diff --git "a/checkpoint-1297440/trainer_state.json" "b/checkpoint-1297440/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1297440/trainer_state.json"
@@ -0,0 +1,15365 @@
+{
+  "best_metric": 3.840970277786255,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/npi-sim-ques/transformer/2/checkpoints/checkpoint-915840",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 1297440,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 11.0359,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8165,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.186,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9717,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8095,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7099,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.603,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5345,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4553,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.407,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3469,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3257,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.273,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2071,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1798,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1367,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1191,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0748,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.045,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0215,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0009,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 4.9793,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9573,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9386,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9166,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.895,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8676,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8534,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.831,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8227,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.8075,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7907,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7807,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7604,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7578,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7533,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7343,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 4.7003,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 4.7084,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 4.6815,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6757,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6741,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6657,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6488,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6506,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6409,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6324,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6245,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.6017,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.5961,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.6023,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.5866,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.572,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.559,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5644,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5467,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5631,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 4.5354,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5416,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5333,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5096,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5184,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5061,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4972,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4925,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5083,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.486,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4748,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4609,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.4629,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4775,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4766,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4508,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4684,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379669187473754e-05,
+      "loss": 4.4616,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 4.4479,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 4.4345,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 4.4393,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 4.4274,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 4.4218,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 4.4224,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 4.4184,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.4214,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 4.4173,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.3987,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.4037,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.4031,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 4.3945,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926231505873767e-05,
+      "loss": 4.3897,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925394549003088e-05,
+      "loss": 4.3871,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924555954252036e-05,
+      "loss": 4.3879,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923717359500984e-05,
+      "loss": 4.3767,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922878764749932e-05,
+      "loss": 4.377,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 4.3698,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.3742,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.3688,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.3636,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918687428875045e-05,
+      "loss": 4.3482,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917848834123993e-05,
+      "loss": 4.3523,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917010239372941e-05,
+      "loss": 4.3551,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916173282502262e-05,
+      "loss": 4.3619,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91533468775121e-05,
+      "loss": 4.347,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914497730880531e-05,
+      "loss": 4.3359,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913659136129479e-05,
+      "loss": 4.3323,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912820541378427e-05,
+      "loss": 4.3405,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911981946627375e-05,
+      "loss": 4.3108,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111449897566966e-05,
+      "loss": 4.3209,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103063950056446e-05,
+      "loss": 4.3224,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094678002545926e-05,
+      "loss": 4.3213,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086292055035406e-05,
+      "loss": 4.314,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9077906107524886e-05,
+      "loss": 4.3208,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9069520160014366e-05,
+      "loss": 4.301,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9061134212503846e-05,
+      "loss": 4.3273,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052764643797055e-05,
+      "loss": 4.3071,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9044378696286535e-05,
+      "loss": 4.3105,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9035992748776015e-05,
+      "loss": 4.312,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027606801265495e-05,
+      "loss": 4.3224,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019237232558704e-05,
+      "loss": 4.3001,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010851285048184e-05,
+      "loss": 4.2933,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900246533753767e-05,
+      "loss": 4.2986,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899407939002715e-05,
+      "loss": 4.2866,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985693442516624e-05,
+      "loss": 4.2941,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977307495006104e-05,
+      "loss": 4.2914,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968921547495584e-05,
+      "loss": 4.2795,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89605519787888e-05,
+      "loss": 4.2982,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895216603127827e-05,
+      "loss": 4.279,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894378008376775e-05,
+      "loss": 4.2733,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893539413625723e-05,
+      "loss": 4.2736,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892702456755045e-05,
+      "loss": 4.2781,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891863862003992e-05,
+      "loss": 4.2633,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891025267252941e-05,
+      "loss": 4.2804,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890186672501889e-05,
+      "loss": 4.2735,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889348077750837e-05,
+      "loss": 4.2599,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888509482999785e-05,
+      "loss": 4.2521,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887670888248733e-05,
+      "loss": 4.2633,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886832293497681e-05,
+      "loss": 4.2586,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885995336627002e-05,
+      "loss": 4.2639,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851583797563226e-05,
+      "loss": 4.2581,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8843197850052706e-05,
+      "loss": 4.2592,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834811902542186e-05,
+      "loss": 4.2561,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8826425955031666e-05,
+      "loss": 4.2511,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8818040007521146e-05,
+      "loss": 4.2427,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8809654060010626e-05,
+      "loss": 4.2408,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8801268112500106e-05,
+      "loss": 4.2472,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879289854379332e-05,
+      "loss": 4.2262,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87845125962828e-05,
+      "loss": 4.2471,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877612664877228e-05,
+      "loss": 4.2419,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876774070126176e-05,
+      "loss": 4.2341,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875935475375124e-05,
+      "loss": 4.2354,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875098518504445e-05,
+      "loss": 4.236,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.196593284606934,
+      "eval_runtime": 303.8622,
+      "eval_samples_per_second": 1255.803,
+      "eval_steps_per_second": 39.245,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874259923753393e-05,
+      "loss": 4.2105,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873421329002341e-05,
+      "loss": 4.2181,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872582734251289e-05,
+      "loss": 4.2297,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871744139500237e-05,
+      "loss": 4.2154,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870905544749185e-05,
+      "loss": 4.2198,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870066949998133e-05,
+      "loss": 4.2098,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869228355247081e-05,
+      "loss": 4.2076,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683913983764027e-05,
+      "loss": 4.2033,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675528036253507e-05,
+      "loss": 4.2113,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667142088742986e-05,
+      "loss": 4.1989,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658772520036196e-05,
+      "loss": 4.2193,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650386572525676e-05,
+      "loss": 4.2135,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642017003818885e-05,
+      "loss": 4.1948,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633631056308365e-05,
+      "loss": 4.1907,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8625245108797845e-05,
+      "loss": 4.1882,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8616859161287324e-05,
+      "loss": 4.1889,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608473213776804e-05,
+      "loss": 4.1836,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860008726626628e-05,
+      "loss": 4.1849,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591701318755764e-05,
+      "loss": 4.1817,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583315371245244e-05,
+      "loss": 4.1943,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574929423734724e-05,
+      "loss": 4.1845,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566543476224204e-05,
+      "loss": 4.1868,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558157528713684e-05,
+      "loss": 4.1822,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854978796000689e-05,
+      "loss": 4.1915,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854140201249637e-05,
+      "loss": 4.1678,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853301606498585e-05,
+      "loss": 4.1723,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852463011747533e-05,
+      "loss": 4.1654,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851624416996481e-05,
+      "loss": 4.1651,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.1644,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.1618,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849110270623698e-05,
+      "loss": 4.1655,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848271675872646e-05,
+      "loss": 4.1649,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847433081121595e-05,
+      "loss": 4.1546,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846594486370543e-05,
+      "loss": 4.1626,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845755891619491e-05,
+      "loss": 4.1692,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844917296868439e-05,
+      "loss": 4.1573,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84408033999776e-05,
+      "loss": 4.1417,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.1519,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842403150495656e-05,
+      "loss": 4.1342,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841564555744604e-05,
+      "loss": 4.1433,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840727598873925e-05,
+      "loss": 4.1439,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839889004122873e-05,
+      "loss": 4.1507,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839050409371821e-05,
+      "loss": 4.1381,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382118146207687e-05,
+      "loss": 4.1448,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373732198697167e-05,
+      "loss": 4.1465,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365346251186646e-05,
+      "loss": 4.1402,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835697668247986e-05,
+      "loss": 4.1409,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834859073496934e-05,
+      "loss": 4.1243,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834020478745882e-05,
+      "loss": 4.1308,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331818839948295e-05,
+      "loss": 4.1387,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832344927124151e-05,
+      "loss": 4.1301,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831506332373099e-05,
+      "loss": 4.1244,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830667737622047e-05,
+      "loss": 4.1106,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298291428709944e-05,
+      "loss": 4.1225,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289905481199424e-05,
+      "loss": 4.1126,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828153591249264e-05,
+      "loss": 4.1275,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827314996498212e-05,
+      "loss": 4.1139,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82647640174716e-05,
+      "loss": 4.1225,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825637806996108e-05,
+      "loss": 4.1241,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824799212245056e-05,
+      "loss": 4.1007,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823960617494004e-05,
+      "loss": 4.1135,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823123660623325e-05,
+      "loss": 4.1113,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822285065872273e-05,
+      "loss": 4.1029,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821446471121221e-05,
+      "loss": 4.103,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820607876370169e-05,
+      "loss": 4.1189,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81977091949949e-05,
+      "loss": 4.1068,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818932324748438e-05,
+      "loss": 4.0955,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818093729997386e-05,
+      "loss": 4.0883,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817255135246334e-05,
+      "loss": 4.0963,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816416540495282e-05,
+      "loss": 4.1091,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155795836246034e-05,
+      "loss": 4.1151,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147409888735514e-05,
+      "loss": 4.0981,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8139023941224994e-05,
+      "loss": 4.1119,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8130637993714474e-05,
+      "loss": 4.1068,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812226842500768e-05,
+      "loss": 4.1002,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811388247749716e-05,
+      "loss": 4.0866,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810549652998664e-05,
+      "loss": 4.1014,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809711058247612e-05,
+      "loss": 4.0911,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80887246349656e-05,
+      "loss": 4.0864,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808035506625881e-05,
+      "loss": 4.0903,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807196911874829e-05,
+      "loss": 4.086,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806358317123777e-05,
+      "loss": 4.1012,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805519722372726e-05,
+      "loss": 4.0875,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804682765502047e-05,
+      "loss": 4.0809,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803844170750995e-05,
+      "loss": 4.0845,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803005575999943e-05,
+      "loss": 4.0873,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802166981248891e-05,
+      "loss": 4.0829,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8013300243782116e-05,
+      "loss": 4.0827,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8004914296271596e-05,
+      "loss": 4.0754,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996528348761076e-05,
+      "loss": 4.0884,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988142401250556e-05,
+      "loss": 4.0739,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979756453740036e-05,
+      "loss": 4.0752,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971386885033245e-05,
+      "loss": 4.0696,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7963000937522725e-05,
+      "loss": 4.0816,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795461499001221e-05,
+      "loss": 4.0798,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794622904250169e-05,
+      "loss": 4.0726,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79378594737949e-05,
+      "loss": 4.0614,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792947352628438e-05,
+      "loss": 4.0705,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792108757877386e-05,
+      "loss": 4.0709,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791270163126334e-05,
