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2950389258649482e-05,
+      "loss": 3.447,
+      "step": 1652224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2942003311138962e-05,
+      "loss": 3.4554,
+      "step": 1652736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2933633742432175e-05,
+      "loss": 3.4548,
+      "step": 1653248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2925247794921655e-05,
+      "loss": 3.4609,
+      "step": 1653760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2916878226214867e-05,
+      "loss": 3.4661,
+      "step": 1654272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2908492278704347e-05,
+      "loss": 3.4553,
+      "step": 1654784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2900106331193827e-05,
+      "loss": 3.4597,
+      "step": 1655296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2891720383683304e-05,
+      "loss": 3.4616,
+      "step": 1655808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2883334436172784e-05,
+      "loss": 3.4464,
+      "step": 1656320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2874964867465996e-05,
+      "loss": 3.4579,
+      "step": 1656832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.286657891995548e-05,
+      "loss": 3.4488,
+      "step": 1657344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2858192972444956e-05,
+      "loss": 3.4622,
+      "step": 1657856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2849807024934436e-05,
+      "loss": 3.4438,
+      "step": 1658368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2841421077423916e-05,
+      "loss": 3.4614,
+      "step": 1658880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2833035129913396e-05,
+      "loss": 3.4519,
+      "step": 1659392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.282466556120661e-05,
+      "loss": 3.4561,
+      "step": 1659904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.281627961369609e-05,
+      "loss": 3.4536,
+      "step": 1660416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2807893666185568e-05,
+      "loss": 3.4595,
+      "step": 1660928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2799507718675048e-05,
+      "loss": 3.4587,
+      "step": 1661440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2791138149968257e-05,
+      "loss": 3.4718,
+      "step": 1661952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2782752202457737e-05,
+      "loss": 3.461,
+      "step": 1662464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2774366254947217e-05,
+      "loss": 3.4649,
+      "step": 1662976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.27659803074367e-05,
+      "loss": 3.4544,
+      "step": 1663488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.275759435992618e-05,
+      "loss": 3.4595,
+      "step": 1664000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.274922479121939e-05,
+      "loss": 3.4483,
+      "step": 1664512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.274083884370887e-05,
+      "loss": 3.4538,
+      "step": 1665024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.273245289619835e-05,
+      "loss": 3.4561,
+      "step": 1665536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.272406694868783e-05,
+      "loss": 3.4568,
+      "step": 1666048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.271568100117731e-05,
+      "loss": 3.4586,
+      "step": 1666560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2707311432470522e-05,
+      "loss": 3.4557,
+      "step": 1667072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2698925484960002e-05,
+      "loss": 3.4573,
+      "step": 1667584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2690539537449482e-05,
+      "loss": 3.4573,
+      "step": 1668096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.268215358993896e-05,
+      "loss": 3.454,
+      "step": 1668608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.267376764242844e-05,
+      "loss": 3.4613,
+      "step": 1669120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.266538169491792e-05,
+      "loss": 3.4567,
+      "step": 1669632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.26569957474074e-05,
+      "loss": 3.45,
+      "step": 1670144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.264862617870061e-05,
+      "loss": 3.4394,
+      "step": 1670656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.264024023119009e-05,
+      "loss": 3.465,
+      "step": 1671168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.263185428367957e-05,
+      "loss": 3.4615,
+      "step": 1671680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.262346833616905e-05,
+      "loss": 3.4622,
+      "step": 1672192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2615098767462263e-05,
+      "loss": 3.4534,
+      "step": 1672704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2606712819951743e-05,
+      "loss": 3.4617,
+      "step": 1673216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2598326872441223e-05,
+      "loss": 3.4547,
+      "step": 1673728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2589940924930703e-05,
+      "loss": 3.4701,
+      "step": 1674240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2581554977420183e-05,
+      "loss": 3.4541,
+      "step": 1674752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2573169029909663e-05,
+      "loss": 3.4558,
+      "step": 1675264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2564799461202872e-05,
+      "loss": 3.4587,
+      "step": 1675776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2556413513692355e-05,
+      "loss": 3.452,
+      "step": 1676288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2548027566181835e-05,
+      "loss": 3.4655,
+      "step": 1676800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2539641618671315e-05,
+      "loss": 3.4596,
+      "step": 1677312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2531255671160792e-05,
+      "loss": 3.4648,
+      "step": 1677824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2522886102454004e-05,
+      "loss": 3.45,
+      "step": 1678336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2514500154943484e-05,
+      "loss": 3.4571,
+      "step": 1678848
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.876386880874634,
+      "eval_runtime": 305.8818,
+      "eval_samples_per_second": 1247.511,
+      "eval_steps_per_second": 38.986,
+      "step": 1679040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2506130586236697e-05,
+      "loss": 3.4598,
+      "step": 1679360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2497744638726177e-05,
+      "loss": 3.4559,
+      "step": 1679872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2489358691215657e-05,
+      "loss": 3.4568,
+      "step": 1680384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2480972743705137e-05,
+      "loss": 3.4557,
+      "step": 1680896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2472586796194613e-05,
+      "loss": 3.469,
+      "step": 1681408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2464200848684093e-05,
+      "loss": 3.4575,
+      "step": 1681920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2455814901173573e-05,
+      "loss": 3.4492,
+      "step": 1682432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2447428953663056e-05,
+      "loss": 3.4428,
+      "step": 1682944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2439059384956266e-05,
+      "loss": 3.4651,
+      "step": 1683456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2430673437445745e-05,
+      "loss": 3.4584,
+      "step": 1683968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2422287489935225e-05,
+      "loss": 3.4525,
+      "step": 1684480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2413901542424705e-05,
+      "loss": 3.4588,
+      "step": 1684992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2405531973717918e-05,
+      "loss": 3.4583,
+      "step": 1685504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2397146026207398e-05,
+      "loss": 3.4504,
+      "step": 1686016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2388760078696878e-05,
+      "loss": 3.4478,
+      "step": 1686528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2380374131186358e-05,
+      "loss": 3.4483,
+      "step": 1687040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2372004562479567e-05,
+      "loss": 3.4472,
+      "step": 1687552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2363618614969047e-05,
+      "loss": 3.449,
+      "step": 1688064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2355232667458527e-05,
+      "loss": 3.4614,
+      "step": 1688576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.234684671994801e-05,
+      "loss": 3.4655,
+      "step": 1689088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.233846077243749e-05,
+      "loss": 3.4631,
+      "step": 1689600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.233007482492697e-05,
+      "loss": 3.4564,
+      "step": 1690112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.232170525622018e-05,
+      "loss": 3.453,
+      "step": 1690624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.231331930870966e-05,
+      "loss": 3.4538,
+      "step": 1691136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.230493336119914e-05,
+      "loss": 3.4512,
+      "step": 1691648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.229654741368862e-05,
+      "loss": 3.4552,
+      "step": 1692160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.228817784498183e-05,
+      "loss": 3.4385,
+      "step": 1692672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.227979189747131e-05,
+      "loss": 3.4597,
+      "step": 1693184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.227140594996079e-05,
+      "loss": 3.4491,
+      "step": 1693696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2263020002450268e-05,
+      "loss": 3.4428,
+      "step": 1694208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2254634054939748e-05,
+      "loss": 3.4519,
+      "step": 1694720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2246248107429228e-05,
+      "loss": 3.46,
+      "step": 1695232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.223786215991871e-05,
+      "loss": 3.4565,
+      "step": 1695744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.222947621240819e-05,
+      "loss": 3.4581,
+      "step": 1696256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.22211066437014e-05,
+      "loss": 3.4465,
+      "step": 1696768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.221272069619088e-05,
+      "loss": 3.4613,
+      "step": 1697280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.220433474868036e-05,
+      "loss": 3.4543,
+      "step": 1697792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2195965179973573e-05,
+      "loss": 3.4409,
+      "step": 1698304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2187579232463053e-05,
+      "loss": 3.4433,
+      "step": 1698816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2179193284952533e-05,
+      "loss": 3.4413,
+      "step": 1699328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2170807337442013e-05,
+      "loss": 3.4501,
+      "step": 1699840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2162421389931492e-05,
+      "loss": 3.4606,
+      "step": 1700352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.21540518212247e-05,
+      "loss": 3.4512,
+      "step": 1700864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.214566587371418e-05,
+      "loss": 3.4484,
+      "step": 1701376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2137279926203665e-05,
+      "loss": 3.4547,
+      "step": 1701888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2128893978693145e-05,
+      "loss": 3.4482,
+      "step": 1702400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2120508031182625e-05,
+      "loss": 3.4613,
+      "step": 1702912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.21121220836721e-05,
+      "loss": 3.4464,
+      "step": 1703424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.210373613616158e-05,
+      "loss": 3.4218,
+      "step": 1703936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2095366567454794e-05,
+      "loss": 3.4662,
+      "step": 1704448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2086980619944274e-05,
+      "loss": 3.4545,
+      "step": 1704960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2078594672433754e-05,
+      "loss": 3.457,
+      "step": 1705472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2070208724923234e-05,
+      "loss": 3.4461,
+      "step": 1705984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2061822777412714e-05,
+      "loss": 3.4445,
+      "step": 1706496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2053453208705923e-05,
+      "loss": 3.4379,
+      "step": 1707008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2045067261195403e-05,
+      "loss": 3.4508,
+      "step": 1707520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2036681313684886e-05,
+      "loss": 3.4457,
+      "step": 1708032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2028295366174366e-05,
+      "loss": 3.4466,
+      "step": 1708544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2019909418663846e-05,
+      "loss": 3.4519,
+      "step": 1709056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2011523471153326e-05,
+      "loss": 3.449,
+      "step": 1709568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2003137523642806e-05,
+      "loss": 3.4363,
+      "step": 1710080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1994751576132282e-05,
+      "loss": 3.4561,
+      "step": 1710592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1986382007425495e-05,
+      "loss": 3.4393,
+      "step": 1711104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1977996059914975e-05,
+      "loss": 3.4406,
+      "step": 1711616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1969610112404455e-05,
+      "loss": 3.455,
+      "step": 1712128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1961224164893935e-05,
+      "loss": 3.4546,
+      "step": 1712640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1952854596187147e-05,
+      "loss": 3.4344,
+      "step": 1713152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1944468648676627e-05,
+      "loss": 3.4434,
+      "step": 1713664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1936082701166104e-05,
+      "loss": 3.4357,
+      "step": 1714176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1927696753655587e-05,
+      "loss": 3.4498,
+      "step": 1714688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.19193271849488e-05,
+      "loss": 3.4524,
+      "step": 1715200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.191094123743828e-05,
+      "loss": 3.4507,
+      "step": 1715712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.190257166873149e-05,
+      "loss": 3.4504,
+      "step": 1716224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.189418572122097e-05,
+      "loss": 3.4487,
+      "step": 1716736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.188579977371045e-05,
+      "loss": 3.4647,
+      "step": 1717248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.187741382619993e-05,
+      "loss": 3.442,
+      "step": 1717760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.186902787868941e-05,
+      "loss": 3.4526,
+      "step": 1718272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.186064193117889e-05,
+      "loss": 3.4559,
+      "step": 1718784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.18522723624721e-05,
+      "loss": 3.4357,
+      "step": 1719296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1843886414961577e-05,
+      "loss": 3.4509,
+      "step": 1719808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1835500467451057e-05,
+      "loss": 3.4521,
+      "step": 1720320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.182711451994054e-05,
+      "loss": 3.4532,
+      "step": 1720832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.181872857243002e-05,
+      "loss": 3.4567,
+      "step": 1721344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.18103426249195e-05,
+      "loss": 3.4407,
+      "step": 1721856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.180195667740898e-05,
+      "loss": 3.4387,
+      "step": 1722368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.179357072989846e-05,
+      "loss": 3.443,
+      "step": 1722880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.178520116119167e-05,
+      "loss": 3.4544,
+      "step": 1723392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.177681521368115e-05,
+      "loss": 3.4464,
+      "step": 1723904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.176842926617063e-05,
+      "loss": 3.4443,
+      "step": 1724416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.176004331866011e-05,
+      "loss": 3.4475,
+      "step": 1724928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1751673749953322e-05,
+      "loss": 3.4505,
+      "step": 1725440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1743287802442802e-05,
+      "loss": 3.4409,
+      "step": 1725952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1734901854932282e-05,
+      "loss": 3.4295,
+      "step": 1726464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.172651590742176e-05,
+      "loss": 3.449,
+      "step": 1726976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1718129959911242e-05,
+      "loss": 3.4491,
+      "step": 1727488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1709760391204454e-05,
+      "loss": 3.4536,
+      "step": 1728000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1701374443693934e-05,
+      "loss": 3.4371,
+      "step": 1728512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.169298849618341e-05,
+      "loss": 3.4464,
+      "step": 1729024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.168460254867289e-05,
+      "loss": 3.445,
+      "step": 1729536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.167621660116237e-05,
+      "loss": 3.4475,
+      "step": 1730048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1667847032455583e-05,
+      "loss": 3.4544,
+      "step": 1730560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1659461084945063e-05,
+      "loss": 3.4471,
+      "step": 1731072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1651075137434543e-05,
+      "loss": 3.4461,
+      "step": 1731584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1642689189924023e-05,
+      "loss": 3.4483,
+      "step": 1732096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1634303242413503e-05,
+      "loss": 3.436,
+      "step": 1732608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1625933673706712e-05,
+      "loss": 3.4469,
+      "step": 1733120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1617547726196196e-05,
+      "loss": 3.4327,
+      "step": 1733632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1609161778685675e-05,
+      "loss": 3.4578,
+      "step": 1734144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1600775831175155e-05,
+      "loss": 3.4324,
+      "step": 1734656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1592389883664635e-05,
+      "loss": 3.449,
+      "step": 1735168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1584020314957844e-05,
+      "loss": 3.4407,
+      "step": 1735680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1575634367447324e-05,
+      "loss": 3.4453,
+      "step": 1736192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1567248419936804e-05,
+      "loss": 3.4432,
+      "step": 1736704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1558862472426288e-05,
+      "loss": 3.449,
+      "step": 1737216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1550476524915764e-05,
+      "loss": 3.4451,
+      "step": 1737728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1542106956208977e-05,
+      "loss": 3.4566,
+      "step": 1738240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1533721008698457e-05,
+      "loss": 3.4474,
+      "step": 1738752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1525335061187937e-05,
+      "loss": 3.4579,
+      "step": 1739264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1516949113677413e-05,
+      "loss": 3.4463,
+      "step": 1739776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1508563166166897e-05,
+      "loss": 3.4455,
+      "step": 1740288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.150019359746011e-05,
+      "loss": 3.442,
+      "step": 1740800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1491807649949586e-05,
+      "loss": 3.4395,
+      "step": 1741312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1483421702439066e-05,
+      "loss": 3.4431,
+      "step": 1741824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1475035754928546e-05,
+      "loss": 3.4468,
+      "step": 1742336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1466666186221758e-05,
+      "loss": 3.4436,
+      "step": 1742848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1458280238711238e-05,
+      "loss": 3.448,
+      "step": 1743360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1449894291200718e-05,
+      "loss": 3.4479,
+      "step": 1743872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1441508343690198e-05,
+      "loss": 3.4478,
+      "step": 1744384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.143313877498341e-05,
+      "loss": 3.4475,
+      "step": 1744896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1424752827472887e-05,
+      "loss": 3.4442,
+      "step": 1745408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1416366879962367e-05,
+      "loss": 3.443,
+      "step": 1745920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.140798093245185e-05,
+      "loss": 3.4438,
+      "step": 1746432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.139959498494133e-05,
+      "loss": 3.4278,
+      "step": 1746944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.139120903743081e-05,
+      "loss": 3.4521,
+      "step": 1747456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.138283946872402e-05,
+      "loss": 3.4492,
+      "step": 1747968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.13744535212135e-05,
+      "loss": 3.4519,
+      "step": 1748480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.136606757370298e-05,
+      "loss": 3.4392,
+      "step": 1748992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.135768162619246e-05,
+      "loss": 3.4533,
+      "step": 1749504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.134931205748567e-05,
+      "loss": 3.4495,
+      "step": 1750016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.134092610997515e-05,
+      "loss": 3.4558,
+      "step": 1750528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.133254016246463e-05,
+      "loss": 3.4394,
+      "step": 1751040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.132415421495411e-05,
+      "loss": 3.4405,
+      "step": 1751552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.131576826744359e-05,
+      "loss": 3.4485,
+      "step": 1752064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1307398698736804e-05,
+      "loss": 3.4447,
+      "step": 1752576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1299012751226284e-05,
+      "loss": 3.4494,
+      "step": 1753088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1290626803715764e-05,
+      "loss": 3.4517,
+      "step": 1753600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.128224085620524e-05,
+      "loss": 3.4522,
+      "step": 1754112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.127385490869472e-05,
+      "loss": 3.4379,
+      "step": 1754624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.12654689611842e-05,
+      "loss": 3.4518,
+      "step": 1755136
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8769567012786865,
+      "eval_runtime": 307.414,
+      "eval_samples_per_second": 1241.294,
+      "eval_steps_per_second": 38.791,
+      "step": 1755360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1257099392477413e-05,
+      "loss": 3.4398,
+      "step": 1755648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1248713444966893e-05,
+      "loss": 3.4486,
+      "step": 1756160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1240327497456373e-05,
+      "loss": 3.4424,
+      "step": 1756672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1231941549945853e-05,
+      "loss": 3.4424,
+      "step": 1757184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1223571981239065e-05,
+      "loss": 3.4554,
+      "step": 1757696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1215186033728542e-05,
+      "loss": 3.4478,
+      "step": 1758208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1206800086218022e-05,
+      "loss": 3.4385,
+      "step": 1758720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1198414138707505e-05,
+      "loss": 3.4367,
+      "step": 1759232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1190028191196985e-05,
+      "loss": 3.4466,
+      "step": 1759744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1181658622490194e-05,
+      "loss": 3.4506,
+      "step": 1760256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1173272674979674e-05,
+      "loss": 3.4392,
+      "step": 1760768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1164886727469154e-05,
+      "loss": 3.4487,
+      "step": 1761280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1156500779958634e-05,
+      "loss": 3.4519,
+      "step": 1761792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1148114832448114e-05,
+      "loss": 3.4378,
+      "step": 1762304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1139728884937597e-05,
+      "loss": 3.4408,
+      "step": 1762816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1131342937427074e-05,
+      "loss": 3.4364,
+      "step": 1763328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1122973368720286e-05,
+      "loss": 3.4361,
+      "step": 1763840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1114587421209766e-05,
+      "loss": 3.4356,
+      "step": 1764352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1106201473699246e-05,
+      "loss": 3.4458,
+      "step": 1764864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1097815526188723e-05,
+      "loss": 3.458,
+      "step": 1765376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1089429578678206e-05,
+      "loss": 3.4546,
+      "step": 1765888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.108106000997142e-05,
+      "loss": 3.4451,
+      "step": 1766400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1072674062460895e-05,
+      "loss": 3.439,
+      "step": 1766912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1064288114950375e-05,
+      "loss": 3.4417,
+      "step": 1767424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1055902167439855e-05,
+      "loss": 3.4406,
+      "step": 1767936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1047516219929335e-05,
+      "loss": 3.4427,
+      "step": 1768448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1039146651222548e-05,
+      "loss": 3.4275,
+      "step": 1768960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1030760703712027e-05,
+      "loss": 3.4501,
+      "step": 1769472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1022374756201507e-05,
+      "loss": 3.4385,
+      "step": 1769984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1013988808690987e-05,
+      "loss": 3.4328,
+      "step": 1770496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1005602861180467e-05,
+      "loss": 3.434,
+      "step": 1771008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0997233292473676e-05,
+      "loss": 3.4505,
+      "step": 1771520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.098884734496316e-05,
+      "loss": 3.4483,
+      "step": 1772032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.098046139745264e-05,
+      "loss": 3.448,
+      "step": 1772544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.097207544994212e-05,
+      "loss": 3.4351,
+      "step": 1773056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.09636895024316e-05,
+      "loss": 3.4461,
+      "step": 1773568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0955303554921076e-05,
+      "loss": 3.4467,
+      "step": 1774080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0946917607410556e-05,
+      "loss": 3.4322,
+      "step": 1774592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0938531659900036e-05,
+      "loss": 3.4319,
+      "step": 1775104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.093017846999698e-05,
+      "loss": 3.4294,
+      "step": 1775616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.092179252248646e-05,
+      "loss": 3.4388,
+      "step": 1776128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.091340657497594e-05,
+      "loss": 3.4457,
+      "step": 1776640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.090502062746542e-05,
+      "loss": 3.4429,
+      "step": 1777152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0896634679954898e-05,
+      "loss": 3.4368,
+      "step": 1777664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0888265111248113e-05,
+      "loss": 3.4445,
+      "step": 1778176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0879879163737593e-05,
+      "loss": 3.4368,
+      "step": 1778688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0871493216227073e-05,
+      "loss": 3.4461,
+      "step": 1779200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.086310726871655e-05,
+      "loss": 3.4387,
+      "step": 1779712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.085472132120603e-05,
+      "loss": 3.4111,
+      "step": 1780224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0846351752499242e-05,
+      "loss": 3.4504,
+      "step": 1780736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0837965804988722e-05,
+      "loss": 3.4445,
+      "step": 1781248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0829579857478202e-05,
+      "loss": 3.4476,
+      "step": 1781760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0821193909967682e-05,
+      "loss": 3.4339,
+      "step": 1782272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0812807962457162e-05,
+      "loss": 3.4314,
+      "step": 1782784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.080443839375037e-05,
+      "loss": 3.4287,
+      "step": 1783296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.079605244623985e-05,
+      "loss": 3.4378,
+      "step": 1783808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0787666498729335e-05,
+      "loss": 3.4386,
+      "step": 1784320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0779280551218815e-05,
+      "loss": 3.4314,
+      "step": 1784832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0770894603708295e-05,
+      "loss": 3.4423,
+      "step": 1785344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0762525035001504e-05,
+      "loss": 3.4397,
+      "step": 1785856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0754139087490984e-05,
+      "loss": 3.4241,
+      "step": 1786368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0745753139980464e-05,
+      "loss": 3.4455,
+      "step": 1786880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0737367192469943e-05,
+      "loss": 3.4294,
+      "step": 1787392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0728997623763156e-05,
+      "loss": 3.4253,
+      "step": 1787904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0720611676252636e-05,
+      "loss": 3.4414,
+      "step": 1788416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0712225728742116e-05,
+      "loss": 3.4454,
+      "step": 1788928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0703839781231596e-05,
+      "loss": 3.4244,
+      "step": 1789440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0695470212524805e-05,
+      "loss": 3.4346,
+      "step": 1789952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0687084265014285e-05,
+      "loss": 3.4219,
+      "step": 1790464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0678698317503768e-05,
+      "loss": 3.4382,
+      "step": 1790976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0670312369993248e-05,
+      "loss": 3.4409,
+      "step": 1791488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0661942801286457e-05,
+      "loss": 3.4376,
+      "step": 1792000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0653556853775937e-05,
+      "loss": 3.4376,
+      "step": 1792512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0645170906265417e-05,
+      "loss": 3.4394,
+      "step": 1793024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0636784958754897e-05,
+      "loss": 3.4552,
+      "step": 1793536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0628399011244377e-05,
+      "loss": 3.4294,
+      "step": 1794048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0620013063733857e-05,
+      "loss": 3.4441,
+      "step": 1794560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0611627116223337e-05,
+      "loss": 3.444,
+      "step": 1795072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.060325754751655e-05,
+      "loss": 3.4254,
+      "step": 1795584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0594871600006026e-05,
+      "loss": 3.4404,
+      "step": 1796096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0586485652495506e-05,
+      "loss": 3.4421,
+      "step": 1796608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.057809970498499e-05,
+      "loss": 3.4409,
+      "step": 1797120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0569730136278202e-05,
+      "loss": 3.4461,
+      "step": 1797632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.056134418876768e-05,
+      "loss": 3.4242,
+      "step": 1798144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.055295824125716e-05,
+      "loss": 3.4278,
+      "step": 1798656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.054457229374664e-05,
+      "loss": 3.4307,
+      "step": 1799168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.053620272503985e-05,
+      "loss": 3.4431,
+      "step": 1799680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.052781677752933e-05,
+      "loss": 3.4357,
+      "step": 1800192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.051943083001881e-05,
+      "loss": 3.4373,
+      "step": 1800704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.051104488250829e-05,
+      "loss": 3.4387,
+      "step": 1801216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.05026753138015e-05,
+      "loss": 3.4338,
+      "step": 1801728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.049428936629098e-05,
+      "loss": 3.4324,
+      "step": 1802240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.048590341878046e-05,
+      "loss": 3.4183,
+      "step": 1802752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0477517471269943e-05,
+      "loss": 3.4356,
+      "step": 1803264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0469147902563152e-05,
+      "loss": 3.4393,
+      "step": 1803776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0460761955052632e-05,
+      "loss": 3.444,
+      "step": 1804288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0452376007542112e-05,
+      "loss": 3.4263,
+      "step": 1804800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0443990060031592e-05,
+      "loss": 3.4328,
+      "step": 1805312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0435604112521072e-05,
+      "loss": 3.4329,
+      "step": 1805824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0427218165010552e-05,
+      "loss": 3.4349,
+      "step": 1806336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0418848596303764e-05,
+      "loss": 3.4411,
+      "step": 1806848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0410462648793244e-05,
+      "loss": 3.4388,
+      "step": 1807360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0402076701282724e-05,
+      "loss": 3.4371,
+      "step": 1807872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0393690753772204e-05,
+      "loss": 3.4355,
+      "step": 1808384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.038530480626168e-05,
+      "loss": 3.4268,
+      "step": 1808896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0376935237554893e-05,
+      "loss": 3.4342,
+      "step": 1809408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0368549290044377e-05,
+      "loss": 3.4257,
+      "step": 1809920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0360163342533857e-05,
+      "loss": 3.4456,
+      "step": 1810432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0351777395023333e-05,
+      "loss": 3.423,
+      "step": 1810944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0343391447512813e-05,
+      "loss": 3.4368,
+      "step": 1811456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0335021878806026e-05,
+      "loss": 3.4299,
+      "step": 1811968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0326635931295506e-05,
+      "loss": 3.4387,
+      "step": 1812480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0318249983784986e-05,
+      "loss": 3.431,
+      "step": 1812992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0309864036274465e-05,
+      "loss": 3.4357,
+      "step": 1813504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0301478088763945e-05,
+      "loss": 3.4332,
+      "step": 1814016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0293108520057155e-05,
+      "loss": 3.449,
+      "step": 1814528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0284722572546634e-05,
+      "loss": 3.4338,
+      "step": 1815040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0276336625036114e-05,
+      "loss": 3.4478,
+      "step": 1815552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0267950677525598e-05,
+      "loss": 3.4327,
+      "step": 1816064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0259581108818807e-05,
+      "loss": 3.435,
+      "step": 1816576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0251195161308287e-05,
+      "loss": 3.4305,
+      "step": 1817088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0242809213797767e-05,
+      "loss": 3.4324,
+      "step": 1817600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0234423266287247e-05,
+      "loss": 3.4291,
+      "step": 1818112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0226037318776727e-05,
+      "loss": 3.438,
+      "step": 1818624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.021766775006994e-05,
+      "loss": 3.4313,
+      "step": 1819136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.020928180255942e-05,
+      "loss": 3.4373,
+      "step": 1819648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.02008958550489e-05,
+      "loss": 3.4387,
+      "step": 1820160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.019250990753838e-05,
+      "loss": 3.4332,
+      "step": 1820672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0184140338831588e-05,
+      "loss": 3.436,
+      "step": 1821184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0175754391321068e-05,
+      "loss": 3.4377,
+      "step": 1821696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.016736844381055e-05,
+      "loss": 3.428,
+      "step": 1822208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.015898249630003e-05,
+      "loss": 3.4358,
+      "step": 1822720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.015059654878951e-05,
+      "loss": 3.4155,
+      "step": 1823232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.014222698008272e-05,
+      "loss": 3.4418,
+      "step": 1823744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.01338410325722e-05,
+      "loss": 3.4378,
+      "step": 1824256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.012545508506168e-05,
+      "loss": 3.4408,
+      "step": 1824768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.011706913755116e-05,
+      "loss": 3.4282,
+      "step": 1825280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0108699568844373e-05,
+      "loss": 3.4393,
+      "step": 1825792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0100313621333853e-05,
+      "loss": 3.4365,
+      "step": 1826304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0091927673823333e-05,
+      "loss": 3.4439,
+      "step": 1826816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.008354172631281e-05,
+      "loss": 3.4305,
+      "step": 1827328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.007515577880229e-05,
+      "loss": 3.4325,
+      "step": 1827840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0066786210095502e-05,
+      "loss": 3.4351,
+      "step": 1828352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0058400262584985e-05,
+      "loss": 3.437,
+      "step": 1828864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.005001431507446e-05,
+      "loss": 3.4375,
+      "step": 1829376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.004162836756394e-05,
+      "loss": 3.4348,
+      "step": 1829888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.003324242005342e-05,
+      "loss": 3.443,
+      "step": 1830400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.002487285134663e-05,
+      "loss": 3.4266,
+      "step": 1830912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0016486903836114e-05,
+      "loss": 3.443,
+      "step": 1831424
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8784711360931396,
+      "eval_runtime": 306.0472,
+      "eval_samples_per_second": 1246.837,
+      "eval_steps_per_second": 38.965,
+      "step": 1831680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0008117335129326e-05,
+      "loss": 3.4366,
+      "step": 1831936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9999731387618806e-05,
+      "loss": 3.433,
+      "step": 1832448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9991345440108283e-05,
+      "loss": 3.4351,
+      "step": 1832960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9982959492597763e-05,
+      "loss": 3.4332,
+      "step": 1833472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9974573545087243e-05,
+      "loss": 3.4472,
+      "step": 1833984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9966203976380455e-05,
+      "loss": 3.4359,
+      "step": 1834496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9957818028869935e-05,
+      "loss": 3.4266,
+      "step": 1835008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9949432081359415e-05,
+      "loss": 3.4245,
+      "step": 1835520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9941046133848895e-05,
+      "loss": 3.4366,
+      "step": 1836032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9932676565142104e-05,
+      "loss": 3.4424,
+      "step": 1836544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9924290617631584e-05,
+      "loss": 3.4308,
+      "step": 1837056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9915904670121068e-05,
+      "loss": 3.4334,
+      "step": 1837568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9907518722610548e-05,
+      "loss": 3.4394,
+      "step": 1838080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9899132775100028e-05,
+      "loss": 3.4265,
+      "step": 1838592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9890763206393237e-05,
+      "loss": 3.4287,
+      "step": 1839104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9882377258882717e-05,
+      "loss": 3.4229,
+      "step": 1839616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9873991311372197e-05,
+      "loss": 3.4285,
+      "step": 1840128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9865605363861677e-05,
+      "loss": 3.4234,
+      "step": 1840640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.985721941635116e-05,
+      "loss": 3.4356,
+      "step": 1841152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.984884984764437e-05,
+      "loss": 3.4457,
+      "step": 1841664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.984046390013385e-05,
+      "loss": 3.4432,
+      "step": 1842176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.983207795262333e-05,
+      "loss": 3.4342,
+      "step": 1842688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.982369200511281e-05,
+      "loss": 3.4294,
+      "step": 1843200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.981532243640602e-05,
+      "loss": 3.4306,
+      "step": 1843712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.98069364888955e-05,
+      "loss": 3.4297,
+      "step": 1844224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.979855054138498e-05,
+      "loss": 3.4324,
+      "step": 1844736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.979016459387446e-05,
+      "loss": 3.4192,
+      "step": 1845248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9781778646363938e-05,
+      "loss": 3.4398,
+      "step": 1845760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.977340907765715e-05,
+      "loss": 3.4289,
+      "step": 1846272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.976502313014663e-05,
+      "loss": 3.42,
+      "step": 1846784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.975663718263611e-05,
+      "loss": 3.4197,
+      "step": 1847296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.974825123512559e-05,
+      "loss": 3.4406,
+      "step": 1847808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.973986528761507e-05,
+      "loss": 3.4392,
+      "step": 1848320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9731495718908283e-05,
+      "loss": 3.4353,
+      "step": 1848832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.972310977139776e-05,
+      "loss": 3.4268,
+      "step": 1849344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.971472382388724e-05,
+      "loss": 3.4282,
+      "step": 1849856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9706337876376722e-05,
+      "loss": 3.44,
+      "step": 1850368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9697968307669935e-05,
+      "loss": 3.4199,
+      "step": 1850880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.968958236015941e-05,
+      "loss": 3.4179,
+      "step": 1851392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.968119641264889e-05,
+      "loss": 3.4218,
+      "step": 1851904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.967281046513837e-05,
+      "loss": 3.4253,
+      "step": 1852416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.966442451762785e-05,
+      "loss": 3.4338,
+      "step": 1852928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9656054948921064e-05,
+      "loss": 3.4273,
+      "step": 1853440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9647669001410544e-05,
+      "loss": 3.4322,
+      "step": 1853952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9639283053900024e-05,
+      "loss": 3.4287,
+      "step": 1854464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9630897106389504e-05,
+      "loss": 3.424,
+      "step": 1854976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9622527537682713e-05,
+      "loss": 3.4388,
+      "step": 1855488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9614141590172193e-05,
+      "loss": 3.4261,
+      "step": 1856000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9605755642661676e-05,
+      "loss": 3.404,
+      "step": 1856512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9597369695151156e-05,
+      "loss": 3.4383,
+      "step": 1857024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9588983747640636e-05,
+      "loss": 3.4289,
+      "step": 1857536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9580614178933845e-05,
+      "loss": 3.439,
+      "step": 1858048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9572228231423325e-05,
+      "loss": 3.4253,
+      "step": 1858560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9563842283912805e-05,
+      "loss": 3.4198,
+      "step": 1859072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9555456336402285e-05,
+      "loss": 3.417,
+      "step": 1859584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9547086767695497e-05,
+      "loss": 3.4255,
+      "step": 1860096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9538700820184977e-05,
+      "loss": 3.4295,
+      "step": 1860608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9530314872674457e-05,
+      "loss": 3.416,
+      "step": 1861120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9521928925163937e-05,
+      "loss": 3.4358,
+      "step": 1861632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9513542977653414e-05,
+      "loss": 3.4275,
+      "step": 1862144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.950517340894663e-05,
+      "loss": 3.417,
+      "step": 1862656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.949678746143611e-05,
+      "loss": 3.434,
+      "step": 1863168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.948840151392559e-05,
+      "loss": 3.4181,
+      "step": 1863680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9480015566415066e-05,
+      "loss": 3.4174,
+      "step": 1864192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.947164599770828e-05,
+      "loss": 3.4292,
+      "step": 1864704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.946326005019776e-05,
+      "loss": 3.4379,
+      "step": 1865216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.945487410268724e-05,
+      "loss": 3.4089,
+      "step": 1865728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.944648815517672e-05,
+      "loss": 3.4275,
+      "step": 1866240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.94381022076662e-05,
+      "loss": 3.4126,
+      "step": 1866752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.942971626015568e-05,
+      "loss": 3.4272,
+      "step": 1867264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9421346691448888e-05,
+      "loss": 3.4323,
+      "step": 1867776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9412960743938368e-05,
+      "loss": 3.4246,
+      "step": 1868288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9404574796427848e-05,
+      "loss": 3.4314,
+      "step": 1868800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.939618884891733e-05,
+      "loss": 3.425,
+      "step": 1869312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.938781928021054e-05,
+      "loss": 3.446,
+      "step": 1869824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.937943333270002e-05,
+      "loss": 3.4161,
+      "step": 1870336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.93710473851895e-05,
+      "loss": 3.4331,
+      "step": 1870848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.936266143767898e-05,
+      "loss": 3.4295,
+      "step": 1871360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.935427549016846e-05,
+      "loss": 3.421,
+      "step": 1871872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9345905921461672e-05,
+      "loss": 3.4289,
+      "step": 1872384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9337519973951152e-05,
+      "loss": 3.4288,
+      "step": 1872896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9329134026440632e-05,
+      "loss": 3.4283,
+      "step": 1873408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9320748078930112e-05,
+      "loss": 3.4357,
+      "step": 1873920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.931237851022332e-05,
+      "loss": 3.4138,
+      "step": 1874432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.93039925627128e-05,
+      "loss": 3.4194,
+      "step": 1874944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9295606615202285e-05,
+      "loss": 3.4196,
+      "step": 1875456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9287220667691764e-05,
+      "loss": 3.4356,
+      "step": 1875968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9278834720181244e-05,
+      "loss": 3.4226,
+      "step": 1876480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.927044877267072e-05,
+      "loss": 3.4256,
+      "step": 1876992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.92620628251602e-05,
+      "loss": 3.43,
+      "step": 1877504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9253693256453413e-05,
+      "loss": 3.4239,
+      "step": 1878016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9245307308942893e-05,
+      "loss": 3.4213,
+      "step": 1878528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9236921361432373e-05,
+      "loss": 3.4081,
+      "step": 1879040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9228535413921853e-05,
+      "loss": 3.4233,
+      "step": 1879552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9220165845215066e-05,
+      "loss": 3.4305,
+      "step": 1880064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9211779897704542e-05,
+      "loss": 3.4327,
+      "step": 1880576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9203393950194022e-05,
+      "loss": 3.4169,
+      "step": 1881088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9195008002683502e-05,
+      "loss": 3.4209,
+      "step": 1881600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9186622055172986e-05,
+      "loss": 3.4227,
+      "step": 1882112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9178252486466195e-05,
+      "loss": 3.4269,
+      "step": 1882624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9169866538955675e-05,
+      "loss": 3.4292,
+      "step": 1883136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9161480591445155e-05,
+      "loss": 3.4259,
+      "step": 1883648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9153094643934635e-05,
+      "loss": 3.4288,
+      "step": 1884160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9144725075227847e-05,
+      "loss": 3.4223,
+      "step": 1884672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9136339127717327e-05,
+      "loss": 3.421,
+      "step": 1885184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9127953180206807e-05,
+      "loss": 3.4193,
+      "step": 1885696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9119567232696287e-05,
+      "loss": 3.4163,
+      "step": 1886208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9111181285185767e-05,
+      "loss": 3.4361,
+      "step": 1886720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9102811716478976e-05,
+      "loss": 3.4101,
+      "step": 1887232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9094425768968456e-05,
+      "loss": 3.4244,
+      "step": 1887744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.908603982145794e-05,
+      "loss": 3.4152,
+      "step": 1888256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.907765387394742e-05,
+      "loss": 3.4299,
+      "step": 1888768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9069267926436896e-05,
+      "loss": 3.4209,
+      "step": 1889280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9060898357730108e-05,
+      "loss": 3.4255,
+      "step": 1889792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9052512410219588e-05,
+      "loss": 3.4226,
+      "step": 1890304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9044126462709068e-05,
+      "loss": 3.4367,
+      "step": 1890816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9035740515198548e-05,
+      "loss": 3.4265,
+      "step": 1891328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.902737094649176e-05,
+      "loss": 3.4319,
+      "step": 1891840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.901898499898124e-05,
+      "loss": 3.4251,
+      "step": 1892352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.901059905147072e-05,
+      "loss": 3.4245,
+      "step": 1892864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9002213103960197e-05,
+      "loss": 3.42,
+      "step": 1893376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.899384353525341e-05,
+      "loss": 3.4202,
+      "step": 1893888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8985457587742893e-05,
+      "loss": 3.4179,
+      "step": 1894400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.897707164023237e-05,
+      "loss": 3.425,
+      "step": 1894912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.896868569272185e-05,
+      "loss": 3.4207,
+      "step": 1895424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8960316124015062e-05,
+      "loss": 3.427,
+      "step": 1895936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8951930176504542e-05,
+      "loss": 3.4241,
+      "step": 1896448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.894354422899402e-05,
+      "loss": 3.4248,
+      "step": 1896960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8935158281483502e-05,
+      "loss": 3.4255,
+      "step": 1897472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8926788712776714e-05,
+      "loss": 3.4286,
+      "step": 1897984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8918402765266194e-05,
+      "loss": 3.4161,
+      "step": 1898496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.891001681775567e-05,
+      "loss": 3.4265,
+      "step": 1899008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.890163087024515e-05,
+      "loss": 3.4039,
+      "step": 1899520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.889324492273463e-05,
+      "loss": 3.4314,
+      "step": 1900032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8884875354027843e-05,
+      "loss": 3.4286,
+      "step": 1900544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8876489406517323e-05,
+      "loss": 3.4331,
+      "step": 1901056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8868103459006803e-05,
+      "loss": 3.417,
+      "step": 1901568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8859717511496283e-05,
+      "loss": 3.4312,
+      "step": 1902080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8851331563985763e-05,
+      "loss": 3.4233,
+      "step": 1902592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8842961995278972e-05,
+      "loss": 3.4383,
+      "step": 1903104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8834576047768455e-05,
+      "loss": 3.4189,
+      "step": 1903616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8826190100257935e-05,
+      "loss": 3.4207,
+      "step": 1904128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8817820531551145e-05,
+      "loss": 3.4226,
+      "step": 1904640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8809434584040624e-05,
+      "loss": 3.4312,
+      "step": 1905152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8801048636530104e-05,
+      "loss": 3.4224,
+      "step": 1905664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8792662689019584e-05,
+      "loss": 3.4249,
+      "step": 1906176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8784276741509064e-05,
+      "loss": 3.432,
+      "step": 1906688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8775907172802277e-05,
+      "loss": 3.4208,
+      "step": 1907200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8767521225291757e-05,
+      "loss": 3.4294,
+      "step": 1907712
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.880861282348633,
+      "eval_runtime": 303.9922,
+      "eval_samples_per_second": 1255.266,
+      "eval_steps_per_second": 39.228,
+      "step": 1908000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8759135277781237e-05,
+      "loss": 3.4135,
+      "step": 1908224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8750749330270717e-05,
+      "loss": 3.4255,
+      "step": 1908736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8742363382760197e-05,
+      "loss": 3.4187,
+      "step": 1909248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.873399381405341e-05,
+      "loss": 3.427,
+      "step": 1909760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.872560786654289e-05,
+      "loss": 3.4402,
+      "step": 1910272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.871722191903237e-05,
+      "loss": 3.4233,
+      "step": 1910784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.870883597152185e-05,
+      "loss": 3.4161,
+      "step": 1911296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8700450024011326e-05,
+      "loss": 3.4157,
+      "step": 1911808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8692080455304538e-05,
+      "loss": 3.4273,
+      "step": 1912320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8683694507794018e-05,
+      "loss": 3.4316,
+      "step": 1912832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8675308560283498e-05,
+      "loss": 3.4175,
+      "step": 1913344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8666922612772978e-05,
+      "loss": 3.4232,
+      "step": 1913856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.865855304406619e-05,
+      "loss": 3.4292,
+      "step": 1914368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.865016709655567e-05,
+      "loss": 3.4174,
+      "step": 1914880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8641781149045147e-05,
+      "loss": 3.4159,
+      "step": 1915392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8633395201534627e-05,
+      "loss": 3.4121,
+      "step": 1915904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8625025632827843e-05,
+      "loss": 3.4178,
+      "step": 1916416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8616639685317323e-05,
+      "loss": 3.4157,
+      "step": 1916928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.86082537378068e-05,
+      "loss": 3.417,
+      "step": 1917440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.859986779029628e-05,
+      "loss": 3.4331,
+      "step": 1917952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.859148184278576e-05,
+      "loss": 3.4326,
+      "step": 1918464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8583112274078972e-05,
+      "loss": 3.4231,
+      "step": 1918976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.857472632656845e-05,
+      "loss": 3.4195,
+      "step": 1919488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.856634037905793e-05,
+      "loss": 3.4207,
+      "step": 1920000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.855795443154741e-05,
+      "loss": 3.423,
+      "step": 1920512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.854956848403689e-05,
+      "loss": 3.4203,
+      "step": 1921024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.85411989153301e-05,
+      "loss": 3.4076,
+      "step": 1921536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.853281296781958e-05,
+      "loss": 3.4256,
+      "step": 1922048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8524427020309064e-05,
+      "loss": 3.4216,
+      "step": 1922560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8516041072798544e-05,
+      "loss": 3.4081,
+      "step": 1923072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8507655125288024e-05,
+      "loss": 3.4159,
+      "step": 1923584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8499285556581233e-05,
+      "loss": 3.4236,
+      "step": 1924096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8490899609070713e-05,
+      "loss": 3.426,
+      "step": 1924608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8482513661560193e-05,
+      "loss": 3.4288,
+      "step": 1925120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8474127714049673e-05,
+      "loss": 3.4187,