+      "loss": 4.08,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790433206255655e-05,
+      "loss": 4.071,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789594611504603e-05,
+      "loss": 4.0596,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788756016753551e-05,
+      "loss": 4.0555,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787917422002499e-05,
+      "loss": 4.07,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78708046513182e-05,
+      "loss": 4.0389,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786241870380768e-05,
+      "loss": 4.0563,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854032756297166e-05,
+      "loss": 4.0582,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845646808786646e-05,
+      "loss": 4.0573,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837277240079855e-05,
+      "loss": 4.0474,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828891292569335e-05,
+      "loss": 4.0569,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820505345058815e-05,
+      "loss": 4.0486,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812119397548295e-05,
+      "loss": 4.0676,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803749828841504e-05,
+      "loss": 4.0525,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795363881330984e-05,
+      "loss": 4.0522,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786977933820464e-05,
+      "loss": 4.0669,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778591986309944e-05,
+      "loss": 4.0723,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777022241760315e-05,
+      "loss": 4.0501,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776183647009263e-05,
+      "loss": 4.0466,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775345052258212e-05,
+      "loss": 4.0565,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77450645750716e-05,
+      "loss": 4.0435,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773667862756108e-05,
+      "loss": 4.048,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772829268005055e-05,
+      "loss": 4.0537,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771990673254003e-05,
+      "loss": 4.04,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771152078502951e-05,
+      "loss": 4.0594,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770315121632272e-05,
+      "loss": 4.0414,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76947652688122e-05,
+      "loss": 4.0441,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768637932130168e-05,
+      "loss": 4.0373,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767799337379116e-05,
+      "loss": 4.0441,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766962380508437e-05,
+      "loss": 4.0351,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766123785757386e-05,
+      "loss": 4.0494,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765285191006334e-05,
+      "loss": 4.0468,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764446596255282e-05,
+      "loss": 4.0372,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7636096393846026e-05,
+      "loss": 4.0204,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7627710446335506e-05,
+      "loss": 4.0444,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7619324498824986e-05,
+      "loss": 4.0351,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610938551314466e-05,
+      "loss": 4.043,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7602568982607675e-05,
+      "loss": 4.0389,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7594183035097155e-05,
+      "loss": 4.0388,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7585797087586635e-05,
+      "loss": 4.0361,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577411140076115e-05,
+      "loss": 4.0412,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569041571369324e-05,
+      "loss": 4.0249,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756065562385881e-05,
+      "loss": 4.0255,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755226967634829e-05,
+      "loss": 4.0353,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754388372883777e-05,
+      "loss": 4.0188,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753549778132725e-05,
+      "loss": 4.0347,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752712821262046e-05,
+      "loss": 4.038,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751874226510994e-05,
+      "loss": 4.0235,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751035631759942e-05,
+      "loss": 4.0254,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75019703700889e-05,
+      "loss": 4.0278,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.027783393859863,
+      "eval_runtime": 307.6267,
+      "eval_samples_per_second": 1240.435,
+      "eval_steps_per_second": 38.765,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749360080138211e-05,
+      "loss": 4.0083,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748521485387159e-05,
+      "loss": 4.0149,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747682890636107e-05,
+      "loss": 4.0269,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746844295885055e-05,
+      "loss": 4.0182,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746005701134003e-05,
+      "loss": 4.0257,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745167106382951e-05,
+      "loss": 4.0084,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744328511631899e-05,
+      "loss": 4.018,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434899168808475e-05,
+      "loss": 4.0034,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426513221297955e-05,
+      "loss": 4.0167,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418127273787435e-05,
+      "loss": 4.0105,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740974132627691e-05,
+      "loss": 4.0209,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401371757570124e-05,
+      "loss": 4.0232,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739300218886333e-05,
+      "loss": 4.004,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738461624135281e-05,
+      "loss": 4.0047,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737623029384229e-05,
+      "loss": 4.0038,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736784434633177e-05,
+      "loss": 4.0013,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735945839882125e-05,
+      "loss": 3.9968,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735107245131073e-05,
+      "loss": 4.002,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734268650380021e-05,
+      "loss": 3.9966,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733431693509343e-05,
+      "loss": 4.0205,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732593098758291e-05,
+      "loss": 3.9975,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731754504007238e-05,
+      "loss": 4.0121,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730915909256186e-05,
+      "loss": 4.0047,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730077314505134e-05,
+      "loss": 4.0125,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729240357634456e-05,
+      "loss": 3.9869,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728401762883403e-05,
+      "loss": 3.9983,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727563168132351e-05,
+      "loss": 3.9879,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726724573381299e-05,
+      "loss": 3.9961,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725887616510621e-05,
+      "loss": 3.9889,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725049021759568e-05,
+      "loss": 3.9893,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7242120648888896e-05,
+      "loss": 3.9987,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723373470137838e-05,
+      "loss": 3.995,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225348753867856e-05,
+      "loss": 3.9836,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216962806357336e-05,
+      "loss": 3.9972,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208576858846816e-05,
+      "loss": 3.9942,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7200190911336296e-05,
+      "loss": 3.9912,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191804963825775e-05,
+      "loss": 3.9792,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183419016315255e-05,
+      "loss": 3.9879,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175049447608465e-05,
+      "loss": 3.9694,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7166663500097945e-05,
+      "loss": 3.983,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158277552587424e-05,
+      "loss": 3.979,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149891605076904e-05,
+      "loss": 3.9889,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714152203637012e-05,
+      "loss": 3.9799,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71331360888596e-05,
+      "loss": 3.9833,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712475014134908e-05,
+      "loss": 3.9877,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711636419383856e-05,
+      "loss": 3.9834,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710799462513177e-05,
+      "loss": 3.9853,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709960867762125e-05,
+      "loss": 3.9664,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709122273011073e-05,
+      "loss": 3.9715,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708283678260021e-05,
+      "loss": 3.9838,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707446721389342e-05,
+      "loss": 3.9754,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70660812663829e-05,
+      "loss": 3.9695,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705769531887238e-05,
+      "loss": 3.9615,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704930937136186e-05,
+      "loss": 3.9709,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040939802655074e-05,
+      "loss": 3.9584,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032553855144554e-05,
+      "loss": 3.9758,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024167907634034e-05,
+      "loss": 3.9638,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015781960123514e-05,
+      "loss": 3.9768,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700741239141672e-05,
+      "loss": 3.9729,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69990264439062e-05,
+      "loss": 3.954,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699064049639568e-05,
+      "loss": 3.9636,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698225454888516e-05,
+      "loss": 3.9637,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697388498017837e-05,
+      "loss": 3.9554,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696549903266785e-05,
+      "loss": 3.9621,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695711308515733e-05,
+      "loss": 3.9682,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694872713764681e-05,
+      "loss": 3.9651,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694035756894003e-05,
+      "loss": 3.9543,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693197162142951e-05,
+      "loss": 3.9452,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692358567391899e-05,
+      "loss": 3.9518,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691519972640847e-05,
+      "loss": 3.965,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906830157701677e-05,
+      "loss": 3.9735,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898444210191157e-05,
+      "loss": 3.9562,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6890058262680636e-05,
+      "loss": 3.9698,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881672315170116e-05,
+      "loss": 3.9671,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873302746463326e-05,
+      "loss": 3.9644,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864916798952806e-05,
+      "loss": 3.9412,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856530851442285e-05,
+      "loss": 3.966,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848144903931765e-05,
+      "loss": 3.9566,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683977533522498e-05,
+      "loss": 3.9482,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683138938771446e-05,
+      "loss": 3.9575,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682300344020394e-05,
+      "loss": 3.9516,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681461749269342e-05,
+      "loss": 3.963,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680624792398663e-05,
+      "loss": 3.9546,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679786197647611e-05,
+      "loss": 3.9449,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678947602896559e-05,
+      "loss": 3.9506,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67811064602588e-05,
+      "loss": 3.9517,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677272051274828e-05,
+      "loss": 3.9526,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676433456523776e-05,
+      "loss": 3.9477,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675594861772724e-05,
+      "loss": 3.9453,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674757904902045e-05,
+      "loss": 3.9536,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739193101509935e-05,
+      "loss": 3.9423,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730807153999415e-05,
+      "loss": 3.9461,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722421206488895e-05,
+      "loss": 3.9402,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6714051637782104e-05,
+      "loss": 3.9551,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705665690271584e-05,
+      "loss": 3.9501,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697279742761064e-05,
+      "loss": 3.9437,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668891017405427e-05,
+      "loss": 3.9372,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668052422654375e-05,
+      "loss": 3.9472,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667213827903323e-05,
+      "loss": 3.9373,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666375233152271e-05,
+      "loss": 3.9566,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665538276281592e-05,
+      "loss": 3.943,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66469968153054e-05,
+      "loss": 3.9373,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663861086779488e-05,
+      "loss": 3.9306,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663022492028437e-05,
+      "loss": 3.9426,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662183897277385e-05,
+      "loss": 3.9207,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661345302526333e-05,
+      "loss": 3.9332,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660508345655654e-05,
+      "loss": 3.9365,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659669750904602e-05,
+      "loss": 3.9353,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65883115615355e-05,
+      "loss": 3.9223,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657992561402498e-05,
+      "loss": 3.9398,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657153966651446e-05,
+      "loss": 3.9228,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656315371900393e-05,
+      "loss": 3.9471,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554784150297146e-05,
+      "loss": 3.9325,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546398202786626e-05,
+      "loss": 3.9303,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6538012255276106e-05,
+      "loss": 3.947,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529626307765586e-05,
+      "loss": 3.954,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65212567390588e-05,
+      "loss": 3.9324,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651287079154828e-05,
+      "loss": 3.9255,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650448484403776e-05,
+      "loss": 3.9367,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6496098896527235e-05,
+      "loss": 3.9278,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487712949016715e-05,
+      "loss": 3.9283,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479327001506195e-05,
+      "loss": 3.9415,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470941053995675e-05,
+      "loss": 3.9189,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462571485288884e-05,
+      "loss": 3.9416,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454185537778364e-05,
+      "loss": 3.9289,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445799590267844e-05,
+      "loss": 3.9288,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437413642757324e-05,
+      "loss": 3.9205,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429027695246804e-05,
+      "loss": 3.9303,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642064174773629e-05,
+      "loss": 3.9183,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641225580022577e-05,
+      "loss": 3.939,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640386985271525e-05,
+      "loss": 3.929,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639550028400846e-05,
+      "loss": 3.9255,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638711433649794e-05,
+      "loss": 3.9086,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637874476779115e-05,
+      "loss": 3.9314,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637035882028063e-05,
+      "loss": 3.9257,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636197287277011e-05,
+      "loss": 3.9276,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635358692525959e-05,
+      "loss": 3.9233,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634520097774907e-05,
+      "loss": 3.9315,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633681503023855e-05,
+      "loss": 3.9261,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632842908272803e-05,
+      "loss": 3.9328,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632004313521751e-05,
+      "loss": 3.9137,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311673566510724e-05,
+      "loss": 3.9152,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303287619000204e-05,
+      "loss": 3.9235,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294901671489684e-05,
+      "loss": 3.9082,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286532102782893e-05,
+      "loss": 3.9244,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6278146155272373e-05,
+      "loss": 3.9311,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626976020776185e-05,
+      "loss": 3.9165,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626137426025133e-05,
+      "loss": 3.9142,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625300469154454e-05,
+      "loss": 3.9233,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9528167247772217,
+      "eval_runtime": 303.7,
+      "eval_samples_per_second": 1256.473,
+      "eval_steps_per_second": 39.266,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624461874403402e-05,
+      "loss": 3.9097,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.62362327965235e-05,
+      "loss": 3.9079,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622784684901298e-05,
+      "loss": 3.9198,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621946090150246e-05,
+      "loss": 3.9091,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621107495399194e-05,
+      "loss": 3.9236,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620268900648142e-05,
+      "loss": 3.903,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61943030589709e-05,
+      "loss": 3.9154,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618593349026412e-05,
+      "loss": 3.8984,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617756392155733e-05,
+      "loss": 3.9122,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616917797404681e-05,
+      "loss": 3.9054,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616079202653629e-05,
+      "loss": 3.917,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615240607902577e-05,
+      "loss": 3.9182,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6144036510318976e-05,
+      "loss": 3.9022,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135650562808456e-05,
+      "loss": 3.9041,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127264615297936e-05,
+      "loss": 3.9019,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6118878667787416e-05,
+      "loss": 3.9026,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110492720276896e-05,
+      "loss": 3.8941,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610212315157011e-05,
+      "loss": 3.8988,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609373720405959e-05,
+      "loss": 3.8959,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085351256549065e-05,
+      "loss": 3.923,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076965309038545e-05,
+      "loss": 3.8998,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068579361528025e-05,
+      "loss": 3.9058,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6060193414017505e-05,
+      "loss": 3.9104,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051807466506985e-05,
+      "loss": 3.9109,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043421518996465e-05,
+      "loss": 3.89,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6035051950289674e-05,
+      "loss": 3.9004,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026666002779154e-05,
+      "loss": 3.8921,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018280055268634e-05,
+      "loss": 3.899,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009894107758114e-05,
+      "loss": 3.8895,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600152453905133e-05,
+      "loss": 3.8911,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599313859154081e-05,
+      "loss": 3.9006,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598475264403029e-05,
+      "loss": 3.8995,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597636669651977e-05,
+      "loss": 3.8932,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596798074900925e-05,
+      "loss": 3.8985,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595959480149873e-05,
+      "loss": 3.8997,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595122523279194e-05,
+      "loss": 3.8932,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594283928528142e-05,
+      "loss": 3.8818,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59344533377709e-05,
+      "loss": 3.8968,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592606739026038e-05,
+      "loss": 3.8728,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591769782155359e-05,
+      "loss": 3.8882,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590931187404307e-05,
+      "loss": 3.8841,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5900925926532554e-05,
+      "loss": 3.8968,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5892539979022034e-05,
+      "loss": 3.8943,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5884154031511514e-05,
+      "loss": 3.8864,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5875768084000994e-05,
+      "loss": 3.8952,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867382136490474e-05,
+      "loss": 3.8928,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858996188979954e-05,
+      "loss": 3.8911,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585062662027316e-05,
+      "loss": 3.8762,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584224067276264e-05,
+      "loss": 3.8782,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583385472525212e-05,
+      "loss": 3.8949,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58254687777416e-05,
+      "loss": 3.8784,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581709920903481e-05,
+      "loss": 3.8759,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580871326152429e-05,
+      "loss": 3.8739,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580032731401377e-05,
+      "loss": 3.8829,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579194136650325e-05,
+      "loss": 3.8699,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578357179779647e-05,
+      "loss": 3.8822,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577518585028595e-05,
+      "loss": 3.8775,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576679990277543e-05,
+      "loss": 3.8822,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57584139552649e-05,
+      "loss": 3.8836,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5750044386558117e-05,
+      "loss": 3.8701,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741658439047596e-05,
+      "loss": 3.8698,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733272491537076e-05,
+      "loss": 3.8811,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572488654402655e-05,
+      "loss": 3.8641,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571650059651603e-05,
+      "loss": 3.8727,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5708131027809245e-05,
+      "loss": 3.88,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699745080298725e-05,
+      "loss": 3.8822,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691359132788205e-05,
+      "loss": 3.8647,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5682973185277685e-05,
+      "loss": 3.859,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674587237767165e-05,
+      "loss": 3.8647,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5666217669060374e-05,
+      "loss": 3.8768,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657831721549854e-05,
+      "loss": 3.8854,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649445774039334e-05,
+      "loss": 3.8714,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5641059826528814e-05,
+      "loss": 3.8828,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632673879018294e-05,
+      "loss": 3.8814,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624287931507774e-05,
+      "loss": 3.8791,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561591836280098e-05,
+      "loss": 3.8583,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560753241529046e-05,
+      "loss": 3.8784,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559914646777994e-05,
+      "loss": 3.8724,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559076052026943e-05,
+      "loss": 3.8633,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558239095156264e-05,
+      "loss": 3.8717,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557400500405212e-05,
+      "loss": 3.8691,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55656190565416e-05,
+      "loss": 3.8757,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555723310903108e-05,
+      "loss": 3.8714,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554884716152056e-05,
+      "loss": 3.8641,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554047759281377e-05,
+      "loss": 3.8704,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553209164530325e-05,
+      "loss": 3.8662,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552370569779273e-05,
+      "loss": 3.8688,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551531975028221e-05,
+      "loss": 3.8667,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550695018157542e-05,
+      "loss": 3.8589,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54985642340649e-05,
+      "loss": 3.8739,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5490178286554384e-05,
+      "loss": 3.8652,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481792339043864e-05,
+      "loss": 3.8588,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547342277033707e-05,
+      "loss": 3.8587,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546503682282655e-05,
+      "loss": 3.8762,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545665087531603e-05,
+      "loss": 3.8699,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544826492780551e-05,
+      "loss": 3.8618,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543989535909872e-05,
+      "loss": 3.8555,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54315094115882e-05,
+      "loss": 3.8642,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542312346407768e-05,
+      "loss": 3.8551,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541473751656716e-05,
+      "loss": 3.8774,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540636794786037e-05,
+      "loss": 3.8629,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539798200034985e-05,
+      "loss": 3.859,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538959605283933e-05,
+      "loss": 3.8523,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538121010532882e-05,
+      "loss": 3.8649,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372840536622026e-05,
+      "loss": 3.8371,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364454589111506e-05,
+      "loss": 3.8562,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5356068641600986e-05,
+      "loss": 3.8564,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347682694090466e-05,
+      "loss": 3.8581,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339313125383675e-05,
+      "loss": 3.8437,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5330927177873155e-05,
+      "loss": 3.8625,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5322541230362635e-05,
+      "loss": 3.8459,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314155282852115e-05,
+      "loss": 3.8667,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305769335341595e-05,
+      "loss": 3.8572,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297399766634804e-05,
+      "loss": 3.8517,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5289013819124284e-05,
+      "loss": 3.8687,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528062787161377e-05,
+      "loss": 3.8745,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527225830290698e-05,
+      "loss": 3.859,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526387235539646e-05,
+      "loss": 3.8486,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525548640788594e-05,
+      "loss": 3.8582,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524710046037542e-05,
+      "loss": 3.8487,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52387145128649e-05,
+      "loss": 3.8509,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523032856535438e-05,
+      "loss": 3.8654,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522194261784386e-05,
+      "loss": 3.8462,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521355667033333e-05,
+      "loss": 3.8628,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520518710162655e-05,
+      "loss": 3.8556,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519680115411603e-05,
+      "loss": 3.8569,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518841520660551e-05,
+      "loss": 3.8401,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518002925909499e-05,
+      "loss": 3.8592,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517164331158447e-05,
+      "loss": 3.8427,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5163273742877684e-05,
+      "loss": 3.8604,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515488779536716e-05,
+      "loss": 3.8535,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514650184785664e-05,
+      "loss": 3.8569,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513811590034612e-05,
+      "loss": 3.8343,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129746331639333e-05,
+      "loss": 3.8599,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121360384128807e-05,
+      "loss": 3.8482,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112974436618287e-05,
+      "loss": 3.8544,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104588489107766e-05,
+      "loss": 3.8493,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096218920400976e-05,
+      "loss": 3.8576,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508783297289046e-05,
+      "loss": 3.8554,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507944702537994e-05,
+      "loss": 3.8571,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507106107786942e-05,
+      "loss": 3.8396,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506269150916263e-05,
+      "loss": 3.8436,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505430556165211e-05,
+      "loss": 3.8474,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504591961414159e-05,
+      "loss": 3.8393,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503753366663107e-05,
+      "loss": 3.8514,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502916409792428e-05,
+      "loss": 3.856,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502077815041376e-05,
+      "loss": 3.8453,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501239220290324e-05,
+      "loss": 3.8384,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500400625539272e-05,
+      "loss": 3.8525,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.910076379776001,
+      "eval_runtime": 306.0279,
+      "eval_samples_per_second": 1246.916,
+      "eval_steps_per_second": 38.967,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499563668668593e-05,
+      "loss": 3.8299,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4987250739175416e-05,
+      "loss": 3.8367,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978864791664896e-05,
+      "loss": 3.8458,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970478844154376e-05,
+      "loss": 3.8411,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4962092896643856e-05,
+      "loss": 3.8526,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953706949133336e-05,
+      "loss": 3.8322,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945321001622816e-05,
+      "loss": 3.842,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936935054112296e-05,
+      "loss": 3.8269,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928565485405505e-05,
+      "loss": 3.8457,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4920179537894985e-05,
+      "loss": 3.8372,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911793590384465e-05,
+      "loss": 3.8409,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903407642873945e-05,
+      "loss": 3.8516,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489505445297088e-05,
+      "loss": 3.8295,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488666850546037e-05,
+      "loss": 3.8401,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487828255794985e-05,
+      "loss": 3.8295,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486989661043933e-05,
+      "loss": 3.834,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486151066292881e-05,
+      "loss": 3.8229,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485312471541829e-05,
+      "loss": 3.8324,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484473876790777e-05,
+      "loss": 3.8304,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483635282039725e-05,
+      "loss": 3.8519,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482798325169046e-05,
+      "loss": 3.8302,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481959730417994e-05,
+      "loss": 3.8384,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481121135666942e-05,
+      "loss": 3.8429,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48028254091589e-05,
+      "loss": 3.8437,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479445584045211e-05,
+      "loss": 3.8208,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478606989294159e-05,
+      "loss": 3.8365,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477768394543107e-05,
+      "loss": 3.8234,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476929799792055e-05,
+      "loss": 3.8317,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476092842921376e-05,
+      "loss": 3.8194,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475254248170324e-05,
+      "loss": 3.8255,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474415653419272e-05,
+      "loss": 3.8318,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47357705866822e-05,
+      "loss": 3.8337,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472740101797541e-05,
+      "loss": 3.8246,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471901507046489e-05,
+      "loss": 3.8341,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471062912295437e-05,
+      "loss": 3.8278,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470224317544385e-05,
+      "loss": 3.8326,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469387360673706e-05,
+      "loss": 3.8167,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468548765922654e-05,
+      "loss": 3.8286,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467710171171602e-05,
+      "loss": 3.8138,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466873214300924e-05,
+      "loss": 3.8197,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466034619549872e-05,
+      "loss": 3.8168,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46519602479882e-05,
+      "loss": 3.8312,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464357430047768e-05,
+      "loss": 3.8281,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4635204731770886e-05,
+      "loss": 3.8256,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626818784260366e-05,
+      "loss": 3.8282,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618432836749846e-05,
+      "loss": 3.829,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4610046889239326e-05,
+      "loss": 3.8232,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4601677320532535e-05,
+      "loss": 3.8165,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4593291373022015e-05,
+      "loss": 3.8055,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584905425511495e-05,
+      "loss": 3.8299,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576519478000975e-05,
+      "loss": 3.8177,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456814990929419e-05,
+      "loss": 3.8112,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455976396178367e-05,
+      "loss": 3.8129,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455137801427315e-05,
+      "loss": 3.8202,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454299206676263e-05,
+      "loss": 3.8064,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453462249805584e-05,
+      "loss": 3.8113,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452623655054532e-05,
+      "loss": 3.8132,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45178506030348e-05,
+      "loss": 3.8149,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450946465552428e-05,
+      "loss": 3.8266,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450109508681749e-05,
+      "loss": 3.8072,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449270913930697e-05,
+      "loss": 3.8075,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448432319179645e-05,
+      "loss": 3.8202,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447593724428593e-05,
+      "loss": 3.8061,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467567675579144e-05,
+      "loss": 3.8098,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4459181728068624e-05,
+      "loss": 3.8181,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4450795780558104e-05,
+      "loss": 3.818,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442409833047584e-05,
+      "loss": 3.8044,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443404026434079e-05,
+      "loss": 3.7978,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442565431683027e-05,
+      "loss": 3.7998,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441726836931975e-05,
+      "loss": 3.8136,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440888242180923e-05,
+      "loss": 3.8247,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440051285310244e-05,
+      "loss": 3.8076,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439212690559192e-05,
+      "loss": 3.8211,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43837409580814e-05,
+      "loss": 3.818,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437535501057088e-05,
+      "loss": 3.8228,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43669854418641e-05,
+      "loss": 3.8006,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435859949435358e-05,
+      "loss": 3.8157,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435021354684306e-05,
+      "loss": 3.81,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434182759933253e-05,
+      "loss": 3.8016,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433345803062575e-05,
+      "loss": 3.8141,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432507208311523e-05,
+      "loss": 3.8093,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431668613560471e-05,
+      "loss": 3.8105,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430830018809418e-05,
+      "loss": 3.8105,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299930619387396e-05,
+      "loss": 3.8079,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291544671876876e-05,