+      "step": 1925632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8465741766539153e-05,
+      "loss": 3.416,
+      "step": 1926144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8457355819028633e-05,
+      "loss": 3.4317,
+      "step": 1926656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8448986250321845e-05,
+      "loss": 3.4083,
+      "step": 1927168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8440600302811325e-05,
+      "loss": 3.409,
+      "step": 1927680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8432214355300802e-05,
+      "loss": 3.4116,
+      "step": 1928192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8423828407790285e-05,
+      "loss": 3.4154,
+      "step": 1928704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8415458839083498e-05,
+      "loss": 3.4213,
+      "step": 1929216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8407072891572974e-05,
+      "loss": 3.4176,
+      "step": 1929728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8398686944062454e-05,
+      "loss": 3.4219,
+      "step": 1930240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8390300996551934e-05,
+      "loss": 3.4187,
+      "step": 1930752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8381931427845146e-05,
+      "loss": 3.4149,
+      "step": 1931264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8373545480334626e-05,
+      "loss": 3.4245,
+      "step": 1931776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8365159532824106e-05,
+      "loss": 3.4206,
+      "step": 1932288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8356773585313586e-05,
+      "loss": 3.396,
+      "step": 1932800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.83484040166068e-05,
+      "loss": 3.4266,
+      "step": 1933312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8340018069096275e-05,
+      "loss": 3.4195,
+      "step": 1933824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8331632121585755e-05,
+      "loss": 3.4306,
+      "step": 1934336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.832324617407524e-05,
+      "loss": 3.4134,
+      "step": 1934848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8314876605368448e-05,
+      "loss": 3.4106,
+      "step": 1935360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8306490657857928e-05,
+      "loss": 3.4045,
+      "step": 1935872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8298104710347408e-05,
+      "loss": 3.4152,
+      "step": 1936384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8289718762836888e-05,
+      "loss": 3.4198,
+      "step": 1936896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.82813491941301e-05,
+      "loss": 3.4077,
+      "step": 1937408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.827296324661958e-05,
+      "loss": 3.4202,
+      "step": 1937920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.826457729910906e-05,
+      "loss": 3.4173,
+      "step": 1938432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.825619135159854e-05,
+      "loss": 3.4082,
+      "step": 1938944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.824782178289175e-05,
+      "loss": 3.4188,
+      "step": 1939456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.823943583538123e-05,
+      "loss": 3.4078,
+      "step": 1939968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.823104988787071e-05,
+      "loss": 3.4088,
+      "step": 1940480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.822266394036019e-05,
+      "loss": 3.4144,
+      "step": 1940992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8214277992849672e-05,
+      "loss": 3.426,
+      "step": 1941504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8205892045339152e-05,
+      "loss": 3.4038,
+      "step": 1942016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.819752247663236e-05,
+      "loss": 3.4183,
+      "step": 1942528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.818913652912184e-05,
+      "loss": 3.3989,
+      "step": 1943040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.818075058161132e-05,
+      "loss": 3.4171,
+      "step": 1943552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.81723646341008e-05,
+      "loss": 3.4212,
+      "step": 1944064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8163995065394014e-05,
+      "loss": 3.4173,
+      "step": 1944576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8155609117883494e-05,
+      "loss": 3.4146,
+      "step": 1945088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8147223170372974e-05,
+      "loss": 3.4161,
+      "step": 1945600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8138837222862454e-05,
+      "loss": 3.4351,
+      "step": 1946112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.813045127535193e-05,
+      "loss": 3.4066,
+      "step": 1946624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.812206532784141e-05,
+      "loss": 3.4229,
+      "step": 1947136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8113695759134626e-05,
+      "loss": 3.4139,
+      "step": 1947648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8105309811624103e-05,
+      "loss": 3.4142,
+      "step": 1948160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8096923864113583e-05,
+      "loss": 3.4183,
+      "step": 1948672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8088537916603062e-05,
+      "loss": 3.4163,
+      "step": 1949184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8080168347896275e-05,
+      "loss": 3.4217,
+      "step": 1949696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8071782400385755e-05,
+      "loss": 3.4282,
+      "step": 1950208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8063396452875235e-05,
+      "loss": 3.4016,
+      "step": 1950720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8055010505364715e-05,
+      "loss": 3.411,
+      "step": 1951232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8046624557854195e-05,
+      "loss": 3.4083,
+      "step": 1951744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8038254989147404e-05,
+      "loss": 3.4229,
+      "step": 1952256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8029869041636884e-05,
+      "loss": 3.4095,
+      "step": 1952768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8021483094126364e-05,
+      "loss": 3.4182,
+      "step": 1953280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8013097146615847e-05,
+      "loss": 3.4172,
+      "step": 1953792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8004727577909056e-05,
+      "loss": 3.4126,
+      "step": 1954304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7996341630398536e-05,
+      "loss": 3.4085,
+      "step": 1954816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7987955682888016e-05,
+      "loss": 3.401,
+      "step": 1955328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7979569735377496e-05,
+      "loss": 3.4096,
+      "step": 1955840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.797120016667071e-05,
+      "loss": 3.4213,
+      "step": 1956352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.796281421916019e-05,
+      "loss": 3.4225,
+      "step": 1956864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.795442827164967e-05,
+      "loss": 3.4042,
+      "step": 1957376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.794604232413915e-05,
+      "loss": 3.4109,
+      "step": 1957888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.793765637662863e-05,
+      "loss": 3.4122,
+      "step": 1958400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7929270429118105e-05,
+      "loss": 3.4158,
+      "step": 1958912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7920900860411317e-05,
+      "loss": 3.416,
+      "step": 1959424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7912514912900797e-05,
+      "loss": 3.4168,
+      "step": 1959936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.790412896539028e-05,
+      "loss": 3.4166,
+      "step": 1960448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7895743017879757e-05,
+      "loss": 3.4144,
+      "step": 1960960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.788737344917297e-05,
+      "loss": 3.4141,
+      "step": 1961472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.787898750166245e-05,
+      "loss": 3.4072,
+      "step": 1961984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.787060155415193e-05,
+      "loss": 3.4083,
+      "step": 1962496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.786221560664141e-05,
+      "loss": 3.4208,
+      "step": 1963008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7853846037934622e-05,
+      "loss": 3.4064,
+      "step": 1963520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7845460090424102e-05,
+      "loss": 3.412,
+      "step": 1964032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.783707414291358e-05,
+      "loss": 3.4063,
+      "step": 1964544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.782870457420679e-05,
+      "loss": 3.4181,
+      "step": 1965056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.782031862669627e-05,
+      "loss": 3.4111,
+      "step": 1965568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.781193267918575e-05,
+      "loss": 3.4128,
+      "step": 1966080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.780354673167523e-05,
+      "loss": 3.4151,
+      "step": 1966592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7795177162968444e-05,
+      "loss": 3.43,
+      "step": 1967104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7786791215457923e-05,
+      "loss": 3.4154,
+      "step": 1967616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7778405267947403e-05,
+      "loss": 3.4158,
+      "step": 1968128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.777001932043688e-05,
+      "loss": 3.4193,
+      "step": 1968640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7761633372926363e-05,
+      "loss": 3.4078,
+      "step": 1969152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7753263804219576e-05,
+      "loss": 3.4144,
+      "step": 1969664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7744877856709052e-05,
+      "loss": 3.4053,
+      "step": 1970176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7736491909198532e-05,
+      "loss": 3.4078,
+      "step": 1970688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7728105961688012e-05,
+      "loss": 3.4142,
+      "step": 1971200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7719736392981225e-05,
+      "loss": 3.4127,
+      "step": 1971712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7711350445470705e-05,
+      "loss": 3.4187,
+      "step": 1972224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7702964497960185e-05,
+      "loss": 3.4133,
+      "step": 1972736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7694578550449665e-05,
+      "loss": 3.4163,
+      "step": 1973248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7686208981742877e-05,
+      "loss": 3.4106,
+      "step": 1973760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7677823034232354e-05,
+      "loss": 3.4201,
+      "step": 1974272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7669437086721834e-05,
+      "loss": 3.408,
+      "step": 1974784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7661051139211317e-05,
+      "loss": 3.4149,
+      "step": 1975296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7652681570504526e-05,
+      "loss": 3.3905,
+      "step": 1975808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7644295622994006e-05,
+      "loss": 3.4227,
+      "step": 1976320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7635909675483486e-05,
+      "loss": 3.4142,
+      "step": 1976832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7627523727972966e-05,
+      "loss": 3.4247,
+      "step": 1977344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.761915415926618e-05,
+      "loss": 3.4066,
+      "step": 1977856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.761076821175566e-05,
+      "loss": 3.4195,
+      "step": 1978368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.760238226424514e-05,
+      "loss": 3.4123,
+      "step": 1978880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.759399631673462e-05,
+      "loss": 3.4302,
+      "step": 1979392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7585626748027827e-05,
+      "loss": 3.4097,
+      "step": 1979904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7577240800517307e-05,
+      "loss": 3.4103,
+      "step": 1980416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7568854853006787e-05,
+      "loss": 3.4129,
+      "step": 1980928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.756046890549627e-05,
+      "loss": 3.4193,
+      "step": 1981440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.755209933678948e-05,
+      "loss": 3.4112,
+      "step": 1981952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.754371338927896e-05,
+      "loss": 3.4132,
+      "step": 1982464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.753532744176844e-05,
+      "loss": 3.4195,
+      "step": 1982976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.752694149425792e-05,
+      "loss": 3.4149,
+      "step": 1983488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7518571925551132e-05,
+      "loss": 3.4177,
+      "step": 1984000
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.881544828414917,
+      "eval_runtime": 303.0069,
+      "eval_samples_per_second": 1259.348,
+      "eval_steps_per_second": 39.356,
+      "step": 1984320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7510202356844345e-05,
+      "loss": 3.4031,
+      "step": 1984512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7501816409333825e-05,
+      "loss": 3.4158,
+      "step": 1985024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.74934304618233e-05,
+      "loss": 3.4115,
+      "step": 1985536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.748504451431278e-05,
+      "loss": 3.414,
+      "step": 1986048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.747665856680226e-05,
+      "loss": 3.4279,
+      "step": 1986560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7468288998095474e-05,
+      "loss": 3.41,
+      "step": 1987072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7459903050584953e-05,
+      "loss": 3.411,
+      "step": 1987584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7451517103074433e-05,
+      "loss": 3.4062,
+      "step": 1988096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7443131155563913e-05,
+      "loss": 3.4127,
+      "step": 1988608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7434745208053393e-05,
+      "loss": 3.418,
+      "step": 1989120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7426375639346602e-05,
+      "loss": 3.4107,
+      "step": 1989632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7417989691836082e-05,
+      "loss": 3.4137,
+      "step": 1990144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7409603744325566e-05,
+      "loss": 3.4191,
+      "step": 1990656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7401217796815046e-05,
+      "loss": 3.4067,
+      "step": 1991168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7392848228108255e-05,
+      "loss": 3.4047,
+      "step": 1991680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7384462280597735e-05,
+      "loss": 3.4007,
+      "step": 1992192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7376076333087215e-05,
+      "loss": 3.4124,
+      "step": 1992704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7367690385576695e-05,
+      "loss": 3.4047,
+      "step": 1993216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7359304438066175e-05,
+      "loss": 3.4094,
+      "step": 1993728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7350934869359387e-05,
+      "loss": 3.4214,
+      "step": 1994240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7342548921848867e-05,
+      "loss": 3.4228,
+      "step": 1994752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7334162974338347e-05,
+      "loss": 3.4127,
+      "step": 1995264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7325777026827827e-05,
+      "loss": 3.4088,
+      "step": 1995776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7317407458121036e-05,
+      "loss": 3.4103,
+      "step": 1996288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.730902151061052e-05,
+      "loss": 3.4138,
+      "step": 1996800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.73006355631e-05,
+      "loss": 3.4111,
+      "step": 1997312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7292249615589476e-05,
+      "loss": 3.3972,
+      "step": 1997824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7283863668078956e-05,
+      "loss": 3.4148,
+      "step": 1998336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.727549409937217e-05,
+      "loss": 3.4116,
+      "step": 1998848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.726710815186165e-05,
+      "loss": 3.3969,
+      "step": 1999360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7258722204351128e-05,
+      "loss": 3.4057,
+      "step": 1999872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7250336256840608e-05,
+      "loss": 3.4149,
+      "step": 2000384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7241950309330088e-05,
+      "loss": 3.4106,
+      "step": 2000896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.72335807406233e-05,
+      "loss": 3.4197,
+      "step": 2001408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7225194793112777e-05,
+      "loss": 3.4092,
+      "step": 2001920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7216808845602257e-05,
+      "loss": 3.4035,
+      "step": 2002432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.720842289809174e-05,
+      "loss": 3.4273,
+      "step": 2002944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.720005332938495e-05,
+      "loss": 3.3948,
+      "step": 2003456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.719166738187443e-05,
+      "loss": 3.3979,
+      "step": 2003968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.718328143436391e-05,
+      "loss": 3.4023,
+      "step": 2004480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.717489548685339e-05,
+      "loss": 3.4018,
+      "step": 2004992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.716650953934287e-05,
+      "loss": 3.4089,
+      "step": 2005504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7158139970636082e-05,
+      "loss": 3.4107,
+      "step": 2006016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7149754023125562e-05,
+      "loss": 3.4104,
+      "step": 2006528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7141368075615042e-05,
+      "loss": 3.4075,
+      "step": 2007040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7132982128104522e-05,
+      "loss": 3.4015,
+      "step": 2007552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.712461255939773e-05,
+      "loss": 3.4122,
+      "step": 2008064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.711622661188721e-05,
+      "loss": 3.409,
+      "step": 2008576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.710784066437669e-05,
+      "loss": 3.3889,
+      "step": 2009088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7099454716866174e-05,
+      "loss": 3.414,
+      "step": 2009600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7091068769355654e-05,
+      "loss": 3.4074,
+      "step": 2010112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7082699200648863e-05,
+      "loss": 3.4172,
+      "step": 2010624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7074313253138343e-05,
+      "loss": 3.4028,
+      "step": 2011136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7065927305627823e-05,
+      "loss": 3.4033,
+      "step": 2011648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7057541358117303e-05,
+      "loss": 3.3922,
+      "step": 2012160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7049171789410516e-05,
+      "loss": 3.4015,
+      "step": 2012672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7040785841899996e-05,
+      "loss": 3.4119,
+      "step": 2013184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7032399894389475e-05,
+      "loss": 3.4,
+      "step": 2013696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7024013946878955e-05,
+      "loss": 3.4101,
+      "step": 2014208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7015627999368432e-05,
+      "loss": 3.4055,
+      "step": 2014720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7007258430661645e-05,
+      "loss": 3.4,
+      "step": 2015232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6998872483151128e-05,
+      "loss": 3.4082,
+      "step": 2015744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6990486535640604e-05,
+      "loss": 3.3977,
+      "step": 2016256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6982100588130084e-05,
+      "loss": 3.401,
+      "step": 2016768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6973714640619564e-05,
+      "loss": 3.4013,
+      "step": 2017280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6965328693109044e-05,
+      "loss": 3.4154,
+      "step": 2017792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6956942745598524e-05,
+      "loss": 3.3934,
+      "step": 2018304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6948573176891737e-05,
+      "loss": 3.4062,
+      "step": 2018816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6940187229381217e-05,
+      "loss": 3.3893,
+      "step": 2019328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6931801281870697e-05,
+      "loss": 3.4057,
+      "step": 2019840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6923415334360177e-05,
+      "loss": 3.4069,
+      "step": 2020352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6915045765653386e-05,
+      "loss": 3.4098,
+      "step": 2020864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6906659818142866e-05,
+      "loss": 3.4062,
+      "step": 2021376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.689827387063235e-05,
+      "loss": 3.4032,
+      "step": 2021888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.688988792312183e-05,
+      "loss": 3.4224,
+      "step": 2022400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.688150197561131e-05,
+      "loss": 3.3986,
+      "step": 2022912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6873132406904518e-05,
+      "loss": 3.4169,
+      "step": 2023424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6864746459393998e-05,
+      "loss": 3.4017,
+      "step": 2023936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6856360511883478e-05,
+      "loss": 3.4006,
+      "step": 2024448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6847974564372958e-05,
+      "loss": 3.4083,
+      "step": 2024960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.683960499566617e-05,
+      "loss": 3.4032,
+      "step": 2025472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.683121904815565e-05,
+      "loss": 3.4128,
+      "step": 2025984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.682283310064513e-05,
+      "loss": 3.4163,
+      "step": 2026496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.681444715313461e-05,
+      "loss": 3.3962,
+      "step": 2027008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.680607758442782e-05,
+      "loss": 3.397,
+      "step": 2027520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6797691636917303e-05,
+      "loss": 3.3985,
+      "step": 2028032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6789305689406783e-05,
+      "loss": 3.4097,
+      "step": 2028544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.678091974189626e-05,
+      "loss": 3.4009,
+      "step": 2029056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.677253379438574e-05,
+      "loss": 3.4088,
+      "step": 2029568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.676414784687522e-05,
+      "loss": 3.41,
+      "step": 2030080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.675577827816843e-05,
+      "loss": 3.4036,
+      "step": 2030592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.674739233065791e-05,
+      "loss": 3.3961,
+      "step": 2031104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.673900638314739e-05,
+      "loss": 3.3917,
+      "step": 2031616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.673062043563687e-05,
+      "loss": 3.3978,
+      "step": 2032128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6722250866930084e-05,
+      "loss": 3.411,
+      "step": 2032640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.671386491941956e-05,
+      "loss": 3.4101,
+      "step": 2033152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.670547897190904e-05,
+      "loss": 3.394,
+      "step": 2033664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.669709302439852e-05,
+      "loss": 3.4011,
+      "step": 2034176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6688707076888004e-05,
+      "loss": 3.4004,
+      "step": 2034688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6680337508181213e-05,
+      "loss": 3.4078,
+      "step": 2035200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6671951560670693e-05,
+      "loss": 3.4082,
+      "step": 2035712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6663565613160173e-05,
+      "loss": 3.4031,
+      "step": 2036224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6655179665649653e-05,
+      "loss": 3.4062,
+      "step": 2036736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6646793718139133e-05,
+      "loss": 3.4019,
+      "step": 2037248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6638424149432345e-05,
+      "loss": 3.4001,
+      "step": 2037760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6630038201921825e-05,
+      "loss": 3.3986,
+      "step": 2038272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6621652254411305e-05,
+      "loss": 3.4023,
+      "step": 2038784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6613266306900785e-05,
+      "loss": 3.4063,
+      "step": 2039296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6604896738193994e-05,
+      "loss": 3.3958,
+      "step": 2039808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6596510790683474e-05,
+      "loss": 3.4009,
+      "step": 2040320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6588124843172957e-05,
+      "loss": 3.397,
+      "step": 2040832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6579755274466167e-05,
+      "loss": 3.4021,
+      "step": 2041344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6571369326955646e-05,
+      "loss": 3.4048,
+      "step": 2041856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6562983379445126e-05,
+      "loss": 3.4014,
+      "step": 2042368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6554597431934606e-05,
+      "loss": 3.4031,
+      "step": 2042880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.654622786322782e-05,
+      "loss": 3.4194,
+      "step": 2043392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.65378419157173e-05,
+      "loss": 3.4092,
+      "step": 2043904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.652945596820678e-05,
+      "loss": 3.4035,
+      "step": 2044416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.652107002069626e-05,
+      "loss": 3.4081,
+      "step": 2044928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6512684073185735e-05,
+      "loss": 3.3997,
+      "step": 2045440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6504314504478948e-05,
+      "loss": 3.4062,
+      "step": 2045952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6495928556968428e-05,
+      "loss": 3.3926,
+      "step": 2046464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.648754260945791e-05,
+      "loss": 3.3979,
+      "step": 2046976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6479156661947388e-05,
+      "loss": 3.4039,
+      "step": 2047488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.64707870932406e-05,
+      "loss": 3.4053,
+      "step": 2048000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.646240114573008e-05,
+      "loss": 3.4035,
+      "step": 2048512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.645401519821956e-05,
+      "loss": 3.4033,
+      "step": 2049024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6445629250709037e-05,
+      "loss": 3.4057,
+      "step": 2049536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6437259682002252e-05,
+      "loss": 3.4021,
+      "step": 2050048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6428873734491732e-05,
+      "loss": 3.4052,
+      "step": 2050560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.642048778698121e-05,
+      "loss": 3.402,
+      "step": 2051072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.641210183947069e-05,
+      "loss": 3.3998,
+      "step": 2051584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.64037322707639e-05,
+      "loss": 3.3828,
+      "step": 2052096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.639534632325338e-05,
+      "loss": 3.4119,
+      "step": 2052608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.638696037574286e-05,
+      "loss": 3.4025,
+      "step": 2053120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.637857442823234e-05,
+      "loss": 3.4131,
+      "step": 2053632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6370204859525554e-05,
+      "loss": 3.3978,
+      "step": 2054144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6361818912015034e-05,
+      "loss": 3.405,
+      "step": 2054656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.635343296450451e-05,
+      "loss": 3.4036,
+      "step": 2055168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.634504701699399e-05,
+      "loss": 3.4215,
+      "step": 2055680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6336677448287206e-05,
+      "loss": 3.3941,
+      "step": 2056192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6328291500776683e-05,
+      "loss": 3.3994,
+      "step": 2056704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6319905553266163e-05,
+      "loss": 3.4039,
+      "step": 2057216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6311519605755643e-05,
+      "loss": 3.4073,
+      "step": 2057728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6303150037048855e-05,
+      "loss": 3.4028,
+      "step": 2058240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6294764089538335e-05,
+      "loss": 3.4033,
+      "step": 2058752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6286378142027815e-05,
+      "loss": 3.404,
+      "step": 2059264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6277992194517295e-05,
+      "loss": 3.4087,
+      "step": 2059776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6269606247006775e-05,
+      "loss": 3.4052,
+      "step": 2060288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8837647438049316,
+      "eval_runtime": 306.8196,
+      "eval_samples_per_second": 1243.698,
+      "eval_steps_per_second": 38.866,
+      "step": 2060640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6261236678299984e-05,
+      "loss": 3.3952,
+      "step": 2060800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6252850730789464e-05,
+      "loss": 3.4038,
+      "step": 2061312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6244464783278944e-05,
+      "loss": 3.4033,
+      "step": 2061824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6236078835768427e-05,
+      "loss": 3.4044,
+      "step": 2062336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6227709267061636e-05,
+      "loss": 3.4124,
+      "step": 2062848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6219323319551116e-05,
+      "loss": 3.4008,
+      "step": 2063360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6210937372040596e-05,
+      "loss": 3.403,
+      "step": 2063872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6202551424530076e-05,
+      "loss": 3.397,
+      "step": 2064384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6194165477019556e-05,
+      "loss": 3.3988,
+      "step": 2064896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.618579590831277e-05,
+      "loss": 3.4085,
+      "step": 2065408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.617740996080225e-05,
+      "loss": 3.4032,
+      "step": 2065920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.616902401329173e-05,
+      "loss": 3.4013,
+      "step": 2066432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.616063806578121e-05,
+      "loss": 3.415,
+      "step": 2066944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.615225211827069e-05,
+      "loss": 3.3901,
+      "step": 2067456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6143882549563898e-05,
+      "loss": 3.3971,
+      "step": 2067968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.613549660205338e-05,
+      "loss": 3.389,
+      "step": 2068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.612711065454286e-05,
+      "loss": 3.4026,
+      "step": 2068992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6118724707032337e-05,
+      "loss": 3.3932,
+      "step": 2069504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.611035513832555e-05,
+      "loss": 3.4023,
+      "step": 2070016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.610196919081503e-05,
+      "loss": 3.41,
+      "step": 2070528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.609358324330451e-05,
+      "loss": 3.4135,
+      "step": 2071040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.608519729579399e-05,
+      "loss": 3.4045,
+      "step": 2071552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6076827727087202e-05,
+      "loss": 3.3978,
+      "step": 2072064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6068441779576682e-05,
+      "loss": 3.4022,
+      "step": 2072576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.606005583206616e-05,
+      "loss": 3.4039,
+      "step": 2073088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.605166988455564e-05,
+      "loss": 3.3999,
+      "step": 2073600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.604330031584885e-05,
+      "loss": 3.3896,
+      "step": 2074112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6034914368338335e-05,
+      "loss": 3.4015,
+      "step": 2074624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.602652842082781e-05,
+      "loss": 3.4015,
+      "step": 2075136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.601814247331729e-05,
+      "loss": 3.3862,
+      "step": 2075648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.600975652580677e-05,
+      "loss": 3.3947,
+      "step": 2076160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6001386957099984e-05,
+      "loss": 3.4039,
+      "step": 2076672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.599300100958946e-05,
+      "loss": 3.3993,
+      "step": 2077184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5984615062078943e-05,
+      "loss": 3.4087,
+      "step": 2077696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5976229114568423e-05,
+      "loss": 3.4012,
+      "step": 2078208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5967843167057903e-05,
+      "loss": 3.3951,
+      "step": 2078720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5959473598351113e-05,
+      "loss": 3.4159,
+      "step": 2079232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5951087650840592e-05,
+      "loss": 3.3813,
+      "step": 2079744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5942701703330072e-05,
+      "loss": 3.3905,
+      "step": 2080256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5934315755819552e-05,
+      "loss": 3.3943,
+      "step": 2080768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5925946187112765e-05,
+      "loss": 3.3925,
+      "step": 2081280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5917560239602245e-05,
+      "loss": 3.3946,
+      "step": 2081792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5909174292091725e-05,
+      "loss": 3.4054,
+      "step": 2082304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5900788344581205e-05,
+      "loss": 3.3976,
+      "step": 2082816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5892418775874414e-05,
+      "loss": 3.3976,
+      "step": 2083328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5884032828363897e-05,
+      "loss": 3.3912,
+      "step": 2083840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5875646880853377e-05,
+      "loss": 3.4006,
+      "step": 2084352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5867260933342857e-05,
+      "loss": 3.4001,
+      "step": 2084864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5858874985832337e-05,
+      "loss": 3.3795,
+      "step": 2085376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5850505417125546e-05,
+      "loss": 3.4018,
+      "step": 2085888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5842119469615026e-05,
+      "loss": 3.3976,
+      "step": 2086400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5833733522104506e-05,
+      "loss": 3.4053,
+      "step": 2086912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.582534757459399e-05,
+      "loss": 3.3926,
+      "step": 2087424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.58169780058872e-05,
+      "loss": 3.391,
+      "step": 2087936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.580859205837668e-05,
+      "loss": 3.3859,
+      "step": 2088448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.580020611086616e-05,
+      "loss": 3.3883,
+      "step": 2088960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.579182016335564e-05,
+      "loss": 3.4034,
+      "step": 2089472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5783434215845115e-05,
+      "loss": 3.3877,
+      "step": 2089984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.577506464713833e-05,
+      "loss": 3.3993,
+      "step": 2090496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.576667869962781e-05,
+      "loss": 3.3921,
+      "step": 2091008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5758292752117287e-05,
+      "loss": 3.3908,
+      "step": 2091520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.57499231834105e-05,
+      "loss": 3.3924,
+      "step": 2092032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.574153723589998e-05,
+      "loss": 3.3899,
+      "step": 2092544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.573315128838946e-05,
+      "loss": 3.3925,
+      "step": 2093056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.572476534087894e-05,
+      "loss": 3.3841,
+      "step": 2093568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.571637939336842e-05,
+      "loss": 3.4099,
+      "step": 2094080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5708009824661632e-05,
+      "loss": 3.3886,
+      "step": 2094592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5699623877151112e-05,
+      "loss": 3.3882,
+      "step": 2095104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.569123792964059e-05,
+      "loss": 3.3818,
+      "step": 2095616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.568285198213007e-05,
+      "loss": 3.3939,
+      "step": 2096128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5674482413423284e-05,
+      "loss": 3.3955,
+      "step": 2096640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.566609646591276e-05,
+      "loss": 3.4007,
+      "step": 2097152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.565771051840224e-05,
+      "loss": 3.399,
+      "step": 2097664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.564932457089172e-05,
+      "loss": 3.3936,
+      "step": 2098176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.56409386233812e-05,
+      "loss": 3.4101,
+      "step": 2098688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5632569054674413e-05,
+      "loss": 3.3925,
+      "step": 2099200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5624183107163893e-05,
+      "loss": 3.4049,
+      "step": 2099712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5615797159653373e-05,
+      "loss": 3.391,
+      "step": 2100224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5607411212142853e-05,
+      "loss": 3.3925,
+      "step": 2100736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5599041643436062e-05,
+      "loss": 3.3935,
+      "step": 2101248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5590655695925542e-05,
+      "loss": 3.3921,
+      "step": 2101760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5582269748415022e-05,
+      "loss": 3.4014,
+      "step": 2102272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5573883800904506e-05,
+      "loss": 3.4092,
+      "step": 2102784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5565497853393986e-05,
+      "loss": 3.3891,
+      "step": 2103296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5557128284687195e-05,
+      "loss": 3.3828,
+      "step": 2103808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5548742337176675e-05,
+      "loss": 3.388,
+      "step": 2104320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5540356389666155e-05,
+      "loss": 3.3993,
+      "step": 2104832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5531970442155635e-05,
+      "loss": 3.3921,
+      "step": 2105344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5523600873448847e-05,
+      "loss": 3.3938,
+      "step": 2105856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5515214925938327e-05,
+      "loss": 3.4023,
+      "step": 2106368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5506828978427807e-05,
+      "loss": 3.3946,
+      "step": 2106880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5498443030917287e-05,
+      "loss": 3.3833,
+      "step": 2107392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5490073462210496e-05,
+      "loss": 3.3811,
+      "step": 2107904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5481687514699976e-05,
+      "loss": 3.3893,
+      "step": 2108416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.547330156718946e-05,
+      "loss": 3.3959,
+      "step": 2108928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.546491561967894e-05,
+      "loss": 3.4059,
+      "step": 2109440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.545654605097215e-05,
+      "loss": 3.3809,
+      "step": 2109952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5448160103461628e-05,
+      "loss": 3.3923,
+      "step": 2110464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5439774155951108e-05,
+      "loss": 3.3946,
+      "step": 2110976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5431388208440588e-05,
+      "loss": 3.3936,
+      "step": 2111488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.54230186397338e-05,
+      "loss": 3.398,
+      "step": 2112000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.541463269222328e-05,
+      "loss": 3.3951,
+      "step": 2112512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.540624674471276e-05,
+      "loss": 3.3989,
+      "step": 2113024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5397860797202237e-05,
+      "loss": 3.3882,
+      "step": 2113536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5389474849691717e-05,
+      "loss": 3.3909,
+      "step": 2114048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.538110528098493e-05,
+      "loss": 3.3892,
+      "step": 2114560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5372719333474413e-05,
+      "loss": 3.3916,
+      "step": 2115072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.536433338596389e-05,
+      "loss": 3.3959,
+      "step": 2115584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.535594743845337e-05,
+      "loss": 3.3874,
+      "step": 2116096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5347577869746582e-05,
+      "loss": 3.3879,
+      "step": 2116608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5339191922236062e-05,
+      "loss": 3.3835,
+      "step": 2117120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.533080597472554e-05,
+      "loss": 3.3923,
+      "step": 2117632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5322420027215022e-05,
+      "loss": 3.3949,
+      "step": 2118144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5314034079704502e-05,
+      "loss": 3.3934,
+      "step": 2118656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.530566451099771e-05,
+      "loss": 3.3908,
+      "step": 2119168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.529727856348719e-05,
+      "loss": 3.4016,
+      "step": 2119680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.528889261597667e-05,
+      "loss": 3.4016,
+      "step": 2120192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.528050666846615e-05,
+      "loss": 3.397,
+      "step": 2120704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5272137099759363e-05,
+      "loss": 3.3977,
+      "step": 2121216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5263751152248843e-05,
+      "loss": 3.3932,
+      "step": 2121728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5255365204738323e-05,
+      "loss": 3.3935,
+      "step": 2122240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5246979257227803e-05,
+      "loss": 3.3841,
+      "step": 2122752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5238609688521014e-05,
+      "loss": 3.3887,
+      "step": 2123264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5230223741010494e-05,
+      "loss": 3.3942,
+      "step": 2123776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5221837793499974e-05,
+      "loss": 3.3944,
+      "step": 2124288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5213451845989454e-05,
+      "loss": 3.3941,
+      "step": 2124800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5205082277282665e-05,
+      "loss": 3.3932,
+      "step": 2125312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5196696329772144e-05,
+      "loss": 3.3944,
+      "step": 2125824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5188310382261624e-05,
+      "loss": 3.3923,
+      "step": 2126336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5179924434751106e-05,
+      "loss": 3.3951,
+      "step": 2126848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5171538487240586e-05,
+      "loss": 3.3928,
+      "step": 2127360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5163168918533797e-05,
+      "loss": 3.3944,
+      "step": 2127872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5154782971023277e-05,
+      "loss": 3.368,
+      "step": 2128384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5146397023512757e-05,
+      "loss": 3.4033,
+      "step": 2128896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5138027454805968e-05,
+      "loss": 3.3913,
+      "step": 2129408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5129641507295447e-05,
+      "loss": 3.4032,
+      "step": 2129920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5121255559784927e-05,
+      "loss": 3.39,
+      "step": 2130432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5112869612274407e-05,
+      "loss": 3.392,
+      "step": 2130944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5104483664763889e-05,
+      "loss": 3.3967,
+      "step": 2131456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5096114096057098e-05,
+      "loss": 3.4086,
+      "step": 2131968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5087728148546578e-05,
+      "loss": 3.3834,
+      "step": 2132480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.507934220103606e-05,
+      "loss": 3.3934,
+      "step": 2132992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.507095625352554e-05,
+      "loss": 3.3928,
+      "step": 2133504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.506258668481875e-05,
+      "loss": 3.3938,
+      "step": 2134016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.505420073730823e-05,
+      "loss": 3.3949,
+      "step": 2134528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.504581478979771e-05,
+      "loss": 3.394,
+      "step": 2135040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.503742884228719e-05,
+      "loss": 3.3918,
+      "step": 2135552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5029059273580401e-05,
+      "loss": 3.3985,
+      "step": 2136064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5020673326069881e-05,
+      "loss": 3.3927,
+      "step": 2136576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8836889266967773,
+      "eval_runtime": 307.1752,
+      "eval_samples_per_second": 1242.259,
+      "eval_steps_per_second": 38.821,
+      "step": 2136960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5012303757363092e-05,
+      "loss": 3.3942,
+      "step": 2137088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5003917809852572e-05,
+      "loss": 3.3929,
+      "step": 2137600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4995531862342052e-05,
+      "loss": 3.3895,
+      "step": 2138112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4987145914831532e-05,
+      "loss": 3.3972,
+      "step": 2138624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4978776346124743e-05,
+      "loss": 3.3993,
+      "step": 2139136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4970390398614223e-05,
+      "loss": 3.395,
+      "step": 2139648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4962004451103702e-05,
+      "loss": 3.3917,
+      "step": 2140160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4953618503593184e-05,
+      "loss": 3.3824,
+      "step": 2140672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4945232556082664e-05,
+      "loss": 3.3862,
+      "step": 2141184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4936862987375875e-05,
+      "loss": 3.4044,
+      "step": 2141696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4928477039865355e-05,
+      "loss": 3.3901,
+      "step": 2142208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4920091092354835e-05,
+      "loss": 3.3885,
+      "step": 2142720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4911705144844313e-05,
+      "loss": 3.4064,
+      "step": 2143232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4903335576137526e-05,
+      "loss": 3.3807,
+      "step": 2143744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4894949628627005e-05,
+      "loss": 3.3847,
+      "step": 2144256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4886563681116485e-05,
+      "loss": 3.3795,
+      "step": 2144768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4878177733605964e-05,
+      "loss": 3.3906,
+      "step": 2145280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4869791786095444e-05,
+      "loss": 3.3825,
+      "step": 2145792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4861422217388656e-05,
+      "loss": 3.3908,
+      "step": 2146304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4853036269878138e-05,
+      "loss": 3.3999,
+      "step": 2146816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4844650322367614e-05,
+      "loss": 3.3995,
+      "step": 2147328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4836264374857094e-05,
+      "loss": 3.3961,
+      "step": 2147840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4827894806150308e-05,
+      "loss": 3.3823,
+      "step": 2148352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4819508858639785e-05,
+      "loss": 3.3903,
+      "step": 2148864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4811122911129267e-05,
+      "loss": 3.3968,
+      "step": 2149376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4802736963618747e-05,
+      "loss": 3.3887,
+      "step": 2149888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4794367394911959e-05,
+      "loss": 3.3801,
+      "step": 2150400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4785981447401437e-05,
+      "loss": 3.3901,
+      "step": 2150912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4777595499890917e-05,
+      "loss": 3.3938,
+      "step": 2151424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4769209552380397e-05,
+      "loss": 3.3753,
+      "step": 2151936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4760839983673608e-05,
+      "loss": 3.3857,
+      "step": 2152448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4752454036163088e-05,
+      "loss": 3.3888,
+      "step": 2152960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4744068088652568e-05,
+      "loss": 3.3943,
+      "step": 2153472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4735682141142048e-05,
+      "loss": 3.3937,
+      "step": 2153984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4727312572435259e-05,
+      "loss": 3.3916,
+      "step": 2154496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4718926624924739e-05,
+      "loss": 3.3908,
+      "step": 2155008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.471054067741422e-05,
+      "loss": 3.4029,
+      "step": 2155520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.47021547299037e-05,
+      "loss": 3.3722,
+      "step": 2156032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.469378516119691e-05,
+      "loss": 3.3802,
+      "step": 2156544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4685399213686391e-05,
+      "loss": 3.3882,
+      "step": 2157056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4677013266175871e-05,
+      "loss": 3.3776,
+      "step": 2157568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4668627318665351e-05,
+      "loss": 3.3887,
+      "step": 2158080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4660241371154831e-05,
+      "loss": 3.3921,
+      "step": 2158592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4651871802448042e-05,
+      "loss": 3.3874,
+      "step": 2159104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4643485854937522e-05,
+      "loss": 3.3873,
+      "step": 2159616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4635099907427002e-05,
+      "loss": 3.3838,
+      "step": 2160128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4626713959916483e-05,
+      "loss": 3.3868,
+      "step": 2160640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4618344391209692e-05,
+      "loss": 3.396,
+      "step": 2161152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4609958443699174e-05,
+      "loss": 3.364,
+      "step": 2161664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4601572496188654e-05,
+      "loss": 3.3894,
+      "step": 2162176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4593186548678134e-05,
+      "loss": 3.3922,
+      "step": 2162688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4584816979971345e-05,
+      "loss": 3.3954,
+      "step": 2163200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4576431032460825e-05,
+      "loss": 3.3834,
+      "step": 2163712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4568045084950305e-05,
+      "loss": 3.3797,
+      "step": 2164224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4559659137439785e-05,
+      "loss": 3.3799,
+      "step": 2164736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4551289568732995e-05,
+      "loss": 3.3767,
+      "step": 2165248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4542903621222475e-05,
+      "loss": 3.3904,
+      "step": 2165760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4534517673711955e-05,
+      "loss": 3.3787,
+      "step": 2166272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4526131726201437e-05,
+      "loss": 3.3909,
+      "step": 2166784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4517762157494646e-05,
+      "loss": 3.3811,
+      "step": 2167296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4509376209984126e-05,
+      "loss": 3.3814,
+      "step": 2167808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4500990262473608e-05,
+      "loss": 3.3782,
+      "step": 2168320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4492604314963088e-05,
+      "loss": 3.3844,
+      "step": 2168832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4484234746256298e-05,
+      "loss": 3.3791,
+      "step": 2169344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4475848798745778e-05,
+      "loss": 3.3779,
+      "step": 2169856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4467462851235258e-05,
+      "loss": 3.397,
+      "step": 2170368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4459076903724737e-05,
+      "loss": 3.3852,
+      "step": 2170880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4450690956214217e-05,
+      "loss": 3.3784,
+      "step": 2171392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4442321387507429e-05,
+      "loss": 3.3703,
+      "step": 2171904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4433935439996909e-05,
+      "loss": 3.3857,
+      "step": 2172416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4425549492486387e-05,
+      "loss": 3.3849,
+      "step": 2172928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4417163544975867e-05,
+      "loss": 3.3916,
+      "step": 2173440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.440879397626908e-05,
+      "loss": 3.3838,
+      "step": 2173952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4400408028758561e-05,
+      "loss": 3.3854,
+      "step": 2174464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4392022081248038e-05,
+      "loss": 3.3969,
+      "step": 2174976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4383636133737518e-05,
+      "loss": 3.3833,
+      "step": 2175488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4375266565030732e-05,
+      "loss": 3.3941,
+      "step": 2176000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4366880617520209e-05,
+      "loss": 3.3831,
+      "step": 2176512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.435849467000969e-05,
+      "loss": 3.3813,
+      "step": 2177024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.435010872249917e-05,
+      "loss": 3.3852,
+      "step": 2177536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4341739153792383e-05,
+      "loss": 3.3861,
+      "step": 2178048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4333353206281861e-05,
+      "loss": 3.3849,
+      "step": 2178560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4324967258771341e-05,
+      "loss": 3.3981,
+      "step": 2179072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4316581311260821e-05,
+      "loss": 3.3856,
+      "step": 2179584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4308211742554033e-05,
+      "loss": 3.3711,
+      "step": 2180096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4299825795043512e-05,
+      "loss": 3.3766,
+      "step": 2180608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4291439847532992e-05,
+      "loss": 3.3872,
+      "step": 2181120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4283053900022472e-05,
+      "loss": 3.3884,
+      "step": 2181632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4274684331315682e-05,
+      "loss": 3.3837,
+      "step": 2182144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4266298383805162e-05,
+      "loss": 3.3895,
+      "step": 2182656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4257912436294644e-05,
+      "loss": 3.3857,
+      "step": 2183168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4249526488784124e-05,
+      "loss": 3.375,
+      "step": 2183680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4241156920077333e-05,
+      "loss": 3.3704,
+      "step": 2184192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4232770972566815e-05,
+      "loss": 3.3803,
+      "step": 2184704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4224385025056295e-05,
+      "loss": 3.3852,
+      "step": 2185216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4215999077545775e-05,
+      "loss": 3.3965,
+      "step": 2185728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4207629508838985e-05,
+      "loss": 3.3722,
+      "step": 2186240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4199243561328465e-05,
+      "loss": 3.3807,
+      "step": 2186752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4190857613817945e-05,
+      "loss": 3.3848,
+      "step": 2187264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4182471666307425e-05,
+      "loss": 3.3848,
+      "step": 2187776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4174102097600636e-05,
+      "loss": 3.3899,
+      "step": 2188288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4165716150090116e-05,
+      "loss": 3.3845,
+      "step": 2188800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4157330202579598e-05,
+      "loss": 3.3815,
+      "step": 2189312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4148944255069078e-05,
+      "loss": 3.3831,
+      "step": 2189824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4140574686362287e-05,
+      "loss": 3.3794,
+      "step": 2190336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4132188738851768e-05,
+      "loss": 3.3794,
+      "step": 2190848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4123802791341248e-05,
+      "loss": 3.3834,
+      "step": 2191360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4115416843830728e-05,
+      "loss": 3.3866,
+      "step": 2191872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4107047275123939e-05,
+      "loss": 3.3785,
+      "step": 2192384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4098661327613419e-05,
+      "loss": 3.3757,
+      "step": 2192896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4090275380102899e-05,
+      "loss": 3.376,
+      "step": 2193408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4081889432592379e-05,
+      "loss": 3.3799,
+      "step": 2193920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.407351986388559e-05,
+      "loss": 3.3839,
+      "step": 2194432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.406513391637507e-05,
+      "loss": 3.3823,
+      "step": 2194944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.405674796886455e-05,
+      "loss": 3.3801,
+      "step": 2195456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4048362021354031e-05,
+      "loss": 3.3899,
+      "step": 2195968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4039976073843511e-05,
+      "loss": 3.3951,
+      "step": 2196480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4031606505136722e-05,
+      "loss": 3.385,
+      "step": 2196992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4023220557626202e-05,
+      "loss": 3.3896,
+      "step": 2197504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4014834610115682e-05,
+      "loss": 3.3854,
+      "step": 2198016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.400644866260516e-05,
+      "loss": 3.3789,
+      "step": 2198528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3998079093898373e-05,
+      "loss": 3.3739,
+      "step": 2199040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3989693146387853e-05,
+      "loss": 3.3801,
+      "step": 2199552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3981307198877333e-05,
+      "loss": 3.383,
+      "step": 2200064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.397292125136681e-05,