+      "loss": 3.8087,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4283158724366356e-05,
+      "loss": 3.8056,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274772776855836e-05,
+      "loss": 3.8103,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426640320814905e-05,
+      "loss": 3.8026,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425801726063853e-05,
+      "loss": 3.8023,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249631313128005e-05,
+      "loss": 3.8116,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4241245365617485e-05,
+      "loss": 3.8095,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42328757969107e-05,
+      "loss": 3.7979,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422448984940018e-05,
+      "loss": 3.8006,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216103901889654e-05,
+      "loss": 3.8173,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207717954379134e-05,
+      "loss": 3.812,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419934838567235e-05,
+      "loss": 3.8052,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419096243816182e-05,
+      "loss": 3.796,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418257649065131e-05,
+      "loss": 3.8033,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417419054314079e-05,
+      "loss": 3.7996,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165820974434005e-05,
+      "loss": 3.8128,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157451405727214e-05,
+      "loss": 3.8116,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4149065458216694e-05,
+      "loss": 3.8004,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4140679510706174e-05,
+      "loss": 3.7955,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4132293563195654e-05,
+      "loss": 3.8083,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412390761568513e-05,
+      "loss": 3.7765,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411552166817461e-05,
+      "loss": 3.8024,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410713572066409e-05,
+      "loss": 3.7932,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409874977315357e-05,
+      "loss": 3.8007,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4090380204446776e-05,
+      "loss": 3.7874,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408199425693626e-05,
+      "loss": 3.8021,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407360830942574e-05,
+      "loss": 3.796,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406522236191522e-05,
+      "loss": 3.801,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405686917201217e-05,
+      "loss": 3.804,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404848322450165e-05,
+      "loss": 3.7933,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404009727699113e-05,
+      "loss": 3.8113,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40317113294806e-05,
+      "loss": 3.8194,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402334176077382e-05,
+      "loss": 3.8015,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40149558132633e-05,
+      "loss": 3.7908,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400656986575277e-05,
+      "loss": 3.8053,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399818391824225e-05,
+      "loss": 3.7956,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398979797073173e-05,
+      "loss": 3.7891,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398141202322122e-05,
+      "loss": 3.8087,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39730260757107e-05,
+      "loss": 3.7891,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396464012820018e-05,
+      "loss": 3.8039,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3956270559493386e-05,
+      "loss": 3.7998,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947884611982866e-05,
+      "loss": 3.802,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3939498664472346e-05,
+      "loss": 3.7906,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3931112716961826e-05,
+      "loss": 3.7974,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922743148255035e-05,
+      "loss": 3.7925,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3914357200744515e-05,
+      "loss": 3.8048,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905971253233995e-05,
+      "loss": 3.796,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897585305723475e-05,
+      "loss": 3.8002,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3889215737016684e-05,
+      "loss": 3.7809,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388082978950617e-05,
+      "loss": 3.8041,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387244384199565e-05,
+      "loss": 3.7923,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386405789448513e-05,
+      "loss": 3.803,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385568832577834e-05,
+      "loss": 3.7933,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384730237826782e-05,
+      "loss": 3.802,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38389164307573e-05,
+      "loss": 3.7985,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383053048324678e-05,
+      "loss": 3.8068,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382216091453999e-05,
+      "loss": 3.7808,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381377496702947e-05,
+      "loss": 3.7932,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380538901951895e-05,
+      "loss": 3.7927,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379700307200843e-05,
+      "loss": 3.7851,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378863350330164e-05,
+      "loss": 3.7995,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3780247555791124e-05,
+      "loss": 3.7986,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771861608280604e-05,
+      "loss": 3.797,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763475660770084e-05,
+      "loss": 3.7785,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375510609206329e-05,
+      "loss": 3.7996,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8851914405822754,
+      "eval_runtime": 306.1573,
+      "eval_samples_per_second": 1246.389,
+      "eval_steps_per_second": 38.951,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374672014455277e-05,
+      "loss": 3.7799,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373833419704225e-05,
+      "loss": 3.7803,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372994824953173e-05,
+      "loss": 3.7953,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372156230202121e-05,
+      "loss": 3.7898,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371317635451069e-05,
+      "loss": 3.7949,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370479040700017e-05,
+      "loss": 3.7823,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369640445948965e-05,
+      "loss": 3.7855,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368801851197913e-05,
+      "loss": 3.777,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367964894327234e-05,
+      "loss": 3.7895,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367126299576182e-05,
+      "loss": 3.7854,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36628770482513e-05,
+      "loss": 3.7867,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365449110074079e-05,
+      "loss": 3.7973,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364613791083773e-05,
+      "loss": 3.7776,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363775196332721e-05,
+      "loss": 3.7877,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3629366015816687e-05,
+      "loss": 3.7813,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3620980068306167e-05,
+      "loss": 3.7797,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3612594120795647e-05,
+      "loss": 3.7726,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3604208173285126e-05,
+      "loss": 3.7819,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3595822225774606e-05,
+      "loss": 3.7757,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358743627826408e-05,
+      "loss": 3.7976,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3579066709557295e-05,
+      "loss": 3.7845,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3570680762046775e-05,
+      "loss": 3.7853,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3562294814536255e-05,
+      "loss": 3.7905,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3553908867025735e-05,
+      "loss": 3.7914,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354553929831895e-05,
+      "loss": 3.7687,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353716972961216e-05,
+      "loss": 3.7857,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352878378210164e-05,
+      "loss": 3.775,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352039783459112e-05,
+      "loss": 3.7793,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35120118870806e-05,
+      "loss": 3.77,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350362593957008e-05,
+      "loss": 3.7728,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349523999205955e-05,
+      "loss": 3.7832,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348685404454903e-05,
+      "loss": 3.7808,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347846809703851e-05,
+      "loss": 3.7772,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347009852833173e-05,
+      "loss": 3.7806,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346171258082121e-05,
+      "loss": 3.7784,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345332663331069e-05,
+      "loss": 3.781,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344494068580017e-05,
+      "loss": 3.7695,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343657111709338e-05,
+      "loss": 3.7692,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342818516958286e-05,
+      "loss": 3.7678,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341979922207234e-05,
+      "loss": 3.7684,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341141327456182e-05,
+      "loss": 3.7674,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.340304370585503e-05,
+      "loss": 3.7806,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339465775834451e-05,
+      "loss": 3.7788,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338627181083399e-05,
+      "loss": 3.7761,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337788586332347e-05,
+      "loss": 3.7843,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336951629461668e-05,
+      "loss": 3.7737,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336113034710616e-05,
+      "loss": 3.7757,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335274439959564e-05,
+      "loss": 3.7673,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334435845208512e-05,
+      "loss": 3.7538,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333598888337833e-05,
+      "loss": 3.782,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332760293586781e-05,
+      "loss": 3.768,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331921698835729e-05,
+      "loss": 3.7668,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331083104084677e-05,
+      "loss": 3.7647,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330246147213998e-05,
+      "loss": 3.7678,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329407552462946e-05,
+      "loss": 3.757,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328568957711894e-05,
+      "loss": 3.7639,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3277320008412156e-05,
+      "loss": 3.767,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3268934060901636e-05,
+      "loss": 3.7655,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3260548113391116e-05,
+      "loss": 3.7764,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3252162165880596e-05,
+      "loss": 3.7593,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3243792597173805e-05,
+      "loss": 3.7592,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3235406649663285e-05,
+      "loss": 3.774,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3227020702152765e-05,
+      "loss": 3.758,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3218634754642245e-05,
+      "loss": 3.7597,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3210265185935454e-05,
+      "loss": 3.7662,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3201879238424934e-05,
+      "loss": 3.7711,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3193493290914414e-05,
+      "loss": 3.7565,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3185107343403894e-05,
+      "loss": 3.7539,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317673777469711e-05,
+      "loss": 3.7475,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3168368205990326e-05,
+      "loss": 3.7665,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31599822584798e-05,
+      "loss": 3.7754,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315159631096928e-05,
+      "loss": 3.7629,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314321036345876e-05,
+      "loss": 3.7688,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313482441594824e-05,
+      "loss": 3.7708,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312643846843772e-05,
+      "loss": 3.7783,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31180525209272e-05,
+      "loss": 3.7541,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310968295222041e-05,
+      "loss": 3.7684,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310129700470989e-05,
+      "loss": 3.7628,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309291105719937e-05,
+      "loss": 3.7517,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308452510968885e-05,
+      "loss": 3.7682,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307613916217833e-05,
+      "loss": 3.7663,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067769593471544e-05,
+      "loss": 3.7634,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3059383645961024e-05,
+      "loss": 3.7655,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3050997698450504e-05,
+      "loss": 3.7612,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042611750939984e-05,
+      "loss": 3.7568,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3034225803429464e-05,
+      "loss": 3.7625,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025839855918944e-05,
+      "loss": 3.7649,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3017453908408424e-05,
+      "loss": 3.7532,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3009067960897903e-05,
+      "loss": 3.7575,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300069839219111e-05,
+      "loss": 3.7667,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299231244468059e-05,
+      "loss": 3.7633,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29839428759738e-05,
+      "loss": 3.7515,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297555692846328e-05,
+      "loss": 3.7534,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296717098095276e-05,
+      "loss": 3.7708,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295878503344224e-05,
+      "loss": 3.7667,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295039908593173e-05,
+      "loss": 3.7629,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294202951722494e-05,
+      "loss": 3.7471,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293364356971442e-05,
+      "loss": 3.7577,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29252576222039e-05,
+      "loss": 3.7545,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291687167469338e-05,
+      "loss": 3.7638,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2908502105986586e-05,
+      "loss": 3.7681,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2900116158476066e-05,
+      "loss": 3.7551,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2891730210965546e-05,
+      "loss": 3.7482,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2883344263455026e-05,
+      "loss": 3.7624,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2874958315944506e-05,
+      "loss": 3.7343,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286657236843398e-05,
+      "loss": 3.7543,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2858186420923466e-05,
+      "loss": 3.7451,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2849800473412946e-05,
+      "loss": 3.759,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284143090470616e-05,
+      "loss": 3.7436,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283306133599937e-05,
+      "loss": 3.7583,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282467538848885e-05,
+      "loss": 3.7483,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281628944097833e-05,
+      "loss": 3.7567,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280790349346781e-05,
+      "loss": 3.7626,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2799517545957284e-05,
+      "loss": 3.7518,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27911479772505e-05,
+      "loss": 3.7649,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278276202973998e-05,
+      "loss": 3.7724,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277437608222945e-05,
+      "loss": 3.7586,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276599013471893e-05,
+      "loss": 3.7501,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275760418720842e-05,
+      "loss": 3.755,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27492182396979e-05,
+      "loss": 3.7554,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274083229218738e-05,
+      "loss": 3.7418,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273244634467686e-05,
+      "loss": 3.7684,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272407677597007e-05,
+      "loss": 3.744,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271569082845955e-05,
+      "loss": 3.7615,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270730488094903e-05,
+      "loss": 3.7554,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269891893343851e-05,
+      "loss": 3.7558,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269054936473172e-05,
+      "loss": 3.7482,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26821634172212e-05,
+      "loss": 3.7541,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267377746971068e-05,
+      "loss": 3.7463,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2665407901003887e-05,
+      "loss": 3.7624,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265702195349337e-05,
+      "loss": 3.7516,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264863600598285e-05,
+      "loss": 3.7584,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264025005847233e-05,
+      "loss": 3.7376,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263186411096181e-05,
+      "loss": 3.7606,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262349454225502e-05,
+      "loss": 3.7485,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26151085947445e-05,
+      "loss": 3.7588,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260672264723398e-05,
+      "loss": 3.7522,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.259833669972346e-05,
+      "loss": 3.7573,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258996713101667e-05,
+      "loss": 3.7538,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258158118350615e-05,
+      "loss": 3.7678,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257319523599563e-05,
+      "loss": 3.7375,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256480928848511e-05,
+      "loss": 3.7488,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255642334097459e-05,
+      "loss": 3.7481,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254805377226781e-05,
+      "loss": 3.7465,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253966782475729e-05,
+      "loss": 3.7482,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253128187724677e-05,
+      "loss": 3.7561,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2522912308539976e-05,
+      "loss": 3.7555,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2514526361029456e-05,
+      "loss": 3.7355,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2506140413518936e-05,
+      "loss": 3.76,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8692798614501953,
+      "eval_runtime": 308.5111,
+      "eval_samples_per_second": 1236.879,
+      "eval_steps_per_second": 38.653,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2497754466008416e-05,
+      "loss": 3.7453,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2489368518497896e-05,
+      "loss": 3.7331,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2480982570987376e-05,
+      "loss": 3.7544,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2472613002280585e-05,
+      "loss": 3.747,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2464227054770065e-05,
+      "loss": 3.7512,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2455841107259545e-05,
+      "loss": 3.7406,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2447455159749025e-05,
+      "loss": 3.7398,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243908559104224e-05,
+      "loss": 3.7367,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243069964353172e-05,
+      "loss": 3.7463,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.24223136960212e-05,
+      "loss": 3.7425,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.241392774851068e-05,
+      "loss": 3.7457,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.240555817980389e-05,
+      "loss": 3.755,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.239717223229337e-05,
+      "loss": 3.7409,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238878628478285e-05,
+      "loss": 3.7413,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238040033727233e-05,
+      "loss": 3.7424,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237203076856554e-05,
+      "loss": 3.7345,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.236364482105502e-05,
+      "loss": 3.73,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23552588735445e-05,
+      "loss": 3.7393,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.234687292603398e-05,
+      "loss": 3.7357,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233848697852346e-05,
+      "loss": 3.7527,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2330101031012945e-05,
+      "loss": 3.7439,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232171508350242e-05,
+      "loss": 3.7455,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23133291359919e-05,
+      "loss": 3.7498,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2304959567285114e-05,
+      "loss": 3.7485,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.229657361977459e-05,
+      "loss": 3.7308,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.22882040510678e-05,
+      "loss": 3.7454,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227981810355728e-05,
+      "loss": 3.7302,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227143215604676e-05,
+      "loss": 3.7362,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2263046208536236e-05,
+      "loss": 3.7343,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.225467663982945e-05,
+      "loss": 3.7306,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.224629069231893e-05,
+      "loss": 3.7434,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.223790474480841e-05,
+      "loss": 3.7385,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222951879729789e-05,
+      "loss": 3.7361,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222113284978737e-05,
+      "loss": 3.7386,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.221274690227685e-05,
+      "loss": 3.7375,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.220437733357006e-05,
+      "loss": 3.7423,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.219599138605954e-05,
+      "loss": 3.7309,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.218760543854902e-05,
+      "loss": 3.7298,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.21792194910385e-05,
+      "loss": 3.7265,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217084992233171e-05,
+      "loss": 3.7277,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.216246397482119e-05,
+      "loss": 3.7282,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.215407802731067e-05,
+      "loss": 3.7384,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.214569207980015e-05,
+      "loss": 3.736,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2137322511093366e-05,
+      "loss": 3.7366,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2128936563582846e-05,
+      "loss": 3.7425,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2120550616072326e-05,
+      "loss": 3.7358,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2112164668561806e-05,
+      "loss": 3.7357,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2103795099855015e-05,
+      "loss": 3.7252,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2095409152344495e-05,
+      "loss": 3.7155,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2087023204833975e-05,
+      "loss": 3.7437,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2078637257323454e-05,
+      "loss": 3.7304,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2070251309812934e-05,
+      "loss": 3.7233,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2061881741106144e-05,
+      "loss": 3.7299,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2053495793595623e-05,
+      "loss": 3.7237,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2045109846085103e-05,
+      "loss": 3.7216,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.203672389857459e-05,
+      "loss": 3.7212,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.202833795106407e-05,
+      "loss": 3.7264,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.201995200355355e-05,
+      "loss": 3.728,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.201156605604303e-05,
+      "loss": 3.7386,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.200318010853251e-05,
+      "loss": 3.7206,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199481053982572e-05,
+      "loss": 3.7175,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19864245923152e-05,
+      "loss": 3.7411,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197803864480468e-05,
+      "loss": 3.7122,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196965269729416e-05,
+      "loss": 3.7203,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196126674978364e-05,
+      "loss": 3.7298,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195289718107685e-05,
+      "loss": 3.7374,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.194451123356633e-05,
+      "loss": 3.7161,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.193612528605581e-05,
+      "loss": 3.7147,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.192773933854529e-05,
+      "loss": 3.7086,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1919353391034775e-05,
+      "loss": 3.7244,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191096744352425e-05,
+      "loss": 3.7393,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1902597874817464e-05,
+      "loss": 3.7266,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1894211927306944e-05,
+      "loss": 3.7264,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1885825979796424e-05,
+      "loss": 3.7325,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18774400322859e-05,
+      "loss": 3.74,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186905408477538e-05,
+      "loss": 3.7172,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186068451606859e-05,
+      "loss": 3.7304,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.185229856855807e-05,
+      "loss": 3.7232,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1843912621047546e-05,
+      "loss": 3.7144,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1835526673537026e-05,
+      "loss": 3.7281,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.182715710483024e-05,
+      "loss": 3.7288,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181877115731972e-05,
+      "loss": 3.7294,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18103852098092e-05,
+      "loss": 3.7236,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180199926229868e-05,
+      "loss": 3.7272,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.179361331478816e-05,
+      "loss": 3.7174,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.178524374608137e-05,
+      "loss": 3.7258,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.177685779857085e-05,
+      "loss": 3.7239,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.176847185106033e-05,
+      "loss": 3.7186,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.176008590354981e-05,
+      "loss": 3.7218,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.175171633484302e-05,
+      "loss": 3.7213,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.17433303873325e-05,
+      "loss": 3.7304,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.173494443982198e-05,
+      "loss": 3.7165,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1726558492311466e-05,
+      "loss": 3.7142,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1718188923604675e-05,
+      "loss": 3.7341,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1709802976094155e-05,
+      "loss": 3.7276,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1701417028583635e-05,
+      "loss": 3.7264,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1693031081073115e-05,
+      "loss": 3.7102,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1684645133562595e-05,
+      "loss": 3.7227,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1676259186052075e-05,
+      "loss": 3.7167,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1667873238541555e-05,
+      "loss": 3.7243,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1659503669834764e-05,
+      "loss": 3.7319,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1651117722324244e-05,
+      "loss": 3.7177,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1642731774813724e-05,
+      "loss": 3.7132,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1634345827303204e-05,
+      "loss": 3.7221,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1625959879792684e-05,
+      "loss": 3.698,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.16175903110859e-05,
+      "loss": 3.7242,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160920436357538e-05,
+      "loss": 3.7048,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160081841606486e-05,
+      "loss": 3.7239,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.159243246855434e-05,
+      "loss": 3.7056,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.158406289984755e-05,
+      "loss": 3.7206,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.157567695233703e-05,
+      "loss": 3.7143,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.156729100482651e-05,
+      "loss": 3.7155,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155890505731599e-05,
+      "loss": 3.728,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155051910980547e-05,
+      "loss": 3.7122,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154214954109868e-05,
+      "loss": 3.7278,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.153376359358816e-05,
+      "loss": 3.7337,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.152537764607764e-05,
+      "loss": 3.7208,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151699169856712e-05,
+      "loss": 3.7183,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150862212986033e-05,
+      "loss": 3.714,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150023618234981e-05,
+      "loss": 3.7173,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.149185023483929e-05,
+      "loss": 3.711,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.148346428732877e-05,
+      "loss": 3.7285,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.147507833981825e-05,
+      "loss": 3.7037,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.146670877111146e-05,
+      "loss": 3.7247,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.145832282360094e-05,
+      "loss": 3.7212,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144993687609042e-05,
+      "loss": 3.7195,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14415509285799e-05,
+      "loss": 3.7091,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.143318135987311e-05,
+      "loss": 3.721,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.142479541236259e-05,
+      "loss": 3.7068,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.141640946485207e-05,
+      "loss": 3.7287,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.140802351734155e-05,
+      "loss": 3.7123,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139963756983103e-05,
+      "loss": 3.7237,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139126800112425e-05,
+      "loss": 3.7007,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.138288205361373e-05,
+      "loss": 3.7268,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.137449610610321e-05,
+      "loss": 3.7141,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.136611015859268e-05,
+      "loss": 3.7221,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1357740589885896e-05,
+      "loss": 3.7158,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1349354642375376e-05,
+      "loss": 3.7149,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1340968694864856e-05,
+      "loss": 3.7205,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.133258274735433e-05,
+      "loss": 3.735,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1324213178647545e-05,
+      "loss": 3.6997,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1315827231137025e-05,
+      "loss": 3.7152,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1307441283626505e-05,
+      "loss": 3.7111,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1299055336115985e-05,
+      "loss": 3.7156,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1290669388605465e-05,
+      "loss": 3.7102,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128229981989868e-05,
+      "loss": 3.7191,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1273913872388154e-05,
+      "loss": 3.7191,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1265527924877634e-05,
+      "loss": 3.7033,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1257141977367114e-05,
+      "loss": 3.7243,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.858419418334961,
+      "eval_runtime": 308.276,
+      "eval_samples_per_second": 1237.823,
+      "eval_steps_per_second": 38.683,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124877240866032e-05,
+      "loss": 3.7082,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.12403864611498e-05,
+      "loss": 3.7005,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123200051363928e-05,
+      "loss": 3.7185,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.122361456612876e-05,
+      "loss": 3.7072,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.121524499742198e-05,
+      "loss": 3.7171,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.120685904991146e-05,
+      "loss": 3.7044,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119847310240094e-05,
+      "loss": 3.7078,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119008715489042e-05,
+      "loss": 3.7025,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118171758618363e-05,
+      "loss": 3.7133,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117333163867311e-05,
+      "loss": 3.7035,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.116494569116259e-05,
+      "loss": 3.7127,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115655974365207e-05,
+      "loss": 3.7182,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1148190174945276e-05,
+      "loss": 3.7105,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1139804227434756e-05,
+      "loss": 3.7066,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1131418279924236e-05,
+      "loss": 3.703,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1123032332413716e-05,
+      "loss": 3.6982,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1114646384903196e-05,
+      "loss": 3.6964,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110627681619641e-05,
+      "loss": 3.7058,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109789086868589e-05,
+      "loss": 3.7044,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108950492117537e-05,
+      "loss": 3.7125,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108111897366485e-05,
+      "loss": 3.7152,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.107274940495806e-05,
+      "loss": 3.7078,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106436345744754e-05,
+      "loss": 3.7139,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.105597750993702e-05,
+      "loss": 3.714,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10475915624265e-05,
+      "loss": 3.6951,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103922199371971e-05,
+      "loss": 3.7127,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103083604620919e-05,
+      "loss": 3.6996,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.102245009869867e-05,
+      "loss": 3.7017,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101406415118815e-05,
+      "loss": 3.6987,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1005694582481366e-05,
+      "loss": 3.7019,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0997308634970846e-05,
+      "loss": 3.7052,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0988922687460326e-05,
+      "loss": 3.7082,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0980536739949806e-05,
+      "loss": 3.7035,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0972150792439286e-05,
+      "loss": 3.7067,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0963781223732495e-05,
+      "loss": 3.6967,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0955395276221975e-05,
+      "loss": 3.709,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0947009328711455e-05,
+      "loss": 3.7052,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0938623381200935e-05,
+      "loss": 3.689,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0930253812494144e-05,
+      "loss": 3.693,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0921867864983624e-05,
+      "loss": 3.6937,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0913481917473104e-05,
+      "loss": 3.7005,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0905095969962583e-05,
+      "loss": 3.7023,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08967264012558e-05,
+      "loss": 3.7041,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.088834045374528e-05,
+      "loss": 3.6999,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087995450623476e-05,
+      "loss": 3.7121,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087156855872424e-05,
+      "loss": 3.7021,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.086319899001745e-05,
+      "loss": 3.7043,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085481304250693e-05,
+      "loss": 3.6926,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084642709499641e-05,
+      "loss": 3.6757,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083804114748589e-05,
+      "loss": 3.7116,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08296715787791e-05,
+      "loss": 3.6975,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082128563126858e-05,
+      "loss": 3.6956,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081289968375806e-05,
+      "loss": 3.6946,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080451373624754e-05,
+      "loss": 3.6908,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079614416754075e-05,
+      "loss": 3.6894,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078775822003023e-05,
+      "loss": 3.6871,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077937227251971e-05,
+      "loss": 3.6941,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077098632500919e-05,
+      "loss": 3.6903,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076260037749867e-05,
+      "loss": 3.7087,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075423080879188e-05,
+      "loss": 3.6895,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.074584486128136e-05,
+      "loss": 3.6851,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.073745891377084e-05,
+      "loss": 3.7067,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072907296626032e-05,
+      "loss": 3.6798,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072070339755353e-05,
+      "loss": 3.6853,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071231745004301e-05,
+      "loss": 3.6984,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.070393150253249e-05,
+      "loss": 3.7017,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069554555502198e-05,
+      "loss": 3.6855,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068717598631519e-05,
+      "loss": 3.6826,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0678790038804667e-05,