+      "loss": 3.3813,
+      "step": 2200576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3964551682660023e-05,
+      "loss": 3.3861,
+      "step": 2201088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3956165735149503e-05,
+      "loss": 3.3829,
+      "step": 2201600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3947779787638985e-05,
+      "loss": 3.3839,
+      "step": 2202112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3939393840128461e-05,
+      "loss": 3.3801,
+      "step": 2202624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3931024271421676e-05,
+      "loss": 3.3874,
+      "step": 2203136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3922638323911156e-05,
+      "loss": 3.3804,
+      "step": 2203648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3914252376400632e-05,
+      "loss": 3.3857,
+      "step": 2204160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3905866428890114e-05,
+      "loss": 3.3596,
+      "step": 2204672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3897496860183326e-05,
+      "loss": 3.3881,
+      "step": 2205184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3889110912672806e-05,
+      "loss": 3.3844,
+      "step": 2205696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3880724965162284e-05,
+      "loss": 3.3914,
+      "step": 2206208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3872339017651764e-05,
+      "loss": 3.3806,
+      "step": 2206720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3863969448944977e-05,
+      "loss": 3.3756,
+      "step": 2207232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3855583501434457e-05,
+      "loss": 3.3874,
+      "step": 2207744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3847197553923935e-05,
+      "loss": 3.3999,
+      "step": 2208256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3838811606413415e-05,
+      "loss": 3.3736,
+      "step": 2208768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.383044203770663e-05,
+      "loss": 3.3858,
+      "step": 2209280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3822056090196106e-05,
+      "loss": 3.3824,
+      "step": 2209792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3813670142685586e-05,
+      "loss": 3.3838,
+      "step": 2210304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3805284195175067e-05,
+      "loss": 3.3837,
+      "step": 2210816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.379691462646828e-05,
+      "loss": 3.3825,
+      "step": 2211328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3788528678957757e-05,
+      "loss": 3.3809,
+      "step": 2211840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3780142731447238e-05,
+      "loss": 3.3879,
+      "step": 2212352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.377177316274045e-05,
+      "loss": 3.3877,
+      "step": 2212864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.885349988937378,
+      "eval_runtime": 309.3665,
+      "eval_samples_per_second": 1233.459,
+      "eval_steps_per_second": 38.547,
+      "step": 2213280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.376338721522993e-05,
+      "loss": 3.3719,
+      "step": 2213376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3755001267719409e-05,
+      "loss": 3.3868,
+      "step": 2213888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3746615320208889e-05,
+      "loss": 3.3787,
+      "step": 2214400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3738245751502101e-05,
+      "loss": 3.3895,
+      "step": 2214912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.372985980399158e-05,
+      "loss": 3.3926,
+      "step": 2215424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.372147385648106e-05,
+      "loss": 3.3831,
+      "step": 2215936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.371308790897054e-05,
+      "loss": 3.3822,
+      "step": 2216448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3704718340263754e-05,
+      "loss": 3.3722,
+      "step": 2216960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.369633239275323e-05,
+      "loss": 3.3799,
+      "step": 2217472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.368794644524271e-05,
+      "loss": 3.3903,
+      "step": 2217984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3679560497732192e-05,
+      "loss": 3.3808,
+      "step": 2218496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3671174550221672e-05,
+      "loss": 3.3823,
+      "step": 2219008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3662804981514883e-05,
+      "loss": 3.3926,
+      "step": 2219520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3654419034004363e-05,
+      "loss": 3.3739,
+      "step": 2220032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3646033086493842e-05,
+      "loss": 3.3734,
+      "step": 2220544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3637647138983322e-05,
+      "loss": 3.3688,
+      "step": 2221056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3629277570276533e-05,
+      "loss": 3.3817,
+      "step": 2221568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3620891622766013e-05,
+      "loss": 3.3742,
+      "step": 2222080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3612505675255493e-05,
+      "loss": 3.3783,
+      "step": 2222592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3604119727744973e-05,
+      "loss": 3.387,
+      "step": 2223104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3595750159038184e-05,
+      "loss": 3.3917,
+      "step": 2223616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3587364211527664e-05,
+      "loss": 3.3848,
+      "step": 2224128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3578978264017145e-05,
+      "loss": 3.3763,
+      "step": 2224640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3570592316506625e-05,
+      "loss": 3.3785,
+      "step": 2225152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3562222747799835e-05,
+      "loss": 3.387,
+      "step": 2225664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3553836800289316e-05,
+      "loss": 3.3782,
+      "step": 2226176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3545450852778796e-05,
+      "loss": 3.3704,
+      "step": 2226688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3537064905268276e-05,
+      "loss": 3.3813,
+      "step": 2227200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3528695336561487e-05,
+      "loss": 3.3763,
+      "step": 2227712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3520309389050967e-05,
+      "loss": 3.37,
+      "step": 2228224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3511923441540447e-05,
+      "loss": 3.3754,
+      "step": 2228736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3503537494029927e-05,
+      "loss": 3.3785,
+      "step": 2229248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3495167925323138e-05,
+      "loss": 3.383,
+      "step": 2229760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3486781977812618e-05,
+      "loss": 3.386,
+      "step": 2230272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.34783960303021e-05,
+      "loss": 3.3793,
+      "step": 2230784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3470010082791579e-05,
+      "loss": 3.3811,
+      "step": 2231296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3461624135281056e-05,
+      "loss": 3.3923,
+      "step": 2231808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.345325456657427e-05,
+      "loss": 3.365,
+      "step": 2232320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.344486861906375e-05,
+      "loss": 3.3724,
+      "step": 2232832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.343648267155323e-05,
+      "loss": 3.3765,
+      "step": 2233344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3428096724042708e-05,
+      "loss": 3.3661,
+      "step": 2233856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.341972715533592e-05,
+      "loss": 3.3789,
+      "step": 2234368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.34113412078254e-05,
+      "loss": 3.3828,
+      "step": 2234880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.340295526031488e-05,
+      "loss": 3.3789,
+      "step": 2235392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3394569312804359e-05,
+      "loss": 3.376,
+      "step": 2235904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3386199744097571e-05,
+      "loss": 3.3787,
+      "step": 2236416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3377813796587053e-05,
+      "loss": 3.3757,
+      "step": 2236928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.336942784907653e-05,
+      "loss": 3.3849,
+      "step": 2237440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.336104190156601e-05,
+      "loss": 3.3566,
+      "step": 2237952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3352672332859224e-05,
+      "loss": 3.3735,
+      "step": 2238464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3344286385348703e-05,
+      "loss": 3.3866,
+      "step": 2238976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.333590043783818e-05,
+      "loss": 3.3859,
+      "step": 2239488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3327514490327662e-05,
+      "loss": 3.373,
+      "step": 2240000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3319144921620874e-05,
+      "loss": 3.3677,
+      "step": 2240512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3310758974110354e-05,
+      "loss": 3.3723,
+      "step": 2241024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3302373026599832e-05,
+      "loss": 3.3667,
+      "step": 2241536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3293987079089312e-05,
+      "loss": 3.3811,
+      "step": 2242048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3285617510382525e-05,
+      "loss": 3.3693,
+      "step": 2242560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3277231562872003e-05,
+      "loss": 3.378,
+      "step": 2243072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3268845615361483e-05,
+      "loss": 3.374,
+      "step": 2243584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3260459667850963e-05,
+      "loss": 3.3743,
+      "step": 2244096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3252090099144177e-05,
+      "loss": 3.368,
+      "step": 2244608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3243704151633654e-05,
+      "loss": 3.376,
+      "step": 2245120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3235318204123134e-05,
+      "loss": 3.3672,
+      "step": 2245632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3226932256612615e-05,
+      "loss": 3.3633,
+      "step": 2246144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3218562687905828e-05,
+      "loss": 3.383,
+      "step": 2246656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3210176740395306e-05,
+      "loss": 3.3764,
+      "step": 2247168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3201790792884786e-05,
+      "loss": 3.3671,
+      "step": 2247680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3193404845374266e-05,
+      "loss": 3.3603,
+      "step": 2248192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3185018897863746e-05,
+      "loss": 3.3747,
+      "step": 2248704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3176649329156957e-05,
+      "loss": 3.3736,
+      "step": 2249216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3168263381646437e-05,
+      "loss": 3.386,
+      "step": 2249728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3159877434135917e-05,
+      "loss": 3.3727,
+      "step": 2250240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3151491486625397e-05,
+      "loss": 3.3749,
+      "step": 2250752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3143121917918607e-05,
+      "loss": 3.3882,
+      "step": 2251264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3134735970408087e-05,
+      "loss": 3.3715,
+      "step": 2251776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3126350022897569e-05,
+      "loss": 3.3805,
+      "step": 2252288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3117964075387049e-05,
+      "loss": 3.3791,
+      "step": 2252800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3109594506680258e-05,
+      "loss": 3.3748,
+      "step": 2253312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.310120855916974e-05,
+      "loss": 3.3714,
+      "step": 2253824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.309282261165922e-05,
+      "loss": 3.3746,
+      "step": 2254336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.30844366641487e-05,
+      "loss": 3.379,
+      "step": 2254848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.307606709544191e-05,
+      "loss": 3.3889,
+      "step": 2255360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.306768114793139e-05,
+      "loss": 3.3735,
+      "step": 2255872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.305929520042087e-05,
+      "loss": 3.3591,
+      "step": 2256384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.305090925291035e-05,
+      "loss": 3.3676,
+      "step": 2256896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3042539684203561e-05,
+      "loss": 3.3759,
+      "step": 2257408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3034153736693041e-05,
+      "loss": 3.3782,
+      "step": 2257920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3025767789182523e-05,
+      "loss": 3.3764,
+      "step": 2258432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3017381841672003e-05,
+      "loss": 3.3753,
+      "step": 2258944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3009012272965212e-05,
+      "loss": 3.3777,
+      "step": 2259456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3000626325454693e-05,
+      "loss": 3.3609,
+      "step": 2259968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2992240377944173e-05,
+      "loss": 3.3611,
+      "step": 2260480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2983854430433653e-05,
+      "loss": 3.3702,
+      "step": 2260992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2975484861726864e-05,
+      "loss": 3.373,
+      "step": 2261504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2967098914216344e-05,
+      "loss": 3.3836,
+      "step": 2262016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2958712966705824e-05,
+      "loss": 3.3683,
+      "step": 2262528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2950327019195304e-05,
+      "loss": 3.369,
+      "step": 2263040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2941941071684782e-05,
+      "loss": 3.3777,
+      "step": 2263552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2933571502977995e-05,
+      "loss": 3.3714,
+      "step": 2264064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2925185555467476e-05,
+      "loss": 3.3788,
+      "step": 2264576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2916799607956956e-05,
+      "loss": 3.3767,
+      "step": 2265088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2908413660446433e-05,
+      "loss": 3.3707,
+      "step": 2265600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2900044091739647e-05,
+      "loss": 3.3703,
+      "step": 2266112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2891658144229127e-05,
+      "loss": 3.3753,
+      "step": 2266624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2883272196718604e-05,
+      "loss": 3.3665,
+      "step": 2267136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2874886249208085e-05,
+      "loss": 3.3721,
+      "step": 2267648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2866516680501298e-05,
+      "loss": 3.3774,
+      "step": 2268160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2858130732990778e-05,
+      "loss": 3.3699,
+      "step": 2268672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2849744785480256e-05,
+      "loss": 3.3647,
+      "step": 2269184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2841358837969736e-05,
+      "loss": 3.3664,
+      "step": 2269696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2832972890459216e-05,
+      "loss": 3.3696,
+      "step": 2270208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2824603321752427e-05,
+      "loss": 3.3736,
+      "step": 2270720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2816217374241907e-05,
+      "loss": 3.3713,
+      "step": 2271232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2807831426731387e-05,
+      "loss": 3.3674,
+      "step": 2271744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2799445479220867e-05,
+      "loss": 3.3809,
+      "step": 2272256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2791075910514077e-05,
+      "loss": 3.3822,
+      "step": 2272768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2782689963003557e-05,
+      "loss": 3.3762,
+      "step": 2273280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2774304015493039e-05,
+      "loss": 3.379,
+      "step": 2273792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2765918067982519e-05,
+      "loss": 3.3754,
+      "step": 2274304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.275754849927573e-05,
+      "loss": 3.3704,
+      "step": 2274816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.274916255176521e-05,
+      "loss": 3.3637,
+      "step": 2275328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.274077660425469e-05,
+      "loss": 3.3716,
+      "step": 2275840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.273239065674417e-05,
+      "loss": 3.3671,
+      "step": 2276352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.272402108803738e-05,
+      "loss": 3.3753,
+      "step": 2276864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.271563514052686e-05,
+      "loss": 3.3808,
+      "step": 2277376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.270724919301634e-05,
+      "loss": 3.3762,
+      "step": 2277888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.269886324550582e-05,
+      "loss": 3.3717,
+      "step": 2278400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2690493676799031e-05,
+      "loss": 3.3694,
+      "step": 2278912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2682107729288511e-05,
+      "loss": 3.38,
+      "step": 2279424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2673721781777993e-05,
+      "loss": 3.3724,
+      "step": 2279936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2665335834267473e-05,
+      "loss": 3.3744,
+      "step": 2280448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2656966265560682e-05,
+      "loss": 3.3478,
+      "step": 2280960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2648580318050163e-05,
+      "loss": 3.3759,
+      "step": 2281472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2640194370539643e-05,
+      "loss": 3.3763,
+      "step": 2281984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2631808423029123e-05,
+      "loss": 3.3803,
+      "step": 2282496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2623438854322334e-05,
+      "loss": 3.3715,
+      "step": 2283008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2615052906811814e-05,
+      "loss": 3.3652,
+      "step": 2283520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2606666959301294e-05,
+      "loss": 3.3762,
+      "step": 2284032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2598281011790774e-05,
+      "loss": 3.3914,
+      "step": 2284544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2589895064280255e-05,
+      "loss": 3.3657,
+      "step": 2285056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2581525495573465e-05,
+      "loss": 3.3772,
+      "step": 2285568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2573139548062946e-05,
+      "loss": 3.3723,
+      "step": 2286080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2564753600552426e-05,
+      "loss": 3.3735,
+      "step": 2286592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2556367653041906e-05,
+      "loss": 3.3758,
+      "step": 2287104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2547998084335117e-05,
+      "loss": 3.3743,
+      "step": 2287616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2539612136824597e-05,
+      "loss": 3.3682,
+      "step": 2288128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2531226189314077e-05,
+      "loss": 3.383,
+      "step": 2288640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2522840241803555e-05,
+      "loss": 3.3741,
+      "step": 2289152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.886814832687378,
+      "eval_runtime": 308.8248,
+      "eval_samples_per_second": 1235.623,
+      "eval_steps_per_second": 38.614,
+      "step": 2289600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2514487051900497e-05,
+      "loss": 3.362,
+      "step": 2289664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2506101104389978e-05,
+      "loss": 3.3749,
+      "step": 2290176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2497715156879458e-05,
+      "loss": 3.3687,
+      "step": 2290688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2489329209368938e-05,
+      "loss": 3.3794,
+      "step": 2291200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.248095964066215e-05,
+      "loss": 3.3805,
+      "step": 2291712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2472573693151629e-05,
+      "loss": 3.3771,
+      "step": 2292224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2464187745641109e-05,
+      "loss": 3.3671,
+      "step": 2292736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2455801798130589e-05,
+      "loss": 3.3628,
+      "step": 2293248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.244741585062007e-05,
+      "loss": 3.3676,
+      "step": 2293760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.243904628191328e-05,
+      "loss": 3.3834,
+      "step": 2294272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2430660334402761e-05,
+      "loss": 3.3692,
+      "step": 2294784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2422274386892241e-05,
+      "loss": 3.3715,
+      "step": 2295296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.241388843938172e-05,
+      "loss": 3.3841,
+      "step": 2295808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.24055024918712e-05,
+      "loss": 3.3667,
+      "step": 2296320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2397132923164412e-05,
+      "loss": 3.3654,
+      "step": 2296832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2388746975653892e-05,
+      "loss": 3.3585,
+      "step": 2297344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.238036102814337e-05,
+      "loss": 3.3697,
+      "step": 2297856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2371975080632852e-05,
+      "loss": 3.3643,
+      "step": 2298368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2363605511926063e-05,
+      "loss": 3.3681,
+      "step": 2298880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2355219564415543e-05,
+      "loss": 3.3759,
+      "step": 2299392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2346833616905023e-05,
+      "loss": 3.3891,
+      "step": 2299904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2338447669394503e-05,
+      "loss": 3.3751,
+      "step": 2300416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2330078100687715e-05,
+      "loss": 3.3682,
+      "step": 2300928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2321692153177193e-05,
+      "loss": 3.3712,
+      "step": 2301440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2313306205666673e-05,
+      "loss": 3.3733,
+      "step": 2301952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2304920258156153e-05,
+      "loss": 3.3709,
+      "step": 2302464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2296550689449366e-05,
+      "loss": 3.3598,
+      "step": 2302976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2288164741938844e-05,
+      "loss": 3.3703,
+      "step": 2303488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2279778794428324e-05,
+      "loss": 3.3677,
+      "step": 2304000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2271392846917806e-05,
+      "loss": 3.3613,
+      "step": 2304512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2263006899407284e-05,
+      "loss": 3.3624,
+      "step": 2305024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2254637330700496e-05,
+      "loss": 3.3754,
+      "step": 2305536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2246251383189976e-05,
+      "loss": 3.368,
+      "step": 2306048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2237865435679456e-05,
+      "loss": 3.3782,
+      "step": 2306560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2229479488168934e-05,
+      "loss": 3.3721,
+      "step": 2307072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2221109919462147e-05,
+      "loss": 3.3687,
+      "step": 2307584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2212723971951627e-05,
+      "loss": 3.3773,
+      "step": 2308096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2204338024441107e-05,
+      "loss": 3.3602,
+      "step": 2308608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2195952076930587e-05,
+      "loss": 3.3642,
+      "step": 2309120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2187582508223798e-05,
+      "loss": 3.3621,
+      "step": 2309632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2179196560713278e-05,
+      "loss": 3.3538,
+      "step": 2310144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2170810613202758e-05,
+      "loss": 3.3696,
+      "step": 2310656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2162424665692237e-05,
+      "loss": 3.3745,
+      "step": 2311168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2154038718181717e-05,
+      "loss": 3.3675,
+      "step": 2311680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.214566914947493e-05,
+      "loss": 3.3691,
+      "step": 2312192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2137283201964408e-05,
+      "loss": 3.3675,
+      "step": 2312704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2128897254453888e-05,
+      "loss": 3.37,
+      "step": 2313216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.212051130694337e-05,
+      "loss": 3.3697,
+      "step": 2313728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2112125359432848e-05,
+      "loss": 3.3499,
+      "step": 2314240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2103755790726059e-05,
+      "loss": 3.3648,
+      "step": 2314752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.209536984321554e-05,
+      "loss": 3.3747,
+      "step": 2315264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.208698389570502e-05,
+      "loss": 3.3733,
+      "step": 2315776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2078597948194499e-05,
+      "loss": 3.3654,
+      "step": 2316288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2070228379487711e-05,
+      "loss": 3.357,
+      "step": 2316800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2061842431977191e-05,
+      "loss": 3.3644,
+      "step": 2317312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.205345648446667e-05,
+      "loss": 3.3585,
+      "step": 2317824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2045070536956151e-05,
+      "loss": 3.3666,
+      "step": 2318336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2036684589445631e-05,
+      "loss": 3.3592,
+      "step": 2318848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2028315020738842e-05,
+      "loss": 3.3692,
+      "step": 2319360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2019929073228322e-05,
+      "loss": 3.3662,
+      "step": 2319872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2011543125717802e-05,
+      "loss": 3.362,
+      "step": 2320384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2003157178207282e-05,
+      "loss": 3.3561,
+      "step": 2320896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1994787609500494e-05,
+      "loss": 3.3693,
+      "step": 2321408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1986401661989972e-05,
+      "loss": 3.3544,
+      "step": 2321920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1978015714479452e-05,
+      "loss": 3.3581,
+      "step": 2322432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1969629766968932e-05,
+      "loss": 3.3704,
+      "step": 2322944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1961260198262143e-05,
+      "loss": 3.3686,
+      "step": 2323456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1952874250751623e-05,
+      "loss": 3.3592,
+      "step": 2323968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1944488303241105e-05,
+      "loss": 3.3497,
+      "step": 2324480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1936102355730585e-05,
+      "loss": 3.366,
+      "step": 2324992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1927732787023794e-05,
+      "loss": 3.3674,
+      "step": 2325504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1919346839513275e-05,
+      "loss": 3.3713,
+      "step": 2326016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1910960892002755e-05,
+      "loss": 3.3624,
+      "step": 2326528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1902574944492234e-05,
+      "loss": 3.3674,
+      "step": 2327040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1894205375785446e-05,
+      "loss": 3.3722,
+      "step": 2327552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1885819428274926e-05,
+      "loss": 3.3669,
+      "step": 2328064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1877433480764406e-05,
+      "loss": 3.3689,
+      "step": 2328576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1869047533253886e-05,
+      "loss": 3.3654,
+      "step": 2329088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1860661585743366e-05,
+      "loss": 3.3703,
+      "step": 2329600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1852292017036577e-05,
+      "loss": 3.3618,
+      "step": 2330112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1843906069526058e-05,
+      "loss": 3.3652,
+      "step": 2330624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1835520122015537e-05,
+      "loss": 3.3673,
+      "step": 2331136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1827134174505017e-05,
+      "loss": 3.379,
+      "step": 2331648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1818748226994497e-05,
+      "loss": 3.3639,
+      "step": 2332160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1810378658287707e-05,
+      "loss": 3.3533,
+      "step": 2332672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1801992710777187e-05,
+      "loss": 3.3608,
+      "step": 2333184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1793606763266667e-05,
+      "loss": 3.3683,
+      "step": 2333696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1785220815756149e-05,
+      "loss": 3.367,
+      "step": 2334208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1776851247049358e-05,
+      "loss": 3.3669,
+      "step": 2334720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.176846529953884e-05,
+      "loss": 3.3671,
+      "step": 2335232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.176007935202832e-05,
+      "loss": 3.3681,
+      "step": 2335744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1751693404517798e-05,
+      "loss": 3.3517,
+      "step": 2336256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.174332383581101e-05,
+      "loss": 3.351,
+      "step": 2336768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.173493788830049e-05,
+      "loss": 3.3562,