+      "loss": 3.6779,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0670404091294147e-05,
+      "loss": 3.6916,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0662018143783627e-05,
+      "loss": 3.7064,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0653648575076836e-05,
+      "loss": 3.6941,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0645262627566316e-05,
+      "loss": 3.6943,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0636876680055796e-05,
+      "loss": 3.7026,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0628490732545275e-05,
+      "loss": 3.7069,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0620104785034755e-05,
+      "loss": 3.6823,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0611735216327965e-05,
+      "loss": 3.7002,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0603349268817444e-05,
+      "loss": 3.6876,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.059496332130693e-05,
+      "loss": 3.6884,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0586577373796404e-05,
+      "loss": 3.6919,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.057820780508962e-05,
+      "loss": 3.7014,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05698218575791e-05,
+      "loss": 3.6944,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056143591006858e-05,
+      "loss": 3.6973,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055304996255805e-05,
+      "loss": 3.6896,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054466401504753e-05,
+      "loss": 3.6824,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053629444634075e-05,
+      "loss": 3.6951,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.052790849883023e-05,
+      "loss": 3.6923,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05195225513197e-05,
+      "loss": 3.6871,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051113660380918e-05,
+      "loss": 3.6886,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.05027670351024e-05,
+      "loss": 3.6898,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049438108759188e-05,
+      "loss": 3.6986,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.048599514008136e-05,
+      "loss": 3.6858,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.047760919257084e-05,
+      "loss": 3.6811,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0469239623864054e-05,
+      "loss": 3.7016,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046085367635353e-05,
+      "loss": 3.6975,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.045246772884301e-05,
+      "loss": 3.6934,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044408178133249e-05,
+      "loss": 3.6815,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04357122126257e-05,
+      "loss": 3.6874,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0427326265115176e-05,
+      "loss": 3.6833,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0418940317604656e-05,
+      "loss": 3.6934,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0410554370094136e-05,
+      "loss": 3.7006,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040216842258362e-05,
+      "loss": 3.6906,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039379885387683e-05,
+      "loss": 3.6838,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038541290636631e-05,
+      "loss": 3.684,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037702695885579e-05,
+      "loss": 3.6722,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036864101134527e-05,
+      "loss": 3.6861,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036027144263848e-05,
+      "loss": 3.6751,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035188549512796e-05,
+      "loss": 3.6942,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034349954761744e-05,
+      "loss": 3.6742,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033511360010692e-05,
+      "loss": 3.6875,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032674403140013e-05,
+      "loss": 3.6813,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031835808388961e-05,
+      "loss": 3.689,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030997213637909e-05,
+      "loss": 3.6935,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0301586188868576e-05,
+      "loss": 3.6825,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0293216620161785e-05,
+      "loss": 3.698,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0284830672651265e-05,
+      "loss": 3.7001,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0276444725140745e-05,
+      "loss": 3.6887,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0268058777630225e-05,
+      "loss": 3.6891,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0259689208923434e-05,
+      "loss": 3.6813,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0251303261412914e-05,
+      "loss": 3.6859,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0242917313902394e-05,
+      "loss": 3.6855,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0234531366391874e-05,
+      "loss": 3.6915,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.022616179768508e-05,
+      "loss": 3.6757,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.021777585017456e-05,
+      "loss": 3.6931,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020938990266404e-05,
+      "loss": 3.6923,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020100395515353e-05,
+      "loss": 3.6887,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.019263438644674e-05,
+      "loss": 3.6761,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.018424843893622e-05,
+      "loss": 3.6909,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01758624914257e-05,
+      "loss": 3.6774,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.016747654391518e-05,
+      "loss": 3.6946,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015910697520839e-05,
+      "loss": 3.6864,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015072102769787e-05,
+      "loss": 3.6889,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014233508018735e-05,
+      "loss": 3.6695,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013394913267683e-05,
+      "loss": 3.6996,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.012557956397004e-05,
+      "loss": 3.68,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011719361645952e-05,
+      "loss": 3.6902,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0108807668949e-05,
+      "loss": 3.686,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0100421721438484e-05,
+      "loss": 3.6819,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009205215273169e-05,
+      "loss": 3.6917,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008366620522117e-05,
+      "loss": 3.7014,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007528025771065e-05,
+      "loss": 3.671,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006689431020013e-05,
+      "loss": 3.6857,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005852474149334e-05,
+      "loss": 3.6812,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005013879398282e-05,
+      "loss": 3.6833,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00417528464723e-05,
+      "loss": 3.6803,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.003336689896178e-05,
+      "loss": 3.6901,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.002499733025499e-05,
+      "loss": 3.6904,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.001661138274447e-05,
+      "loss": 3.6726,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000822543523395e-05,
+      "loss": 3.6919,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.852022171020508,
+      "eval_runtime": 308.9743,
+      "eval_samples_per_second": 1235.025,
+      "eval_steps_per_second": 38.595,
+      "step": 610560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999983948772343e-05,
+      "loss": 3.6791,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999145354021292e-05,
+      "loss": 3.6737,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9983083971506126e-05,
+      "loss": 3.6831,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9974698023995606e-05,
+      "loss": 3.6766,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9966312076485086e-05,
+      "loss": 3.6884,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9957926128974566e-05,
+      "loss": 3.6699,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9949556560267775e-05,
+      "loss": 3.6848,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9941170612757255e-05,
+      "loss": 3.67,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9932784665246735e-05,
+      "loss": 3.6814,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9924398717736215e-05,
+      "loss": 3.675,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9916029149029424e-05,
+      "loss": 3.6788,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9907643201518904e-05,
+      "loss": 3.6897,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9899257254008384e-05,
+      "loss": 3.6782,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989087130649787e-05,
+      "loss": 3.6769,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.988250173779108e-05,
+      "loss": 3.6775,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.987411579028056e-05,
+      "loss": 3.6661,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.986572984277004e-05,
+      "loss": 3.6678,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.985734389525952e-05,
+      "loss": 3.6793,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984897432655273e-05,
+      "loss": 3.6749,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984058837904221e-05,
+      "loss": 3.6782,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983220243153169e-05,
+      "loss": 3.6853,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.982381648402117e-05,
+      "loss": 3.6805,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.981544691531438e-05,
+      "loss": 3.6853,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.980706096780386e-05,
+      "loss": 3.6814,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.979867502029334e-05,
+      "loss": 3.6676,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9790289072782825e-05,
+      "loss": 3.6822,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9781919504076034e-05,
+      "loss": 3.6688,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9773533556565514e-05,
+      "loss": 3.6729,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9765147609054994e-05,
+      "loss": 3.6735,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9756761661544474e-05,
+      "loss": 3.6706,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.974839209283768e-05,
+      "loss": 3.6737,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.974000614532716e-05,
+      "loss": 3.6792,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973162019781664e-05,
+      "loss": 3.675,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.972323425030612e-05,
+      "loss": 3.6779,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.971486468159933e-05,
+      "loss": 3.6675,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.970647873408881e-05,
+      "loss": 3.6805,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.969809278657829e-05,
+      "loss": 3.6749,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968970683906778e-05,
+      "loss": 3.6623,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968133727036099e-05,
+      "loss": 3.6661,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.967295132285047e-05,
+      "loss": 3.6622,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.966456537533995e-05,
+      "loss": 3.6718,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.965617942782943e-05,
+      "loss": 3.6744,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9647809859122636e-05,
+      "loss": 3.6744,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9639423911612116e-05,
+      "loss": 3.6747,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9631037964101596e-05,
+      "loss": 3.6791,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9622652016591076e-05,
+      "loss": 3.6751,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9614282447884285e-05,
+      "loss": 3.6759,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9605896500373765e-05,
+      "loss": 3.666,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9597510552863245e-05,
+      "loss": 3.6422,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958912460535273e-05,
+      "loss": 3.6878,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958075503664594e-05,
+      "loss": 3.6663,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957236908913542e-05,
+      "loss": 3.669,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.95639831416249e-05,
+      "loss": 3.6648,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9555597194114374e-05,
+      "loss": 3.6628,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.954722762540759e-05,
+      "loss": 3.659,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953884167789707e-05,
+      "loss": 3.6573,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953045573038655e-05,
+      "loss": 3.6684,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952206978287602e-05,
+      "loss": 3.6624,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.951370021416924e-05,
+      "loss": 3.6802,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.950531426665872e-05,
+      "loss": 3.6634,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.94969283191482e-05,
+      "loss": 3.6558,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948854237163768e-05,
+      "loss": 3.6763,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9480172802930895e-05,
+      "loss": 3.6535,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9471786855420375e-05,
+      "loss": 3.6578,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.946340090790985e-05,
+      "loss": 3.6712,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.945501496039933e-05,
+      "loss": 3.6721,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9446645391692544e-05,
+      "loss": 3.6571,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9438259444182024e-05,
+      "loss": 3.6535,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.94298734966715e-05,
+      "loss": 3.6483,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942148754916098e-05,
+      "loss": 3.6629,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.941311798045419e-05,
+      "loss": 3.6779,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.940473203294367e-05,
+      "loss": 3.6642,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.939634608543315e-05,
+      "loss": 3.6694,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.938796013792263e-05,
+      "loss": 3.67,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937957419041211e-05,
+      "loss": 3.6773,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937120462170532e-05,
+      "loss": 3.6588,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.93628186741948e-05,
+      "loss": 3.6723,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.935443272668428e-05,
+      "loss": 3.6564,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.934604677917376e-05,
+      "loss": 3.6611,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.933767721046697e-05,
+      "loss": 3.6631,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932929126295645e-05,
+      "loss": 3.671,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932090531544593e-05,
+      "loss": 3.6646,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.931251936793541e-05,
+      "loss": 3.6725,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9304149799228626e-05,
+      "loss": 3.6593,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9295763851718106e-05,
+      "loss": 3.6567,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9287377904207586e-05,
+      "loss": 3.6682,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9278991956697066e-05,
+      "loss": 3.6639,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9270622387990275e-05,
+      "loss": 3.6545,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9262236440479755e-05,
+      "loss": 3.6649,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9253850492969235e-05,
+      "loss": 3.6655,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9245464545458715e-05,
+      "loss": 3.6707,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9237094976751924e-05,
+      "loss": 3.6554,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9228709029241404e-05,
+      "loss": 3.6533,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9220323081730884e-05,
+      "loss": 3.6733,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9211937134220364e-05,
+      "loss": 3.6683,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.920356756551358e-05,
+      "loss": 3.6658,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.919518161800306e-05,
+      "loss": 3.655,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.918679567049254e-05,
+      "loss": 3.661,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.917840972298202e-05,
+      "loss": 3.6573,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.917004015427523e-05,
+      "loss": 3.6661,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.916165420676471e-05,
+      "loss": 3.6723,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.915326825925419e-05,
+      "loss": 3.6635,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.914488231174367e-05,
+      "loss": 3.6557,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.913651274303688e-05,
+      "loss": 3.6527,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.912812679552636e-05,
+      "loss": 3.651,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911974084801584e-05,
+      "loss": 3.6547,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9111371279309054e-05,
+      "loss": 3.6483,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9102985331798534e-05,
+      "loss": 3.6669,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9094599384288014e-05,
+      "loss": 3.6486,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9086213436777494e-05,
+      "loss": 3.6579,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9077827489266973e-05,
+      "loss": 3.6524,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906945792056018e-05,
+      "loss": 3.6645,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906107197304966e-05,
+      "loss": 3.6662,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.905268602553914e-05,
+      "loss": 3.6526,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.904431645683235e-05,
+      "loss": 3.6729,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.903593050932183e-05,
+      "loss": 3.6741,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.902754456181131e-05,
+      "loss": 3.6631,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.901915861430079e-05,
+      "loss": 3.662,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9010789045594e-05,
+      "loss": 3.6547,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.900240309808349e-05,
+      "loss": 3.6533,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.899401715057297e-05,
+      "loss": 3.6662,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.898563120306245e-05,
+      "loss": 3.6616,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.897724525555193e-05,
+      "loss": 3.6496,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8968875686845136e-05,
+      "loss": 3.6671,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8960489739334616e-05,
+      "loss": 3.6666,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8952103791824096e-05,
+      "loss": 3.6618,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8943717844313576e-05,
+      "loss": 3.6513,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8935348275606785e-05,
+      "loss": 3.6591,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8926962328096265e-05,
+      "loss": 3.6511,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8918576380585745e-05,
+      "loss": 3.6684,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8910190433075225e-05,
+      "loss": 3.6587,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.890182086436844e-05,
+      "loss": 3.6656,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.889343491685792e-05,
+      "loss": 3.6438,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.88850489693474e-05,
+      "loss": 3.6714,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.887666302183688e-05,
+      "loss": 3.6497,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.886829345313009e-05,