+      "step": 2337280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.172655194078997e-05,
+      "loss": 3.3629,
+      "step": 2337792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1718165993279449e-05,
+      "loss": 3.3744,
+      "step": 2338304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1709796424572661e-05,
+      "loss": 3.3598,
+      "step": 2338816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1701410477062141e-05,
+      "loss": 3.3534,
+      "step": 2339328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1693024529551621e-05,
+      "loss": 3.372,
+      "step": 2339840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1684638582041101e-05,
+      "loss": 3.3607,
+      "step": 2340352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1676269013334312e-05,
+      "loss": 3.3722,
+      "step": 2340864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1667883065823793e-05,
+      "loss": 3.3671,
+      "step": 2341376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1659497118313272e-05,
+      "loss": 3.3602,
+      "step": 2341888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1651111170802752e-05,
+      "loss": 3.3629,
+      "step": 2342400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1642741602095964e-05,
+      "loss": 3.3657,
+      "step": 2342912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1634355654585444e-05,
+      "loss": 3.3527,
+      "step": 2343424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1625969707074922e-05,
+      "loss": 3.3638,
+      "step": 2343936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1617583759564402e-05,
+      "loss": 3.3655,
+      "step": 2344448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1609214190857615e-05,
+      "loss": 3.3617,
+      "step": 2344960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1600828243347095e-05,
+      "loss": 3.355,
+      "step": 2345472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1592442295836575e-05,
+      "loss": 3.3607,
+      "step": 2345984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1584056348326055e-05,
+      "loss": 3.3557,
+      "step": 2346496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1575686779619265e-05,
+      "loss": 3.3644,
+      "step": 2347008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1567300832108745e-05,
+      "loss": 3.3666,
+      "step": 2347520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1558914884598225e-05,
+      "loss": 3.3562,
+      "step": 2348032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1550528937087705e-05,
+      "loss": 3.3727,
+      "step": 2348544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1542142989577183e-05,
+      "loss": 3.3738,
+      "step": 2349056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1533773420870396e-05,
+      "loss": 3.3659,
+      "step": 2349568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1525387473359876e-05,
+      "loss": 3.3688,
+      "step": 2350080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1517001525849356e-05,
+      "loss": 3.3678,
+      "step": 2350592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1508615578338836e-05,
+      "loss": 3.3604,
+      "step": 2351104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1500246009632047e-05,
+      "loss": 3.3504,
+      "step": 2351616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1491860062121528e-05,
+      "loss": 3.3637,
+      "step": 2352128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1483474114611008e-05,
+      "loss": 3.36,
+      "step": 2352640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1475088167100486e-05,
+      "loss": 3.3643,
+      "step": 2353152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1466718598393699e-05,
+      "loss": 3.3667,
+      "step": 2353664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1458332650883179e-05,
+      "loss": 3.3683,
+      "step": 2354176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1449946703372657e-05,
+      "loss": 3.3595,
+      "step": 2354688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1441560755862137e-05,
+      "loss": 3.3585,
+      "step": 2355200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.143319118715535e-05,
+      "loss": 3.3703,
+      "step": 2355712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.142480523964483e-05,
+      "loss": 3.3664,
+      "step": 2356224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.141641929213431e-05,
+      "loss": 3.3644,
+      "step": 2356736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.140803334462379e-05,
+      "loss": 3.3427,
+      "step": 2357248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1399663775917e-05,
+      "loss": 3.3613,
+      "step": 2357760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1391277828406482e-05,
+      "loss": 3.3675,
+      "step": 2358272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.138289188089596e-05,
+      "loss": 3.3725,
+      "step": 2358784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.137450593338544e-05,
+      "loss": 3.3637,
+      "step": 2359296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1366136364678653e-05,
+      "loss": 3.3538,
+      "step": 2359808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1357750417168131e-05,
+      "loss": 3.3674,
+      "step": 2360320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1349364469657611e-05,
+      "loss": 3.379,
+      "step": 2360832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.134097852214709e-05,
+      "loss": 3.3603,
+      "step": 2361344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1332608953440303e-05,
+      "loss": 3.3635,
+      "step": 2361856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1324223005929782e-05,
+      "loss": 3.3641,
+      "step": 2362368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1315837058419263e-05,
+      "loss": 3.3675,
+      "step": 2362880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1307451110908743e-05,
+      "loss": 3.3638,
+      "step": 2363392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1299081542201954e-05,
+      "loss": 3.3637,
+      "step": 2363904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1290695594691434e-05,
+      "loss": 3.3571,
+      "step": 2364416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1282309647180914e-05,
+      "loss": 3.3757,
+      "step": 2364928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1273923699670394e-05,
+      "loss": 3.3619,
+      "step": 2365440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.888003349304199,
+      "eval_runtime": 304.2968,
+      "eval_samples_per_second": 1254.009,
+      "eval_steps_per_second": 39.189,
+      "step": 2365920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1265554130963605e-05,
+      "loss": 3.3569,
+      "step": 2365952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1257168183453085e-05,
+      "loss": 3.366,
+      "step": 2366464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1248782235942565e-05,
+      "loss": 3.3584,
+      "step": 2366976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1240396288432044e-05,
+      "loss": 3.37,
+      "step": 2367488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1232026719725255e-05,
+      "loss": 3.3708,
+      "step": 2368000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1223640772214735e-05,
+      "loss": 3.3711,
+      "step": 2368512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1215254824704217e-05,
+      "loss": 3.3578,
+      "step": 2369024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1206868877193695e-05,
+      "loss": 3.3525,
+      "step": 2369536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1198499308486906e-05,
+      "loss": 3.3536,
+      "step": 2370048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1190113360976388e-05,
+      "loss": 3.3797,
+      "step": 2370560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1181727413465868e-05,
+      "loss": 3.3587,
+      "step": 2371072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1173341465955346e-05,
+      "loss": 3.3627,
+      "step": 2371584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1164971897248558e-05,
+      "loss": 3.3686,
+      "step": 2372096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1156585949738038e-05,
+      "loss": 3.3639,
+      "step": 2372608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1148200002227518e-05,
+      "loss": 3.3515,
+      "step": 2373120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1139814054716998e-05,
+      "loss": 3.3527,
+      "step": 2373632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1131444486010209e-05,
+      "loss": 3.3546,
+      "step": 2374144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1123058538499689e-05,
+      "loss": 3.3565,
+      "step": 2374656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1114672590989169e-05,
+      "loss": 3.3603,
+      "step": 2375168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1106286643478649e-05,
+      "loss": 3.3675,
+      "step": 2375680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.109791707477186e-05,
+      "loss": 3.3796,
+      "step": 2376192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1089531127261341e-05,
+      "loss": 3.3624,
+      "step": 2376704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.108114517975082e-05,
+      "loss": 3.3609,
+      "step": 2377216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.10727592322403e-05,
+      "loss": 3.3574,
+      "step": 2377728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1064389663533512e-05,
+      "loss": 3.3671,
+      "step": 2378240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1056003716022992e-05,
+      "loss": 3.3576,
+      "step": 2378752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.104761776851247e-05,
+      "loss": 3.351,
+      "step": 2379264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1039231821001952e-05,
+      "loss": 3.3578,
+      "step": 2379776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1030862252295163e-05,
+      "loss": 3.3595,
+      "step": 2380288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1022476304784641e-05,
+      "loss": 3.3522,
+      "step": 2380800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1014090357274123e-05,
+      "loss": 3.3512,
+      "step": 2381312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1005704409763602e-05,
+      "loss": 3.3666,
+      "step": 2381824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0997334841056813e-05,
+      "loss": 3.3592,
+      "step": 2382336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0988948893546293e-05,
+      "loss": 3.3629,
+      "step": 2382848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0980562946035773e-05,
+      "loss": 3.37,
+      "step": 2383360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0972176998525253e-05,
+      "loss": 3.3597,
+      "step": 2383872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0963791051014733e-05,
+      "loss": 3.3644,
+      "step": 2384384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0955421482307944e-05,
+      "loss": 3.3521,
+      "step": 2384896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0947035534797424e-05,
+      "loss": 3.3545,
+      "step": 2385408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0938649587286905e-05,
+      "loss": 3.3519,
+      "step": 2385920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0930263639776384e-05,
+      "loss": 3.3508,
+      "step": 2386432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0921894071069595e-05,
+      "loss": 3.354,
+      "step": 2386944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0913508123559076e-05,
+      "loss": 3.3676,
+      "step": 2387456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0905122176048556e-05,
+      "loss": 3.3519,
+      "step": 2387968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0896736228538034e-05,
+      "loss": 3.3625,
+      "step": 2388480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0888366659831247e-05,
+      "loss": 3.3574,
+      "step": 2388992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0879980712320727e-05,
+      "loss": 3.3594,
+      "step": 2389504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0871594764810205e-05,
+      "loss": 3.3613,
+      "step": 2390016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0863208817299687e-05,
+      "loss": 3.3419,
+      "step": 2390528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0854839248592898e-05,
+      "loss": 3.3539,
+      "step": 2391040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0846453301082377e-05,
+      "loss": 3.362,
+      "step": 2391552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0838067353571857e-05,
+      "loss": 3.3671,
+      "step": 2392064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0829681406061337e-05,
+      "loss": 3.3523,
+      "step": 2392576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0821295458550817e-05,
+      "loss": 3.3508,
+      "step": 2393088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0812925889844028e-05,
+      "loss": 3.3531,
+      "step": 2393600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0804539942333508e-05,
+      "loss": 3.3478,
+      "step": 2394112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0796153994822988e-05,
+      "loss": 3.3564,
+      "step": 2394624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0787768047312468e-05,
+      "loss": 3.3502,
+      "step": 2395136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0779398478605679e-05,
+      "loss": 3.361,
+      "step": 2395648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0771012531095159e-05,
+      "loss": 3.3537,
+      "step": 2396160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.076262658358464e-05,
+      "loss": 3.3538,
+      "step": 2396672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0754240636074119e-05,
+      "loss": 3.3459,
+      "step": 2397184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.074587106736733e-05,
+      "loss": 3.3611,
+      "step": 2397696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0737485119856811e-05,
+      "loss": 3.343,
+      "step": 2398208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0729099172346291e-05,
+      "loss": 3.3519,
+      "step": 2398720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.072071322483577e-05,
+      "loss": 3.3603,
+      "step": 2399232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.071232727732525e-05,
+      "loss": 3.3598,
+      "step": 2399744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0703957708618462e-05,
+      "loss": 3.346,
+      "step": 2400256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0695571761107942e-05,
+      "loss": 3.3441,
+      "step": 2400768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0687185813597422e-05,
+      "loss": 3.3552,
+      "step": 2401280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0678799866086902e-05,
+      "loss": 3.3545,
+      "step": 2401792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0670430297380112e-05,
+      "loss": 3.3652,
+      "step": 2402304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0662044349869592e-05,
+      "loss": 3.3504,
+      "step": 2402816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0653658402359072e-05,
+      "loss": 3.3627,
+      "step": 2403328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0645272454848552e-05,
+      "loss": 3.3601,
+      "step": 2403840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0636902886141765e-05,
+      "loss": 3.3607,
+      "step": 2404352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0628516938631243e-05,
+      "loss": 3.3572,
+      "step": 2404864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0620130991120723e-05,
+      "loss": 3.356,
+      "step": 2405376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0611745043610203e-05,
+      "loss": 3.3599,
+      "step": 2405888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0603375474903415e-05,
+      "loss": 3.3482,
+      "step": 2406400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0594989527392894e-05,
+      "loss": 3.3557,
+      "step": 2406912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0586603579882375e-05,
+      "loss": 3.3565,
+      "step": 2407424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0578217632371855e-05,
+      "loss": 3.3682,
+      "step": 2407936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0569831684861334e-05,
+      "loss": 3.3518,
+      "step": 2408448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0561462116154546e-05,
+      "loss": 3.3498,
+      "step": 2408960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0553076168644026e-05,
+      "loss": 3.3445,
+      "step": 2409472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0544690221133506e-05,
+      "loss": 3.3591,
+      "step": 2409984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0536304273622984e-05,
+      "loss": 3.3542,
+      "step": 2410496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0527934704916197e-05,
+      "loss": 3.3633,
+      "step": 2411008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0519548757405677e-05,
+      "loss": 3.3555,
+      "step": 2411520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0511162809895157e-05,
+      "loss": 3.3564,
+      "step": 2412032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0502776862384637e-05,
+      "loss": 3.3431,
+      "step": 2412544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0494407293677847e-05,
+      "loss": 3.3465,
+      "step": 2413056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0486021346167329e-05,
+      "loss": 3.3447,
+      "step": 2413568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0477635398656807e-05,
+      "loss": 3.3514,
+      "step": 2414080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0469249451146287e-05,
+      "loss": 3.3633,
+      "step": 2414592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.04608798824395e-05,
+      "loss": 3.3548,
+      "step": 2415104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.045249393492898e-05,
+      "loss": 3.3457,
+      "step": 2415616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0444107987418458e-05,
+      "loss": 3.3587,
+      "step": 2416128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0435722039907938e-05,
+      "loss": 3.3543,
+      "step": 2416640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.042735247120115e-05,
+      "loss": 3.3577,
+      "step": 2417152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0418966523690629e-05,
+      "loss": 3.3577,
+      "step": 2417664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.041058057618011e-05,
+      "loss": 3.3515,
+      "step": 2418176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.040219462866959e-05,
+      "loss": 3.3534,
+      "step": 2418688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0393825059962801e-05,
+      "loss": 3.359,
+      "step": 2419200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0385439112452281e-05,
+      "loss": 3.3404,
+      "step": 2419712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0377053164941761e-05,
+      "loss": 3.3542,
+      "step": 2420224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0368667217431241e-05,
+      "loss": 3.3577,
+      "step": 2420736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.036028126992072e-05,
+      "loss": 3.3489,
+      "step": 2421248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0351911701213932e-05,
+      "loss": 3.3469,
+      "step": 2421760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0343525753703412e-05,
+      "loss": 3.3526,
+      "step": 2422272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0335139806192892e-05,
+      "loss": 3.3453,
+      "step": 2422784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0326753858682372e-05,
+      "loss": 3.3563,
+      "step": 2423296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0318384289975582e-05,
+      "loss": 3.3545,
+      "step": 2423808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0309998342465064e-05,
+      "loss": 3.3459,
+      "step": 2424320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0301612394954544e-05,
+      "loss": 3.3655,
+      "step": 2424832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0293226447444022e-05,
+      "loss": 3.364,
+      "step": 2425344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0284856878737235e-05,
+      "loss": 3.3548,
+      "step": 2425856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0276470931226715e-05,
+      "loss": 3.3601,
+      "step": 2426368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0268084983716193e-05,
+      "loss": 3.3549,
+      "step": 2426880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0259699036205673e-05,
+      "loss": 3.3517,
+      "step": 2427392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0251329467498885e-05,
+      "loss": 3.3468,
+      "step": 2427904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0242943519988365e-05,
+      "loss": 3.3483,
+      "step": 2428416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0234557572477845e-05,
+      "loss": 3.3532,
+      "step": 2428928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0226171624967325e-05,
+      "loss": 3.3521,
+      "step": 2429440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0217802056260536e-05,
+      "loss": 3.3539,
+      "step": 2429952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0209416108750016e-05,
+      "loss": 3.3597,
+      "step": 2430464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0201030161239496e-05,
+      "loss": 3.3528,
+      "step": 2430976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0192644213728976e-05,
+      "loss": 3.3483,
+      "step": 2431488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0184274645022188e-05,
+      "loss": 3.3567,
+      "step": 2432000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0175888697511667e-05,
+      "loss": 3.3575,
+      "step": 2432512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0167502750001147e-05,
+      "loss": 3.3539,
+      "step": 2433024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0159116802490626e-05,
+      "loss": 3.3368,
+      "step": 2433536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0150747233783839e-05,
+      "loss": 3.3508,
+      "step": 2434048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0142361286273317e-05,
+      "loss": 3.3593,
+      "step": 2434560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0133975338762799e-05,
+      "loss": 3.3596,
+      "step": 2435072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0125589391252279e-05,
+      "loss": 3.3513,
+      "step": 2435584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0117219822545488e-05,
+      "loss": 3.3433,
+      "step": 2436096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.010883387503497e-05,
+      "loss": 3.3581,
+      "step": 2436608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.010044792752445e-05,
+      "loss": 3.3674,
+      "step": 2437120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.009206198001393e-05,
+      "loss": 3.3529,
+      "step": 2437632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.008369241130714e-05,
+      "loss": 3.3569,
+      "step": 2438144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.007530646379662e-05,
+      "loss": 3.353,
+      "step": 2438656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.00669205162861e-05,
+      "loss": 3.3538,
+      "step": 2439168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.005853456877558e-05,
+      "loss": 3.3505,
+      "step": 2439680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0050165000068791e-05,
+      "loss": 3.3614,
+      "step": 2440192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0041779052558271e-05,
+      "loss": 3.3467,
+      "step": 2440704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0033393105047751e-05,
+      "loss": 3.3649,
+      "step": 2441216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.002500715753723e-05,
+      "loss": 3.3532,
+      "step": 2441728
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.0016637588830442e-05,
+      "loss": 3.3485,
+      "step": 2442240
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.887634754180908,
+      "eval_runtime": 304.291,
+      "eval_samples_per_second": 1254.033,
+      "eval_steps_per_second": 39.189,
+      "step": 2442240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0008268020123654e-05,
+      "loss": 3.3562,
+      "step": 2442752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.999882072613134e-06,
+      "loss": 3.3511,
+      "step": 2443264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.991496125102614e-06,
+      "loss": 3.3599,
+      "step": 2443776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.983110177592094e-06,
+      "loss": 3.3584,
+      "step": 2444288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.974740608885305e-06,
+      "loss": 3.3616,
+      "step": 2444800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.966354661374785e-06,
+      "loss": 3.3499,
+      "step": 2445312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.957968713864265e-06,
+      "loss": 3.3412,
+      "step": 2445824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.949582766353745e-06,
+      "loss": 3.345,
+      "step": 2446336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.941213197646957e-06,
+      "loss": 3.3683,
+      "step": 2446848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.932827250136435e-06,
+      "loss": 3.3527,
+      "step": 2447360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.924441302625915e-06,
+      "loss": 3.3498,
+      "step": 2447872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.916055355115395e-06,
+      "loss": 3.3579,
+      "step": 2448384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.907685786408608e-06,
+      "loss": 3.3548,
+      "step": 2448896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.899299838898086e-06,
+      "loss": 3.3429,
+      "step": 2449408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.890913891387566e-06,
+      "loss": 3.3453,
+      "step": 2449920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.882527943877048e-06,
+      "loss": 3.3436,
+      "step": 2450432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.874158375170258e-06,
+      "loss": 3.3464,
+      "step": 2450944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.865772427659738e-06,
+      "loss": 3.3483,
+      "step": 2451456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.857386480149218e-06,
+      "loss": 3.3599,
+      "step": 2451968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.849000532638698e-06,
+      "loss": 3.3711,
+      "step": 2452480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.840630963931909e-06,
+      "loss": 3.3527,
+      "step": 2452992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.832245016421389e-06,
+      "loss": 3.3457,
+      "step": 2453504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.823859068910869e-06,
+      "loss": 3.3518,
+      "step": 2454016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.815473121400349e-06,
+      "loss": 3.3577,
+      "step": 2454528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.80710355269356e-06,
+      "loss": 3.3441,
+      "step": 2455040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.79871760518304e-06,
+      "loss": 3.3454,
+      "step": 2455552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.79033165767252e-06,
+      "loss": 3.3478,
+      "step": 2456064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.781945710162e-06,
+      "loss": 3.3569,
+      "step": 2456576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.77355976265148e-06,
+      "loss": 3.3391,
+      "step": 2457088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.765190193944692e-06,
+      "loss": 3.3399,
+      "step": 2457600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.756804246434172e-06,
+      "loss": 3.3551,
+      "step": 2458112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.74841829892365e-06,
+      "loss": 3.351,
+      "step": 2458624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.74003235141313e-06,
+      "loss": 3.3548,
+      "step": 2459136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.731662782706343e-06,
+      "loss": 3.3591,
+      "step": 2459648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.723276835195823e-06,
+      "loss": 3.3466,
+      "step": 2460160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.714890887685301e-06,
+      "loss": 3.3551,
+      "step": 2460672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.706504940174783e-06,
+      "loss": 3.3415,
+      "step": 2461184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.698118992664263e-06,
+      "loss": 3.3468,
+      "step": 2461696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.689749423957473e-06,
+      "loss": 3.3412,
+      "step": 2462208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.681363476446953e-06,
+      "loss": 3.3407,
+      "step": 2462720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.672977528936433e-06,
+      "loss": 3.3478,
+      "step": 2463232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.664591581425913e-06,
+      "loss": 3.3593,
+      "step": 2463744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.656222012719124e-06,
+      "loss": 3.3466,
+      "step": 2464256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.647836065208604e-06,
+      "loss": 3.3496,
+      "step": 2464768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.639450117698084e-06,
+      "loss": 3.3485,
+      "step": 2465280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.631064170187564e-06,
+      "loss": 3.3496,
+      "step": 2465792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.622694601480775e-06,
+      "loss": 3.3521,
+      "step": 2466304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.614308653970255e-06,
+      "loss": 3.3366,
+      "step": 2466816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.605922706459736e-06,
+      "loss": 3.3397,
+      "step": 2467328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.597536758949214e-06,
+      "loss": 3.3522,
+      "step": 2467840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.589150811438694e-06,
+      "loss": 3.3585,
+      "step": 2468352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.580781242731907e-06,
+      "loss": 3.3417,
+      "step": 2468864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.572395295221385e-06,
+      "loss": 3.3437,
+      "step": 2469376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.564009347710865e-06,
+      "loss": 3.3464,
+      "step": 2469888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.555623400200347e-06,
+      "loss": 3.3365,
+      "step": 2470400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.547253831493558e-06,
+      "loss": 3.3462,
+      "step": 2470912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.538867883983038e-06,
+      "loss": 3.3402,
+      "step": 2471424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.530481936472518e-06,
+      "loss": 3.3494,
+      "step": 2471936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.522095988961997e-06,
+      "loss": 3.3472,
+      "step": 2472448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.513726420255208e-06,
+      "loss": 3.3446,
+      "step": 2472960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.505340472744688e-06,
+      "loss": 3.3346,
+      "step": 2473472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.496954525234168e-06,
+      "loss": 3.3513,
+      "step": 2473984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.488568577723648e-06,