+      "loss": 3.6669,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885990750561957e-05,
+      "loss": 3.6623,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885152155810905e-05,
+      "loss": 3.6514,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.884313561059853e-05,
+      "loss": 3.6645,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.883476604189174e-05,
+      "loss": 3.6782,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.882638009438122e-05,
+      "loss": 3.6431,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.88179941468707e-05,
+      "loss": 3.6624,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880960819936018e-05,
+      "loss": 3.654,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8801238630653395e-05,
+      "loss": 3.6542,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8792852683142875e-05,
+      "loss": 3.6553,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8784466735632355e-05,
+      "loss": 3.6619,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8776080788121834e-05,
+      "loss": 3.6611,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8767711219415044e-05,
+      "loss": 3.6526,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8759325271904524e-05,
+      "loss": 3.6631,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.846881151199341,
+      "eval_runtime": 307.4882,
+      "eval_samples_per_second": 1240.994,
+      "eval_steps_per_second": 38.782,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8750939324394004e-05,
+      "loss": 3.6466,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8742553376883483e-05,
+      "loss": 3.6474,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8734167429372963e-05,
+      "loss": 3.6584,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.872578148186244e-05,
+      "loss": 3.6523,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.871741191315565e-05,
+      "loss": 3.6622,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870902596564513e-05,
+      "loss": 3.6455,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870064001813461e-05,
+      "loss": 3.6577,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869225407062409e-05,
+      "loss": 3.6474,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.868388450191731e-05,
+      "loss": 3.6522,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.867549855440679e-05,
+      "loss": 3.6533,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.866711260689627e-05,
+      "loss": 3.6505,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865872665938575e-05,
+      "loss": 3.6598,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8650373469482686e-05,
+      "loss": 3.6625,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8641987521972166e-05,
+      "loss": 3.645,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8633601574461646e-05,
+      "loss": 3.6504,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8625215626951126e-05,
+      "loss": 3.6418,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8616829679440606e-05,
+      "loss": 3.64,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8608443731930086e-05,
+      "loss": 3.6531,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8600057784419566e-05,
+      "loss": 3.6461,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8591671836909046e-05,
+      "loss": 3.6536,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.858330226820226e-05,
+      "loss": 3.6603,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.857491632069174e-05,
+      "loss": 3.6579,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.856653037318122e-05,
+      "loss": 3.6603,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8558144425670695e-05,
+      "loss": 3.6561,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854977485696391e-05,
+      "loss": 3.6426,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854138890945339e-05,
+      "loss": 3.6558,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.853300296194287e-05,
+      "loss": 3.6443,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8524617014432344e-05,
+      "loss": 3.6517,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.851624744572556e-05,
+      "loss": 3.6445,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.850786149821504e-05,
+      "loss": 3.6444,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849947555070452e-05,
+      "loss": 3.6471,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8491089603194e-05,
+      "loss": 3.652,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8482720034487216e-05,
+      "loss": 3.6502,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8474334086976695e-05,
+      "loss": 3.6498,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.846594813946617e-05,
+      "loss": 3.6427,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.845756219195565e-05,
+      "loss": 3.6512,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8449192623248865e-05,
+      "loss": 3.6521,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8440806675738344e-05,
+      "loss": 3.6391,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843242072822782e-05,
+      "loss": 3.6421,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84240347807173e-05,
+      "loss": 3.6382,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8415665212010513e-05,
+      "loss": 3.6411,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8407279264499993e-05,
+      "loss": 3.6473,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839889331698947e-05,
+      "loss": 3.6521,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839050736947895e-05,
+      "loss": 3.6504,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838213780077217e-05,
+      "loss": 3.6523,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837375185326164e-05,
+      "loss": 3.6523,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836536590575112e-05,
+      "loss": 3.6483,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.83569799582406e-05,
+      "loss": 3.6414,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834861038953382e-05,
+      "loss": 3.6199,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834022444202329e-05,
+      "loss": 3.6614,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833183849451277e-05,
+      "loss": 3.6404,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.832345254700225e-05,
+      "loss": 3.6439,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.831508297829547e-05,
+      "loss": 3.6442,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.830669703078495e-05,
+      "loss": 3.6379,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.829831108327443e-05,
+      "loss": 3.6331,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828992513576391e-05,
+      "loss": 3.6341,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8281555567057116e-05,
+      "loss": 3.6453,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8273169619546596e-05,
+      "loss": 3.636,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8264783672036076e-05,
+      "loss": 3.6562,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8256397724525556e-05,
+      "loss": 3.64,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8248028155818765e-05,
+      "loss": 3.6274,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8239642208308245e-05,
+      "loss": 3.6536,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8231256260797725e-05,
+      "loss": 3.6337,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8222870313287205e-05,
+      "loss": 3.631,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8214500744580414e-05,
+      "loss": 3.6465,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82061147970699e-05,
+      "loss": 3.651,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.819772884955938e-05,
+      "loss": 3.6319,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818934290204886e-05,
+      "loss": 3.6294,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818097333334207e-05,
+      "loss": 3.6254,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817258738583155e-05,
+      "loss": 3.6385,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.816420143832103e-05,
+      "loss": 3.6507,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.815581549081051e-05,
+      "loss": 3.6428,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.814744592210372e-05,
+      "loss": 3.6453,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.81390599745932e-05,
+      "loss": 3.6411,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813067402708268e-05,
+      "loss": 3.6505,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812228807957216e-05,
+      "loss": 3.6363,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.811391851086537e-05,
+      "loss": 3.6512,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810553256335485e-05,
+      "loss": 3.629,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8097146615844334e-05,
+      "loss": 3.6364,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8088760668333814e-05,
+      "loss": 3.6377,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8080391099627023e-05,
+      "loss": 3.6432,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.80720051521165e-05,
+      "loss": 3.6426,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.806361920460598e-05,
+      "loss": 3.6463,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.805523325709546e-05,
+      "loss": 3.6349,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.804686368838867e-05,
+      "loss": 3.6317,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.803847774087815e-05,
+      "loss": 3.6427,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.803009179336763e-05,
+      "loss": 3.646,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802170584585711e-05,
+      "loss": 3.6298,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.801333627715032e-05,
+      "loss": 3.6366,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.80049503296398e-05,
+      "loss": 3.642,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799656438212929e-05,
+      "loss": 3.6428,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.798817843461877e-05,
+      "loss": 3.6352,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797980886591198e-05,
+      "loss": 3.6297,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797142291840146e-05,
+      "loss": 3.648,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796303697089094e-05,
+      "loss": 3.6461,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7954667402184146e-05,
+      "loss": 3.6434,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7946281454673626e-05,
+      "loss": 3.6297,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7937895507163106e-05,
+      "loss": 3.6373,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7929509559652586e-05,
+      "loss": 3.6315,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7921139990945795e-05,
+      "loss": 3.6446,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7912754043435275e-05,
+      "loss": 3.6442,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7904368095924755e-05,
+      "loss": 3.6365,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789598214841424e-05,
+      "loss": 3.6368,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.788761257970745e-05,
+      "loss": 3.6278,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787922663219693e-05,
+      "loss": 3.6299,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787084068468641e-05,
+      "loss": 3.6275,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786245473717589e-05,
+      "loss": 3.6297,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78540851684691e-05,
+      "loss": 3.6388,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784569922095858e-05,
+      "loss": 3.627,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.783731327344806e-05,
+      "loss": 3.6315,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782892732593754e-05,
+      "loss": 3.6312,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782055775723075e-05,
+      "loss": 3.6326,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781217180972023e-05,
+      "loss": 3.6448,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.780378586220971e-05,
+      "loss": 3.6297,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7795399914699195e-05,
+      "loss": 3.6459,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7787030345992404e-05,
+      "loss": 3.653,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7778644398481884e-05,
+      "loss": 3.6395,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7770258450971364e-05,
+      "loss": 3.635,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7761872503460844e-05,
+      "loss": 3.6325,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7753502934754053e-05,
+      "loss": 3.6333,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7745116987243533e-05,
+      "loss": 3.6417,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.773673103973301e-05,
+      "loss": 3.6367,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.772834509222249e-05,
+      "loss": 3.6274,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.77199755235157e-05,
+      "loss": 3.6439,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771158957600518e-05,
+      "loss": 3.6454,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.770320362849466e-05,
+      "loss": 3.6321,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.769481768098415e-05,
+      "loss": 3.6304,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768644811227736e-05,
+      "loss": 3.6386,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.767806216476684e-05,
+      "loss": 3.6255,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766967621725632e-05,
+      "loss": 3.6479,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76612902697458e-05,
+      "loss": 3.6334,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.765292070103901e-05,
+      "loss": 3.6445,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.764453475352849e-05,
+      "loss": 3.6197,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.763614880601797e-05,
+      "loss": 3.6416,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.762776285850745e-05,
+      "loss": 3.636,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7619393289800656e-05,
+      "loss": 3.6347,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7611007342290136e-05,
+      "loss": 3.6393,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7602621394779616e-05,
+      "loss": 3.6281,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.75942354472691e-05,
+      "loss": 3.6411,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758586587856231e-05,
+      "loss": 3.6576,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757747993105179e-05,
+      "loss": 3.6178,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756909398354127e-05,
+      "loss": 3.6397,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756070803603075e-05,
+      "loss": 3.6316,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755233846732396e-05,
+      "loss": 3.6287,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754395251981344e-05,
+      "loss": 3.6338,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753556657230292e-05,
+      "loss": 3.6367,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.75271806247924e-05,
+      "loss": 3.6369,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751881105608561e-05,
+      "loss": 3.6315,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751042510857509e-05,
+      "loss": 3.6379,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8443191051483154,
+      "eval_runtime": 308.6372,
+      "eval_samples_per_second": 1236.374,
+      "eval_steps_per_second": 38.638,
+      "step": 763200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.750203916106457e-05,
+      "loss": 3.6262,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7493653213554056e-05,
+      "loss": 3.6232,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7485283644847265e-05,
+      "loss": 3.6325,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7476897697336745e-05,
+      "loss": 3.6326,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7468511749826225e-05,
+      "loss": 3.6389,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7460125802315705e-05,
+      "loss": 3.6235,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7451756233608914e-05,
+      "loss": 3.6324,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7443370286098394e-05,
+      "loss": 3.622,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7434984338587874e-05,
+      "loss": 3.6318,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7426598391077354e-05,
+      "loss": 3.6285,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7418228822370563e-05,
+      "loss": 3.624,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740984287486004e-05,
+      "loss": 3.6385,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.740147330615326e-05,
+      "loss": 3.6433,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.739308735864274e-05,
+      "loss": 3.6211,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.738470141113222e-05,
+      "loss": 3.6292,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.73763154636217e-05,
+      "loss": 3.6161,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.736792951611118e-05,
+      "loss": 3.6191,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735954356860066e-05,
+      "loss": 3.6276,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735115762109014e-05,
+      "loss": 3.6214,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.734277167357961e-05,
+      "loss": 3.6314,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.733440210487283e-05,
+      "loss": 3.638,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.732601615736231e-05,
+      "loss": 3.6325,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.731763020985179e-05,
+      "loss": 3.6361,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730924426234126e-05,
+      "loss": 3.6296,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730087469363448e-05,
+      "loss": 3.6239,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729248874612396e-05,
+      "loss": 3.6307,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.728410279861344e-05,
+      "loss": 3.6232,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.727571685110292e-05,
+      "loss": 3.6301,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.726734728239613e-05,
+      "loss": 3.62,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.725896133488561e-05,
+      "loss": 3.6227,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7250575387375086e-05,
+      "loss": 3.6243,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7242189439864566e-05,
+      "loss": 3.6282,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.723381987115778e-05,
+      "loss": 3.6325,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.722543392364726e-05,
+      "loss": 3.6243,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7217047976136735e-05,
+      "loss": 3.6219,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7208662028626215e-05,
+      "loss": 3.6267,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.720029245991943e-05,
+      "loss": 3.631,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719190651240891e-05,
+      "loss": 3.6162,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.718352056489839e-05,
+      "loss": 3.618,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.717513461738787e-05,
+      "loss": 3.6175,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7166765048681086e-05,
+      "loss": 3.6189,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.715837910117056e-05,
+      "loss": 3.6216,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714999315366004e-05,
+      "loss": 3.6295,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714160720614952e-05,
+      "loss": 3.6259,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7133237637442735e-05,
+      "loss": 3.629,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.712485168993221e-05,
+      "loss": 3.6275,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.711646574242169e-05,
+      "loss": 3.626,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7108096173714904e-05,
+      "loss": 3.6202,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7099710226204384e-05,
+      "loss": 3.5956,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7091324278693864e-05,
+      "loss": 3.6369,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7082938331183344e-05,
+      "loss": 3.6176,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.707456876247656e-05,