+      "loss": 3.3334,
+      "step": 2474496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.480199009016859e-06,
+      "loss": 3.3437,
+      "step": 2475008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.471813061506339e-06,
+      "loss": 3.3503,
+      "step": 2475520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.463427113995819e-06,
+      "loss": 3.3506,
+      "step": 2476032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.4550411664853e-06,
+      "loss": 3.3344,
+      "step": 2476544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.44667159777851e-06,
+      "loss": 3.336,
+      "step": 2477056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.43828565026799e-06,
+      "loss": 3.3399,
+      "step": 2477568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.429899702757471e-06,
+      "loss": 3.348,
+      "step": 2478080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.42151375524695e-06,
+      "loss": 3.3462,
+      "step": 2478592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.413144186540162e-06,
+      "loss": 3.3431,
+      "step": 2479104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.404758239029642e-06,
+      "loss": 3.3515,
+      "step": 2479616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.396372291519122e-06,
+      "loss": 3.3548,
+      "step": 2480128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.3879863440086e-06,
+      "loss": 3.3471,
+      "step": 2480640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.379600396498082e-06,
+      "loss": 3.3462,
+      "step": 2481152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.371230827791293e-06,
+      "loss": 3.3473,
+      "step": 2481664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.362844880280772e-06,
+      "loss": 3.3556,
+      "step": 2482176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.354458932770252e-06,
+      "loss": 3.3348,
+      "step": 2482688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.346089364063463e-06,
+      "loss": 3.3468,
+      "step": 2483200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.337703416552943e-06,
+      "loss": 3.3461,
+      "step": 2483712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.329317469042423e-06,
+      "loss": 3.3609,
+      "step": 2484224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.320931521531903e-06,
+      "loss": 3.3465,
+      "step": 2484736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.312561952825116e-06,
+      "loss": 3.3362,
+      "step": 2485248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.304176005314596e-06,
+      "loss": 3.3369,
+      "step": 2485760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.295790057804074e-06,
+      "loss": 3.3443,
+      "step": 2486272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.287404110293554e-06,
+      "loss": 3.3508,
+      "step": 2486784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.279034541586766e-06,
+      "loss": 3.351,
+      "step": 2487296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.270648594076246e-06,
+      "loss": 3.3466,
+      "step": 2487808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.262262646565724e-06,
+      "loss": 3.3448,
+      "step": 2488320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.253876699055206e-06,
+      "loss": 3.3369,
+      "step": 2488832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.245507130348417e-06,
+      "loss": 3.3332,
+      "step": 2489344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.237121182837897e-06,
+      "loss": 3.331,
+      "step": 2489856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.228735235327377e-06,
+      "loss": 3.3472,
+      "step": 2490368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.220349287816857e-06,
+      "loss": 3.3511,
+      "step": 2490880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.21197971911007e-06,
+      "loss": 3.3451,
+      "step": 2491392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.203593771599548e-06,
+      "loss": 3.3386,
+      "step": 2491904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.195207824089027e-06,
+      "loss": 3.3462,
+      "step": 2492416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.186821876578507e-06,
+      "loss": 3.3447,
+      "step": 2492928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.17845230787172e-06,
+      "loss": 3.3514,
+      "step": 2493440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.170066360361198e-06,
+      "loss": 3.347,
+      "step": 2493952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.161680412850678e-06,
+      "loss": 3.3429,
+      "step": 2494464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.15329446534016e-06,
+      "loss": 3.3412,
+      "step": 2494976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.144924896633369e-06,
+      "loss": 3.35,
+      "step": 2495488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.13653894912285e-06,
+      "loss": 3.3348,
+      "step": 2496000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.12815300161233e-06,
+      "loss": 3.3425,
+      "step": 2496512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.11976705410181e-06,
+      "loss": 3.3498,
+      "step": 2497024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.111381106591289e-06,
+      "loss": 3.3403,
+      "step": 2497536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.103011537884501e-06,
+      "loss": 3.3357,
+      "step": 2498048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.094625590373981e-06,
+      "loss": 3.3451,
+      "step": 2498560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.08623964286346e-06,
+      "loss": 3.34,
+      "step": 2499072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.077853695352941e-06,
+      "loss": 3.3433,
+      "step": 2499584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.069484126646152e-06,
+      "loss": 3.3436,
+      "step": 2500096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.061098179135632e-06,
+      "loss": 3.3404,
+      "step": 2500608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.052712231625112e-06,
+      "loss": 3.3531,
+      "step": 2501120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.044326284114592e-06,
+      "loss": 3.3542,
+      "step": 2501632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.035956715407804e-06,
+      "loss": 3.3457,
+      "step": 2502144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.027570767897284e-06,
+      "loss": 3.3542,
+      "step": 2502656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.019184820386762e-06,
+      "loss": 3.3462,
+      "step": 2503168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.010798872876242e-06,
+      "loss": 3.3417,
+      "step": 2503680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.002429304169455e-06,
+      "loss": 3.3359,
+      "step": 2504192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.994043356658933e-06,
+      "loss": 3.3404,
+      "step": 2504704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.985657409148413e-06,
+      "loss": 3.3444,
+      "step": 2505216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.977271461637895e-06,
+      "loss": 3.3428,
+      "step": 2505728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.968901892931106e-06,
+      "loss": 3.3469,
+      "step": 2506240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.960515945420585e-06,
+      "loss": 3.348,
+      "step": 2506752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.952129997910065e-06,
+      "loss": 3.3451,
+      "step": 2507264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.943744050399545e-06,
+      "loss": 3.3376,
+      "step": 2507776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.935374481692758e-06,
+      "loss": 3.3465,
+      "step": 2508288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.926988534182236e-06,
+      "loss": 3.3502,
+      "step": 2508800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.918602586671716e-06,
+      "loss": 3.3417,
+      "step": 2509312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.910216639161196e-06,
+      "loss": 3.3297,
+      "step": 2509824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.901847070454407e-06,
+      "loss": 3.339,
+      "step": 2510336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.893461122943887e-06,
+      "loss": 3.3534,
+      "step": 2510848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.885075175433367e-06,
+      "loss": 3.3522,
+      "step": 2511360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.876689227922847e-06,
+      "loss": 3.3361,
+      "step": 2511872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.868319659216057e-06,
+      "loss": 3.3379,
+      "step": 2512384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.859933711705539e-06,
+      "loss": 3.3523,
+      "step": 2512896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.851547764195019e-06,
+      "loss": 3.3526,
+      "step": 2513408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.843161816684497e-06,
+      "loss": 3.3473,
+      "step": 2513920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.83479224797771e-06,
+      "loss": 3.3448,
+      "step": 2514432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.82640630046719e-06,
+      "loss": 3.3454,
+      "step": 2514944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.81802035295667e-06,
+      "loss": 3.3439,
+      "step": 2515456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.809634405446148e-06,
+      "loss": 3.3375,
+      "step": 2515968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.80126483673936e-06,
+      "loss": 3.3524,
+      "step": 2516480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.79287888922884e-06,
+      "loss": 3.3405,
+      "step": 2516992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.78449294171832e-06,
+      "loss": 3.3548,
+      "step": 2517504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.7761069942078e-06,
+      "loss": 3.3444,
+      "step": 2518016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.767737425501011e-06,
+      "loss": 3.3367,
+      "step": 2518528
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8876395225524902,
+      "eval_runtime": 303.3322,
+      "eval_samples_per_second": 1257.997,
+      "eval_steps_per_second": 39.313,
+      "step": 2518560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.759351477990493e-06,
+      "loss": 3.345,
+      "step": 2519040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.750965530479971e-06,
+      "loss": 3.3433,
+      "step": 2519552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.742579582969451e-06,
+      "loss": 3.3492,
+      "step": 2520064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.734210014262664e-06,
+      "loss": 3.3471,
+      "step": 2520576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.725824066752143e-06,
+      "loss": 3.3536,
+      "step": 2521088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.717438119241622e-06,
+      "loss": 3.3396,
+      "step": 2521600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.709052171731102e-06,
+      "loss": 3.335,
+      "step": 2522112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.700666224220583e-06,
+      "loss": 3.3292,
+      "step": 2522624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.692296655513792e-06,
+      "loss": 3.3604,
+      "step": 2523136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.683910708003274e-06,
+      "loss": 3.3458,
+      "step": 2523648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.675524760492754e-06,
+      "loss": 3.3384,
+      "step": 2524160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.667138812982234e-06,
+      "loss": 3.3522,
+      "step": 2524672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.658769244275445e-06,
+      "loss": 3.3427,
+      "step": 2525184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.650383296764925e-06,
+      "loss": 3.3365,
+      "step": 2525696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.641997349254405e-06,
+      "loss": 3.3353,
+      "step": 2526208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.633611401743883e-06,
+      "loss": 3.3389,
+      "step": 2526720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.625241833037095e-06,
+      "loss": 3.3348,
+      "step": 2527232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.616855885526575e-06,
+      "loss": 3.3363,
+      "step": 2527744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.608469938016055e-06,
+      "loss": 3.3524,
+      "step": 2528256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.600083990505535e-06,
+      "loss": 3.361,
+      "step": 2528768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.591714421798746e-06,
+      "loss": 3.3443,
+      "step": 2529280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.583328474288228e-06,
+      "loss": 3.3365,
+      "step": 2529792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.574942526777708e-06,
+      "loss": 3.342,
+      "step": 2530304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.566556579267186e-06,
+      "loss": 3.3464,
+      "step": 2530816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.558187010560398e-06,
+      "loss": 3.3416,
+      "step": 2531328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.549801063049878e-06,
+      "loss": 3.3347,
+      "step": 2531840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.541415115539357e-06,
+      "loss": 3.339,
+      "step": 2532352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.533029168028837e-06,
+      "loss": 3.3495,
+      "step": 2532864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.524659599322049e-06,
+      "loss": 3.3254,
+      "step": 2533376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.516273651811529e-06,
+      "loss": 3.3303,
+      "step": 2533888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.507887704301009e-06,
+      "loss": 3.3449,
+      "step": 2534400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.499501756790489e-06,
+      "loss": 3.3427,
+      "step": 2534912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.4911321880837e-06,
+      "loss": 3.3493,
+      "step": 2535424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.482746240573181e-06,
+      "loss": 3.3439,
+      "step": 2535936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.47436029306266e-06,
+      "loss": 3.3377,
+      "step": 2536448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.46597434555214e-06,
+      "loss": 3.3454,
+      "step": 2536960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.457604776845352e-06,
+      "loss": 3.3295,
+      "step": 2537472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.44921882933483e-06,
+      "loss": 3.3405,
+      "step": 2537984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.44083288182431e-06,
+      "loss": 3.3299,
+      "step": 2538496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.43244693431379e-06,
+      "loss": 3.3347,
+      "step": 2539008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.424077365607003e-06,
+      "loss": 3.3378,
+      "step": 2539520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.415691418096481e-06,
+      "loss": 3.3466,
+      "step": 2540032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.407305470585963e-06,
+      "loss": 3.3411,
+      "step": 2540544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.398919523075443e-06,
+      "loss": 3.3373,
+      "step": 2541056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.390549954368653e-06,
+      "loss": 3.3411,
+      "step": 2541568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.382164006858133e-06,
+      "loss": 3.341,
+      "step": 2542080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.373778059347613e-06,
+      "loss": 3.3422,
+      "step": 2542592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.365392111837093e-06,
+      "loss": 3.3291,
+      "step": 2543104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.357022543130304e-06,
+      "loss": 3.3297,
+      "step": 2543616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.348636595619784e-06,
+      "loss": 3.3412,
+      "step": 2544128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.340250648109264e-06,
+      "loss": 3.3457,
+      "step": 2544640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.331864700598744e-06,
+      "loss": 3.3351,
+      "step": 2545152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.323495131891955e-06,
+      "loss": 3.3336,
+      "step": 2545664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.315109184381435e-06,
+      "loss": 3.3318,
+      "step": 2546176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.306723236870916e-06,
+      "loss": 3.33,
+      "step": 2546688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.298337289360395e-06,
+      "loss": 3.3363,
+      "step": 2547200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.289967720653605e-06,
+      "loss": 3.3297,
+      "step": 2547712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.281581773143087e-06,
+      "loss": 3.3395,
+      "step": 2548224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.273195825632567e-06,
+      "loss": 3.3386,
+      "step": 2548736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.264809878122045e-06,
+      "loss": 3.3362,
+      "step": 2549248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.256440309415258e-06,
+      "loss": 3.3247,
+      "step": 2549760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.248054361904738e-06,
+      "loss": 3.3468,
+      "step": 2550272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.239668414394216e-06,
+      "loss": 3.3236,
+      "step": 2550784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.231282466883698e-06,
+      "loss": 3.3328,
+      "step": 2551296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.222896519373178e-06,
+      "loss": 3.3428,
+      "step": 2551808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.214526950666388e-06,
+      "loss": 3.342,
+      "step": 2552320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.206141003155868e-06,
+      "loss": 3.3279,
+      "step": 2552832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.197755055645348e-06,
+      "loss": 3.3278,
+      "step": 2553344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.189369108134828e-06,
+      "loss": 3.3294,
+      "step": 2553856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.18099953942804e-06,
+      "loss": 3.3376,
+      "step": 2554368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.172613591917519e-06,
+      "loss": 3.337,
+      "step": 2554880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.164227644406999e-06,
+      "loss": 3.3364,
+      "step": 2555392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.155841696896479e-06,
+      "loss": 3.3452,
+      "step": 2555904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.14747212818969e-06,
+      "loss": 3.3418,
+      "step": 2556416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.13908618067917e-06,
+      "loss": 3.3451,
+      "step": 2556928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.130700233168651e-06,
+      "loss": 3.3296,
+      "step": 2557440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.122314285658131e-06,
+      "loss": 3.3432,
+      "step": 2557952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.11394471695134e-06,
+      "loss": 3.3446,
+      "step": 2558464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.105558769440822e-06,
+      "loss": 3.3232,
+      "step": 2558976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.097172821930302e-06,
+      "loss": 3.3471,
+      "step": 2559488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.08878687441978e-06,
+      "loss": 3.3323,
+      "step": 2560000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.080417305712993e-06,
+      "loss": 3.3559,
+      "step": 2560512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.072031358202473e-06,
+      "loss": 3.3371,
+      "step": 2561024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.063645410691953e-06,
+      "loss": 3.3191,
+      "step": 2561536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.055259463181433e-06,
+      "loss": 3.3335,
+      "step": 2562048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.046889894474643e-06,
+      "loss": 3.3306,
+      "step": 2562560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.038503946964123e-06,
+      "loss": 3.3431,
+      "step": 2563072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.030117999453605e-06,
+      "loss": 3.3369,
+      "step": 2563584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 8.021732051943083e-06,
+      "loss": 3.3376,
+      "step": 2564096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.013362483236294e-06,
+      "loss": 3.3383,
+      "step": 2564608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 8.004976535725776e-06,
+      "loss": 3.3266,
+      "step": 2565120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.996590588215254e-06,
+      "loss": 3.3262,
+      "step": 2565632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.988204640704734e-06,
+      "loss": 3.3249,
+      "step": 2566144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.979835071997946e-06,
+      "loss": 3.3348,
+      "step": 2566656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.971449124487426e-06,
+      "loss": 3.3445,
+      "step": 2567168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.963063176976905e-06,
+      "loss": 3.3359,
+      "step": 2567680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.954677229466386e-06,
+      "loss": 3.3261,
+      "step": 2568192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.946307660759597e-06,
+      "loss": 3.3361,
+      "step": 2568704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.937921713249077e-06,
+      "loss": 3.3336,
+      "step": 2569216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.929535765738557e-06,
+      "loss": 3.3422,
+      "step": 2569728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.921149818228037e-06,
+      "loss": 3.3368,
+      "step": 2570240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.912780249521248e-06,
+      "loss": 3.3342,
+      "step": 2570752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.904394302010728e-06,
+      "loss": 3.3348,
+      "step": 2571264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.896008354500208e-06,
+      "loss": 3.3419,
+      "step": 2571776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.887622406989688e-06,
+      "loss": 3.3212,
+      "step": 2572288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.8792528382829e-06,
+      "loss": 3.3363,
+      "step": 2572800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.870866890772378e-06,
+      "loss": 3.3389,
+      "step": 2573312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.862480943261858e-06,
+      "loss": 3.3324,
+      "step": 2573824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.85409499575134e-06,
+      "loss": 3.326,
+      "step": 2574336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.84572542704455e-06,
+      "loss": 3.3353,
+      "step": 2574848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.837339479534029e-06,
+      "loss": 3.331,
+      "step": 2575360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.82895353202351e-06,
+      "loss": 3.3324,
+      "step": 2575872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.82056758451299e-06,
+      "loss": 3.3355,
+      "step": 2576384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.812181637002469e-06,
+      "loss": 3.3296,
+      "step": 2576896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.803812068295681e-06,
+      "loss": 3.3452,
+      "step": 2577408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.795426120785161e-06,
+      "loss": 3.3457,
+      "step": 2577920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.787040173274641e-06,
+      "loss": 3.3354,
+      "step": 2578432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.778654225764121e-06,
+      "loss": 3.3424,
+      "step": 2578944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.770284657057332e-06,
+      "loss": 3.3408,
+      "step": 2579456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.761898709546812e-06,
+      "loss": 3.3304,
+      "step": 2579968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.753512762036292e-06,
+      "loss": 3.327,
+      "step": 2580480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.745126814525772e-06,
+      "loss": 3.329,
+      "step": 2580992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.736757245818983e-06,
+      "loss": 3.3348,
+      "step": 2581504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.728371298308464e-06,
+      "loss": 3.3379,
+      "step": 2582016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.719985350797943e-06,
+      "loss": 3.3369,
+      "step": 2582528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.711599403287422e-06,
+      "loss": 3.3351,
+      "step": 2583040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.703229834580635e-06,
+      "loss": 3.3314,
+      "step": 2583552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.694843887070115e-06,
+      "loss": 3.3365,
+      "step": 2584064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.686457939559593e-06,
+      "loss": 3.3347,
+      "step": 2584576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.678071992049075e-06,
+      "loss": 3.342,
+      "step": 2585088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.669702423342286e-06,
+      "loss": 3.334,
+      "step": 2585600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.661316475831764e-06,
+      "loss": 3.3217,
+      "step": 2586112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.652930528321246e-06,
+      "loss": 3.3257,
+      "step": 2586624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.644544580810725e-06,
+      "loss": 3.3429,
+      "step": 2587136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.636175012103936e-06,
+      "loss": 3.3383,
+      "step": 2587648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.627789064593415e-06,
+      "loss": 3.3315,
+      "step": 2588160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.619403117082896e-06,
+      "loss": 3.3267,
+      "step": 2588672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.611017169572376e-06,
+      "loss": 3.3461,
+      "step": 2589184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.602647600865587e-06,
+      "loss": 3.3432,
+      "step": 2589696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.594261653355067e-06,
+      "loss": 3.3391,
+      "step": 2590208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.585875705844548e-06,
+      "loss": 3.3355,
+      "step": 2590720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.577489758334028e-06,
+      "loss": 3.3335,
+      "step": 2591232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.569120189627238e-06,
+      "loss": 3.3366,
+      "step": 2591744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.560734242116718e-06,
+      "loss": 3.3279,
+      "step": 2592256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.552348294606198e-06,
+      "loss": 3.3455,
+      "step": 2592768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.5439623470956775e-06,
+      "loss": 3.3331,
+      "step": 2593280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.535592778388889e-06,
+      "loss": 3.3463,
+      "step": 2593792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.527206830878369e-06,
+      "loss": 3.3347,
+      "step": 2594304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.51882088336785e-06,
+      "loss": 3.3265,
+      "step": 2594816
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.888495922088623,
+      "eval_runtime": 302.4779,
+      "eval_samples_per_second": 1261.55,
+      "eval_steps_per_second": 39.424,
+      "step": 2594880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.51045131466106e-06,
+      "loss": 3.3391,
+      "step": 2595328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.502065367150541e-06,
+      "loss": 3.3345,
+      "step": 2595840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.4936794196400206e-06,
+      "loss": 3.3364,
+      "step": 2596352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.485293472129501e-06,
+      "loss": 3.337,
+      "step": 2596864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.476923903422711e-06,
+      "loss": 3.3471,
+      "step": 2597376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.468537955912191e-06,
+      "loss": 3.3319,
+      "step": 2597888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.460152008401672e-06,
+      "loss": 3.3255,
+      "step": 2598400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.451766060891151e-06,
+      "loss": 3.3185,
+      "step": 2598912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.443396492184363e-06,
+      "loss": 3.3515,
+      "step": 2599424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.435010544673843e-06,
+      "loss": 3.3362,
+      "step": 2599936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.426624597163323e-06,
+      "loss": 3.3251,
+      "step": 2600448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.418238649652802e-06,
+      "loss": 3.3447,
+      "step": 2600960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.4098690809460135e-06,
+      "loss": 3.331,
+      "step": 2601472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.401483133435494e-06,
+      "loss": 3.3297,
+      "step": 2601984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.393097185924974e-06,
+      "loss": 3.3247,
+      "step": 2602496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.384711238414453e-06,
+      "loss": 3.3277,
+      "step": 2603008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.376325290903933e-06,
+      "loss": 3.3307,
+      "step": 2603520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.367955722197145e-06,
+      "loss": 3.3239,
+      "step": 2604032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.359569774686624e-06,
+      "loss": 3.3438,
+      "step": 2604544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.351183827176104e-06,
+      "loss": 3.3471,
+      "step": 2605056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.342814258469316e-06,
+      "loss": 3.336,
+      "step": 2605568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.3344283109587964e-06,
+      "loss": 3.329,
+      "step": 2606080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.3260423634482755e-06,
+      "loss": 3.3331,
+      "step": 2606592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.3176564159377555e-06,
+      "loss": 3.3357,
+      "step": 2607104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.309270468427236e-06,
+      "loss": 3.3296,
+      "step": 2607616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.300900899720448e-06,
+      "loss": 3.3296,
+      "step": 2608128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.292514952209926e-06,
+      "loss": 3.3252,
+      "step": 2608640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.284129004699407e-06,
+      "loss": 3.3417,
+      "step": 2609152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.275743057188887e-06,
+      "loss": 3.3141,
+      "step": 2609664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.267373488482098e-06,
+      "loss": 3.3246,
+      "step": 2610176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.258987540971578e-06,
+      "loss": 3.3343,
+      "step": 2610688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.250601593461058e-06,
+      "loss": 3.3352,
+      "step": 2611200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.2422156459505385e-06,
+      "loss": 3.336,
+      "step": 2611712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.2338296984400176e-06,
+      "loss": 3.3369,
+      "step": 2612224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.225460129733229e-06,
+      "loss": 3.3269,
+      "step": 2612736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.217074182222709e-06,
+      "loss": 3.3398,
+      "step": 2613248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.208688234712188e-06,
+      "loss": 3.3246,
+      "step": 2613760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.200302287201668e-06,
+      "loss": 3.33,
+      "step": 2614272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.19193271849488e-06,
+      "loss": 3.3172,
+      "step": 2614784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.183546770984361e-06,
+      "loss": 3.3242,
+      "step": 2615296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.175160823473839e-06,
+      "loss": 3.325,
+      "step": 2615808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.16677487596332e-06,
+      "loss": 3.341,
+      "step": 2616320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.158405307256531e-06,
+      "loss": 3.3314,
+      "step": 2616832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.150019359746011e-06,
+      "loss": 3.3276,
+      "step": 2617344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.1416334122354904e-06,
+      "loss": 3.3362,
+      "step": 2617856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.133247464724971e-06,
+      "loss": 3.3304,
+      "step": 2618368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.124877896018183e-06,
+      "loss": 3.3311,
+      "step": 2618880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.116491948507661e-06,
+      "loss": 3.3212,
+      "step": 2619392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.108106000997142e-06,
+      "loss": 3.3154,
+      "step": 2619904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.099720053486622e-06,
+      "loss": 3.3377,
+      "step": 2620416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.0913504847798335e-06,
+      "loss": 3.334,
+      "step": 2620928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.082964537269313e-06,
+      "loss": 3.3293,
+      "step": 2621440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.074578589758793e-06,
+      "loss": 3.3275,
+      "step": 2621952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.066192642248273e-06,
+      "loss": 3.3196,
+      "step": 2622464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.057823073541485e-06,
+      "loss": 3.3213,
+      "step": 2622976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.049437126030964e-06,
+      "loss": 3.3278,
+      "step": 2623488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.041051178520444e-06,
+      "loss": 3.3221,
+      "step": 2624000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.032665231009924e-06,
+      "loss": 3.3299,
+      "step": 2624512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.024295662303135e-06,
+      "loss": 3.3269,
+      "step": 2625024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.015909714792615e-06,
+      "loss": 3.3237,
+      "step": 2625536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.007523767282096e-06,
+      "loss": 3.3181,
+      "step": 2626048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.9991378197715756e-06,
+      "loss": 3.3355,
+      "step": 2626560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.990768251064786e-06,
+      "loss": 3.3163,
+      "step": 2627072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.982382303554266e-06,
+      "loss": 3.3236,
+      "step": 2627584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.973996356043746e-06,
+      "loss": 3.3295,
+      "step": 2628096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.965610408533225e-06,
+      "loss": 3.3334,
+      "step": 2628608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.957224461022706e-06,
+      "loss": 3.3186,
+      "step": 2629120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.948854892315918e-06,
+      "loss": 3.3237,
+      "step": 2629632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.940468944805398e-06,
+      "loss": 3.315,
+      "step": 2630144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.932082997294877e-06,
+      "loss": 3.3301,
+      "step": 2630656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.9237134285880885e-06,
+      "loss": 3.3293,
+      "step": 2631168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.9153274810775685e-06,
+      "loss": 3.3277,
+      "step": 2631680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.9069415335670476e-06,
+      "loss": 3.3321,
+      "step": 2632192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.8985555860565275e-06,
+      "loss": 3.3292,
+      "step": 2632704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.890169638546008e-06,
+      "loss": 3.3393,
+      "step": 2633216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.88180006983922e-06,
+      "loss": 3.321,
+      "step": 2633728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.873414122328699e-06,
+      "loss": 3.3305,
+      "step": 2634240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.865028174818179e-06,
+      "loss": 3.3384,
+      "step": 2634752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.85664222730766e-06,
+      "loss": 3.3175,
+      "step": 2635264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.8482726586008715e-06,
+      "loss": 3.3331,
+      "step": 2635776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.83988671109035e-06,
+      "loss": 3.3215,
+      "step": 2636288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.8315007635798305e-06,
+      "loss": 3.3464,
+      "step": 2636800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.8231148160693105e-06,
+      "loss": 3.3285,
+      "step": 2637312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.814745247362521e-06,
+      "loss": 3.314,
+      "step": 2637824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.806359299852001e-06,
+      "loss": 3.3209,
+      "step": 2638336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.797973352341481e-06,
+      "loss": 3.3249,
+      "step": 2638848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.789587404830962e-06,
+      "loss": 3.3314,
+      "step": 2639360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.781217836124172e-06,
+      "loss": 3.331,
+      "step": 2639872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.772831888613653e-06,
+      "loss": 3.3269,
+      "step": 2640384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.764445941103133e-06,
+      "loss": 3.3269,
+      "step": 2640896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.756059993592612e-06,
+      "loss": 3.3197,
+      "step": 2641408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.7476904248858234e-06,
+      "loss": 3.3193,
+      "step": 2641920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.739304477375303e-06,
+      "loss": 3.311,
+      "step": 2642432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.730918529864784e-06,
+      "loss": 3.3274,
+      "step": 2642944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.7225325823542625e-06,
+      "loss": 3.3331,
+      "step": 2643456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.714163013647474e-06,
+      "loss": 3.3291,
+      "step": 2643968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.705777066136955e-06,
+      "loss": 3.3184,
+      "step": 2644480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.697391118626435e-06,
+      "loss": 3.3284,
+      "step": 2644992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.689005171115914e-06,
+      "loss": 3.318,
+      "step": 2645504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.680635602409126e-06,
+      "loss": 3.3366,
+      "step": 2646016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.672249654898606e-06,
+      "loss": 3.3259,
+      "step": 2646528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.663863707388085e-06,
+      "loss": 3.3285,
+      "step": 2647040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.6554777598775655e-06,
+      "loss": 3.3235,
+      "step": 2647552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.647108191170777e-06,
+      "loss": 3.3334,
+      "step": 2648064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.638722243660257e-06,
+      "loss": 3.3131,
+      "step": 2648576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.630336296149736e-06,
+      "loss": 3.3304,
+      "step": 2649088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.621950348639216e-06,
+      "loss": 3.3235,
+      "step": 2649600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.613580779932428e-06,
+      "loss": 3.3284,
+      "step": 2650112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.6051948324219086e-06,
+      "loss": 3.3143,
+      "step": 2650624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.596808884911388e-06,
+      "loss": 3.3246,
+      "step": 2651136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.588422937400868e-06,
+      "loss": 3.3219,
+      "step": 2651648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.580053368694079e-06,
+      "loss": 3.3231,
+      "step": 2652160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.571667421183558e-06,
+      "loss": 3.3278,
+      "step": 2652672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.563281473673038e-06,
+      "loss": 3.3223,
+      "step": 2653184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.554895526162519e-06,
+      "loss": 3.33,
+      "step": 2653696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.546509578651999e-06,
+      "loss": 3.3368,
+      "step": 2654208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.53814000994521e-06,
+      "loss": 3.3308,
+      "step": 2654720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.52975406243469e-06,
+      "loss": 3.332,
+      "step": 2655232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.52136811492417e-06,
+      "loss": 3.3288,
+      "step": 2655744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.512982167413649e-06,
+      "loss": 3.3198,
+      "step": 2656256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.5046125987068605e-06,
+      "loss": 3.3206,
+      "step": 2656768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.496226651196341e-06,
+      "loss": 3.3231,
+      "step": 2657280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.487840703685821e-06,
+      "loss": 3.3229,
+      "step": 2657792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.4794547561753e-06,
+      "loss": 3.3282,
+      "step": 2658304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.471085187468512e-06,
+      "loss": 3.3255,
+      "step": 2658816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.462699239957992e-06,
+      "loss": 3.3294,
+      "step": 2659328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.454313292447473e-06,
+      "loss": 3.3236,
+      "step": 2659840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.445927344936951e-06,
+      "loss": 3.3238,
+      "step": 2660352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.437557776230163e-06,
+      "loss": 3.3233,
+      "step": 2660864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.4291718287196435e-06,
+      "loss": 3.3352,
+      "step": 2661376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.420785881209123e-06,
+      "loss": 3.3233,
+      "step": 2661888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.4123999336986026e-06,
+      "loss": 3.3176,
+      "step": 2662400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.404030364991814e-06,
+      "loss": 3.312,
+      "step": 2662912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.395644417481295e-06,
+      "loss": 3.3309,
+      "step": 2663424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.387258469970773e-06,
+      "loss": 3.3301,
+      "step": 2663936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.378872522460254e-06,
+      "loss": 3.3261,
+      "step": 2664448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.370502953753466e-06,
+      "loss": 3.3193,
+      "step": 2664960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.362117006242946e-06,
+      "loss": 3.3371,
+      "step": 2665472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.353731058732425e-06,
+      "loss": 3.329,
+      "step": 2665984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.345345111221905e-06,
+      "loss": 3.3312,
+      "step": 2666496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.336975542515116e-06,
+      "loss": 3.3251,
+      "step": 2667008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.3285895950045955e-06,
+      "loss": 3.327,
+      "step": 2667520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.320203647494076e-06,
+      "loss": 3.3288,
+      "step": 2668032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.311817699983556e-06,
+      "loss": 3.316,
+      "step": 2668544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.303448131276768e-06,
+      "loss": 3.3355,
+      "step": 2669056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.295062183766247e-06,
+      "loss": 3.3282,
+      "step": 2669568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.286676236255727e-06,
+      "loss": 3.3314,
+      "step": 2670080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.278290288745208e-06,
+      "loss": 3.3255,
+      "step": 2670592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.269920720038418e-06,
+      "loss": 3.3158,
+      "step": 2671104
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.888662338256836,
+      "eval_runtime": 303.3447,
+      "eval_samples_per_second": 1257.945,
+      "eval_steps_per_second": 39.312,
+      "step": 2671200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.261534772527898e-06,
+      "loss": 3.3299,
+      "step": 2671616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.2531488250173784e-06,
+      "loss": 3.3263,
+      "step": 2672128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.244762877506858e-06,
+      "loss": 3.326,
+      "step": 2672640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.23639330880007e-06,
+      "loss": 3.3282,
+      "step": 2673152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.228007361289549e-06,
+      "loss": 3.3382,
+      "step": 2673664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.21962141377903e-06,
+      "loss": 3.3238,
+      "step": 2674176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.211235466268509e-06,
+      "loss": 3.3139,
+      "step": 2674688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.202865897561721e-06,
+      "loss": 3.3116,
+      "step": 2675200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.194479950051201e-06,
+      "loss": 3.3381,
+      "step": 2675712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.18609400254068e-06,
+      "loss": 3.3311,
+      "step": 2676224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.1777080550301606e-06,
+      "loss": 3.3115,
+      "step": 2676736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.169338486323371e-06,
+      "loss": 3.3354,
+      "step": 2677248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.160952538812851e-06,
+      "loss": 3.3226,
+      "step": 2677760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.152566591302331e-06,
+      "loss": 3.3202,
+      "step": 2678272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.144180643791811e-06,
+      "loss": 3.3191,
+      "step": 2678784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.135811075085023e-06,
+      "loss": 3.3188,
+      "step": 2679296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.127425127574503e-06,
+      "loss": 3.3182,
+      "step": 2679808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.119039180063983e-06,
+      "loss": 3.3152,
+      "step": 2680320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.110653232553462e-06,
+      "loss": 3.3344,
+      "step": 2680832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.102267285042943e-06,
+      "loss": 3.3376,
+      "step": 2681344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.0938977163361535e-06,
+      "loss": 3.3291,
+      "step": 2681856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.0855117688256334e-06,
+      "loss": 3.3222,
+      "step": 2682368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.077125821315113e-06,
+      "loss": 3.3212,
+      "step": 2682880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.068739873804593e-06,
+      "loss": 3.3302,
+      "step": 2683392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.060370305097805e-06,
+      "loss": 3.318,
+      "step": 2683904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.051984357587285e-06,
+      "loss": 3.3228,
+      "step": 2684416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.043598410076765e-06,
+      "loss": 3.3173,
+      "step": 2684928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.035212462566244e-06,
+      "loss": 3.329,
+      "step": 2685440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.026842893859456e-06,
+      "loss": 3.3085,
+      "step": 2685952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.018456946348936e-06,
+      "loss": 3.3171,
+      "step": 2686464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.0100709988384155e-06,
+      "loss": 3.3288,
+      "step": 2686976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.0016850513278955e-06,
+      "loss": 3.3217,
+      "step": 2687488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.993315482621107e-06,
+      "loss": 3.3282,
+      "step": 2688000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.984929535110586e-06,
+      "loss": 3.3296,
+      "step": 2688512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.976543587600066e-06,
+      "loss": 3.3178,
+      "step": 2689024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.968157640089546e-06,
+      "loss": 3.3306,
+      "step": 2689536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.959788071382758e-06,
+      "loss": 3.3175,
+      "step": 2690048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.951402123872238e-06,
+      "loss": 3.3141,
+      "step": 2690560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.943016176361718e-06,
+      "loss": 3.3126,
+      "step": 2691072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.934630228851198e-06,
+      "loss": 3.3142,
+      "step": 2691584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.9262606601444085e-06,
+      "loss": 3.3157,
+      "step": 2692096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.917874712633889e-06,
+      "loss": 3.3336,
+      "step": 2692608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.909488765123368e-06,
+      "loss": 3.3209,
+      "step": 2693120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.901102817612848e-06,
+      "loss": 3.319,
+      "step": 2693632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.89273324890606e-06,
+      "loss": 3.3286,
+      "step": 2694144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.88434730139554e-06,
+      "loss": 3.3178,
+      "step": 2694656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.87596135388502e-06,
+      "loss": 3.3277,
+      "step": 2695168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.8675754063745e-06,
+      "loss": 3.3145,
+      "step": 2695680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.8592058376677115e-06,
+      "loss": 3.302,
+      "step": 2696192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.8508198901571906e-06,
+      "loss": 3.3339,
+      "step": 2696704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.842433942646671e-06,
+      "loss": 3.3232,
+      "step": 2697216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.8340479951361505e-06,
+      "loss": 3.3247,
+      "step": 2697728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.825678426429362e-06,
+      "loss": 3.3144,
+      "step": 2698240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.817292478918842e-06,
+      "loss": 3.3111,
+      "step": 2698752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.808906531408321e-06,
+      "loss": 3.3126,
+      "step": 2699264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.800520583897802e-06,
+      "loss": 3.3195,
+      "step": 2699776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.792134636387281e-06,
+      "loss": 3.3136,
+      "step": 2700288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.783765067680494e-06,
+      "loss": 3.3164,
+      "step": 2700800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.775379120169973e-06,
+      "loss": 3.3256,
+      "step": 2701312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.7669931726594535e-06,
+      "loss": 3.3136,
+      "step": 2701824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.758607225148933e-06,
+      "loss": 3.3054,
+      "step": 2702336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.750237656442144e-06,
+      "loss": 3.3296,
+      "step": 2702848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.741851708931624e-06,
+      "loss": 3.3076,
+      "step": 2703360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.733465761421103e-06,
+      "loss": 3.3119,
+      "step": 2703872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.725079813910584e-06,
+      "loss": 3.3248,
+      "step": 2704384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.716710245203795e-06,
+      "loss": 3.3265,
+      "step": 2704896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.708324297693275e-06,
+      "loss": 3.3091,
+      "step": 2705408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.699938350182755e-06,
+      "loss": 3.3144,
+      "step": 2705920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.691552402672235e-06,
+      "loss": 3.3021,
+      "step": 2706432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.683182833965446e-06,
+      "loss": 3.3255,
+      "step": 2706944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.674796886454926e-06,
+      "loss": 3.318,
+      "step": 2707456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.666410938944406e-06,
+      "loss": 3.3198,
+      "step": 2707968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.6580249914338854e-06,
+      "loss": 3.3232,
+      "step": 2708480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.649639043923366e-06,
+      "loss": 3.3218,
+      "step": 2708992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.641269475216577e-06,
+      "loss": 3.3342,
+      "step": 2709504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.632883527706057e-06,
+      "loss": 3.3106,
+      "step": 2710016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.624497580195537e-06,
+      "loss": 3.3209,
+      "step": 2710528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.616111632685017e-06,
+      "loss": 3.3291,
+      "step": 2711040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.6077420639782285e-06,
+      "loss": 3.3085,
+      "step": 2711552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5993561164677085e-06,
+      "loss": 3.3244,
+      "step": 2712064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5909701689571884e-06,
+      "loss": 3.3175,
+      "step": 2712576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5825842214466675e-06,
+      "loss": 3.3306,
+      "step": 2713088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.574198273936148e-06,
+      "loss": 3.3208,
+      "step": 2713600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.565828705229359e-06,
+      "loss": 3.3072,
+      "step": 2714112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.557442757718839e-06,
+      "loss": 3.315,
+      "step": 2714624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.549056810208319e-06,
+      "loss": 3.3131,
+      "step": 2715136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.540670862697799e-06,
+      "loss": 3.3268,
+      "step": 2715648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.53230129399101e-06,
+      "loss": 3.319,
+      "step": 2716160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.523915346480491e-06,
+      "loss": 3.3152,
+      "step": 2716672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.51552939896997e-06,
+      "loss": 3.3193,
+      "step": 2717184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.50714345145945e-06,
+      "loss": 3.3175,
+      "step": 2717696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.498773882752661e-06,
+      "loss": 3.3107,
+      "step": 2718208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.490387935242141e-06,
+      "loss": 3.3031,
+      "step": 2718720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.482001987731621e-06,
+      "loss": 3.318,
+      "step": 2719232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.473616040221101e-06,
+      "loss": 3.3266,
+      "step": 2719744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.465246471514313e-06,
+      "loss": 3.3225,
+      "step": 2720256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.456860524003792e-06,
+      "loss": 3.3063,
+      "step": 2720768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.448474576493273e-06,
+      "loss": 3.3197,
+      "step": 2721280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.440088628982752e-06,
+      "loss": 3.3101,
+      "step": 2721792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.4317190602759635e-06,
+      "loss": 3.3258,
+      "step": 2722304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.423333112765443e-06,
+      "loss": 3.3168,
+      "step": 2722816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.414947165254923e-06,
+      "loss": 3.3192,
+      "step": 2723328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.406561217744403e-06,
+      "loss": 3.3155,
+      "step": 2723840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.398191649037614e-06,
+      "loss": 3.3242,
+      "step": 2724352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.389805701527095e-06,
+      "loss": 3.3032,
+      "step": 2724864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.381419754016574e-06,
+      "loss": 3.3208,
+      "step": 2725376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.373033806506055e-06,
+      "loss": 3.3115,
+      "step": 2725888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.364664237799266e-06,
+      "loss": 3.3183,
+      "step": 2726400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.356278290288746e-06,
+      "loss": 3.3051,
+      "step": 2726912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.3478923427782255e-06,
+      "loss": 3.3225,
+      "step": 2727424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.339506395267705e-06,
+      "loss": 3.309,
+      "step": 2727936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.331136826560916e-06,
+      "loss": 3.314,
+      "step": 2728448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.322750879050396e-06,
+      "loss": 3.3157,
+      "step": 2728960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.314364931539877e-06,
+      "loss": 3.3151,
+      "step": 2729472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.305978984029356e-06,
+      "loss": 3.3213,
+      "step": 2729984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.297593036518836e-06,
+      "loss": 3.3288,
+      "step": 2730496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.289223467812048e-06,
+      "loss": 3.3201,
+      "step": 2731008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.280837520301527e-06,
+      "loss": 3.3259,
+      "step": 2731520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.272451572791008e-06,
+      "loss": 3.3198,
+      "step": 2732032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.264065625280487e-06,
+      "loss": 3.3122,
+      "step": 2732544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.255696056573698e-06,
+      "loss": 3.3084,
+      "step": 2733056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.247310109063178e-06,
+      "loss": 3.3129,
+      "step": 2733568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.238924161552658e-06,
+      "loss": 3.3171,
+      "step": 2734080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.230538214042138e-06,
+      "loss": 3.3207,
+      "step": 2734592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.22216864533535e-06,
+      "loss": 3.3158,
+      "step": 2735104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.21378269782483e-06,
+      "loss": 3.3218,
+      "step": 2735616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.205396750314309e-06,
+      "loss": 3.3142,
+      "step": 2736128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.19701080280379e-06,
+      "loss": 3.313,
+      "step": 2736640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.1886412340970006e-06,
+      "loss": 3.3148,
+      "step": 2737152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.1802552865864805e-06,
+      "loss": 3.3256,
+      "step": 2737664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.1718693390759605e-06,
+      "loss": 3.3159,
+      "step": 2738176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.1634833915654404e-06,
+      "loss": 3.3082,
+      "step": 2738688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.155113822858652e-06,
+      "loss": 3.3031,
+      "step": 2739200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.146727875348132e-06,
+      "loss": 3.3214,
+      "step": 2739712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.138341927837612e-06,
+      "loss": 3.3206,
+      "step": 2740224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.129955980327091e-06,
+      "loss": 3.3197,
+      "step": 2740736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.121586411620303e-06,
+      "loss": 3.3103,
+      "step": 2741248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.113200464109783e-06,
+      "loss": 3.3297,
+      "step": 2741760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.104814516599263e-06,
+      "loss": 3.3189,
+      "step": 2742272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.096428569088743e-06,
+      "loss": 3.3233,
+      "step": 2742784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.088059000381954e-06,
+      "loss": 3.3161,
+      "step": 2743296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.079673052871433e-06,
+      "loss": 3.3172,
+      "step": 2743808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.071287105360914e-06,
+      "loss": 3.3213,
+      "step": 2744320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.062901157850393e-06,
+      "loss": 3.3054,
+      "step": 2744832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.054531589143605e-06,
+      "loss": 3.3268,
+      "step": 2745344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.046145641633085e-06,
+      "loss": 3.322,
+      "step": 2745856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.037759694122565e-06,
+      "loss": 3.3208,
+      "step": 2746368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.029373746612045e-06,
+      "loss": 3.316,
+      "step": 2746880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.0210041779052555e-06,
+      "loss": 3.3131,
+      "step": 2747392
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.88900089263916,
+      "eval_runtime": 303.2284,
+      "eval_samples_per_second": 1258.428,
+      "eval_steps_per_second": 39.327,
+      "step": 2747520
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.8352569175136502e+18,
+  "trial_name": null,
+  "trial_params": null
+}