+      "loss": 3.623,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.706618281496603e-05,
+      "loss": 3.6214,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.705779686745551e-05,
+      "loss": 3.6122,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704941091994499e-05,
+      "loss": 3.6142,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704104135123821e-05,
+      "loss": 3.6111,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703265540372768e-05,
+      "loss": 3.6241,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.702426945621716e-05,
+      "loss": 3.6103,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.701588350870664e-05,
+      "loss": 3.6331,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.700751393999986e-05,
+      "loss": 3.6195,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.699912799248934e-05,
+      "loss": 3.603,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.699074204497882e-05,
+      "loss": 3.6332,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69823560974683e-05,
+      "loss": 3.6094,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.697398652876151e-05,
+      "loss": 3.6076,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.696560058125099e-05,
+      "loss": 3.6246,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.695721463374047e-05,
+      "loss": 3.6288,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694882868622995e-05,
+      "loss": 3.6082,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6940459117523156e-05,
+      "loss": 3.6084,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6932073170012636e-05,
+      "loss": 3.6028,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6923687222502116e-05,
+      "loss": 3.6153,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6915301274991596e-05,
+      "loss": 3.6257,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.690693170628481e-05,
+      "loss": 3.6228,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.689854575877429e-05,
+      "loss": 3.6224,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.689015981126377e-05,
+      "loss": 3.6206,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688177386375325e-05,
+      "loss": 3.6295,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.687340429504646e-05,
+      "loss": 3.619,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.686501834753594e-05,
+      "loss": 3.6269,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.685663240002542e-05,
+      "loss": 3.6093,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68482464525149e-05,
+      "loss": 3.6136,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683987688380811e-05,
+      "loss": 3.6173,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683149093629759e-05,
+      "loss": 3.6213,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.682310498878707e-05,
+      "loss": 3.6246,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.681471904127655e-05,
+      "loss": 3.6265,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6806349472569765e-05,
+      "loss": 3.6133,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6797963525059245e-05,
+      "loss": 3.6084,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6789577577548725e-05,
+      "loss": 3.6192,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6781191630038205e-05,
+      "loss": 3.6229,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6772822061331414e-05,
+      "loss": 3.6081,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6764436113820894e-05,
+      "loss": 3.6141,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6756050166310374e-05,
+      "loss": 3.6268,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6747664218799854e-05,
+      "loss": 3.6186,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.673929465009306e-05,
+      "loss": 3.6137,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.673090870258254e-05,
+      "loss": 3.6085,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672252275507202e-05,
+      "loss": 3.6236,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.67141368075615e-05,
+      "loss": 3.6269,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.670576723885472e-05,
+      "loss": 3.6208,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.66973812913442e-05,
+      "loss": 3.612,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668899534383368e-05,
+      "loss": 3.6139,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668060939632316e-05,
+      "loss": 3.6116,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667223982761637e-05,
+      "loss": 3.6221,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.666385388010585e-05,
+      "loss": 3.6255,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.665546793259533e-05,
+      "loss": 3.6146,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664708198508481e-05,
+      "loss": 3.6127,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663871241637802e-05,
+      "loss": 3.6113,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.66303264688675e-05,
+      "loss": 3.6056,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662194052135698e-05,
+      "loss": 3.609,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.661355457384646e-05,
+      "loss": 3.6085,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.660518500513967e-05,
+      "loss": 3.6124,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.659679905762915e-05,
+      "loss": 3.6095,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.658841311011863e-05,
+      "loss": 3.6077,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.658004354141184e-05,
+      "loss": 3.6114,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.657165759390132e-05,
+      "loss": 3.6093,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.65632716463908e-05,
+      "loss": 3.6239,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.655488569888028e-05,
+      "loss": 3.6071,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.654651613017349e-05,
+      "loss": 3.6278,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.653813018266297e-05,
+      "loss": 3.629,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652974423515245e-05,
+      "loss": 3.6208,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652135828764193e-05,
+      "loss": 3.611,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6513005097738876e-05,
+      "loss": 3.6132,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6504619150228356e-05,
+      "loss": 3.6116,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6496233202717835e-05,
+      "loss": 3.6224,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6487847255207315e-05,
+      "loss": 3.6145,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6479461307696795e-05,
+      "loss": 3.608,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6471075360186275e-05,
+      "loss": 3.6194,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6462689412675755e-05,
+      "loss": 3.6268,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6454303465165235e-05,
+      "loss": 3.6087,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6445933896458444e-05,
+      "loss": 3.6108,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6437547948947924e-05,
+      "loss": 3.6165,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6429162001437404e-05,
+      "loss": 3.602,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6420776053926884e-05,
+      "loss": 3.6297,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641240648522009e-05,
+      "loss": 3.6084,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.640402053770958e-05,
+      "loss": 3.6257,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.639563459019906e-05,
+      "loss": 3.599,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.638724864268854e-05,
+      "loss": 3.6186,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637887907398175e-05,
+      "loss": 3.613,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637049312647123e-05,
+      "loss": 3.6175,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636210717896071e-05,
+      "loss": 3.6148,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.635372123145019e-05,
+      "loss": 3.6062,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63453516627434e-05,
+      "loss": 3.6196,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.633696571523288e-05,
+      "loss": 3.6352,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.632857976772236e-05,
+      "loss": 3.5995,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.632019382021184e-05,
+      "loss": 3.6143,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631182425150505e-05,
+      "loss": 3.6154,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.630343830399453e-05,
+      "loss": 3.6098,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6295052356484014e-05,
+      "loss": 3.6128,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6286666408973494e-05,
+      "loss": 3.6165,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.62782968402667e-05,
+      "loss": 3.6145,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626991089275618e-05,
+      "loss": 3.6135,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626152494524566e-05,
+      "loss": 3.6153,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8424148559570312,
+      "eval_runtime": 308.8693,
+      "eval_samples_per_second": 1235.445,
+      "eval_steps_per_second": 38.609,
+      "step": 839520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.625315537653887e-05,
+      "loss": 3.5951,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.624476942902835e-05,
+      "loss": 3.6022,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.623638348151783e-05,
+      "loss": 3.6148,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.622799753400731e-05,
+      "loss": 3.6092,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621962796530052e-05,
+      "loss": 3.6179,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621124201779e-05,
+      "loss": 3.6036,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.620285607027948e-05,
+      "loss": 3.6097,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.619447012276897e-05,
+      "loss": 3.6069,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6186100554062176e-05,
+      "loss": 3.6109,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6177714606551656e-05,
+      "loss": 3.6028,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6169328659041136e-05,
+      "loss": 3.604,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6160942711530616e-05,
+      "loss": 3.622,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6152589521627555e-05,
+      "loss": 3.6174,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6144203574117034e-05,
+      "loss": 3.6013,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6135817626606514e-05,
+      "loss": 3.6084,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6127431679095994e-05,
+      "loss": 3.5938,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6119045731585474e-05,
+      "loss": 3.5999,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6110659784074954e-05,
+      "loss": 3.6118,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6102273836564434e-05,
+      "loss": 3.5998,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.609388788905392e-05,
+      "loss": 3.6074,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.608551832034713e-05,
+      "loss": 3.6156,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.607713237283661e-05,
+      "loss": 3.6118,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606874642532609e-05,
+      "loss": 3.6183,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606036047781557e-05,
+      "loss": 3.6118,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605199090910878e-05,
+      "loss": 3.6047,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.604360496159826e-05,
+      "loss": 3.6076,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.603521901408774e-05,
+      "loss": 3.6007,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.602683306657722e-05,
+      "loss": 3.6106,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.601846349787043e-05,
+      "loss": 3.5992,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.601007755035991e-05,
+      "loss": 3.6035,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.600169160284939e-05,
+      "loss": 3.6058,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5993305655338875e-05,
+      "loss": 3.6065,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5984936086632084e-05,
+      "loss": 3.6073,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5976550139121564e-05,
+      "loss": 3.6074,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5968164191611044e-05,
+      "loss": 3.6004,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5959778244100524e-05,
+      "loss": 3.6059,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595140867539373e-05,
+      "loss": 3.6125,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.594302272788321e-05,
+      "loss": 3.5941,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.593463678037269e-05,
+      "loss": 3.5968,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.592625083286217e-05,
+      "loss": 3.5967,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.591788126415538e-05,
+      "loss": 3.5948,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590949531664486e-05,
+      "loss": 3.6019,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590110936913434e-05,
+      "loss": 3.6075,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589272342162383e-05,
+      "loss": 3.6092,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.588435385291704e-05,
+      "loss": 3.6076,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.587596790540652e-05,
+      "loss": 3.6081,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5867581957896e-05,
+      "loss": 3.606,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585919601038548e-05,
+      "loss": 3.6028,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5850826441678686e-05,
+      "loss": 3.5774,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5842440494168166e-05,
+      "loss": 3.6152,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5834054546657646e-05,
+      "loss": 3.5977,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5825684977950855e-05,
+      "loss": 3.6043,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5817299030440335e-05,
+      "loss": 3.597,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5808913082929815e-05,
+      "loss": 3.5924,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5800527135419295e-05,
+      "loss": 3.5944,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579215756671251e-05,
+      "loss": 3.5926,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.578377161920199e-05,
+      "loss": 3.605,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.577538567169147e-05,
+      "loss": 3.5884,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.576699972418095e-05,
+      "loss": 3.6138,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.575863015547416e-05,
+      "loss": 3.603,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.575024420796364e-05,
+      "loss": 3.5781,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574185826045312e-05,
+      "loss": 3.612,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.573347231294259e-05,
+      "loss": 3.5898,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.572510274423581e-05,
+      "loss": 3.5904,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.571671679672529e-05,
+      "loss": 3.6004,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.570833084921477e-05,
+      "loss": 3.6086,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569994490170425e-05,
+      "loss": 3.5855,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5691575332997465e-05,
+      "loss": 3.5944,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5683189385486945e-05,
+      "loss": 3.5816,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5674803437976425e-05,
+      "loss": 3.5928,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.56664174904659e-05,
+      "loss": 3.6071,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5658047921759114e-05,
+      "loss": 3.6011,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5649661974248594e-05,
+      "loss": 3.6041,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564127602673807e-05,
+      "loss": 3.5986,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563289007922755e-05,
+      "loss": 3.6102,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.562452051052076e-05,
+      "loss": 3.6015,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.561613456301024e-05,
+      "loss": 3.606,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.560774861549972e-05,
+      "loss": 3.5887,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55993626679892e-05,
+      "loss": 3.5923,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559099309928242e-05,
+      "loss": 3.5966,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55826071517719e-05,
+      "loss": 3.598,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.557422120426137e-05,
+      "loss": 3.6011,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.556583525675085e-05,
+      "loss": 3.6067,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.555746568804407e-05,
+      "loss": 3.5999,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554907974053354e-05,
+      "loss": 3.5858,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554069379302302e-05,
+      "loss": 3.6016,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55323078455125e-05,
+      "loss": 3.6034,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5523938276805716e-05,
+      "loss": 3.5899,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5515552329295196e-05,
+      "loss": 3.5939,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5507166381784676e-05,
+      "loss": 3.604,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5498780434274156e-05,
+      "loss": 3.5981,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.549041086556737e-05,
+      "loss": 3.5955,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5482024918056845e-05,
+      "loss": 3.5871,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5473638970546325e-05,
+      "loss": 3.6054,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5465253023035805e-05,
+      "loss": 3.6036,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5456883454329014e-05,
+      "loss": 3.6064,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5448497506818494e-05,
+      "loss": 3.5834,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5440111559307974e-05,
+      "loss": 3.5998,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5431725611797454e-05,
+      "loss": 3.5921,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.542335604309066e-05,
+      "loss": 3.6011,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.541498647438388e-05,
+      "loss": 3.6071,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.540660052687336e-05,
+      "loss": 3.5927,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5398214579362846e-05,
+      "loss": 3.5933,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538982863185232e-05,
+      "loss": 3.5899,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53814426843418e-05,
+      "loss": 3.5875,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.537305673683128e-05,
+      "loss": 3.5887,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.536467078932076e-05,
+      "loss": 3.5893,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.535628484181024e-05,
+      "loss": 3.5937,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.534791527310345e-05,
+      "loss": 3.591,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.533952932559293e-05,
+      "loss": 3.5902,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.533114337808241e-05,
+      "loss": 3.5857,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532275743057189e-05,
+      "loss": 3.5897,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53143878618651e-05,
+      "loss": 3.6069,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5306001914354584e-05,
+      "loss": 3.5915,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5297615966844064e-05,
+      "loss": 3.6021,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5289230019333544e-05,
+      "loss": 3.605,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528086045062675e-05,
+      "loss": 3.6074,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527247450311623e-05,
+      "loss": 3.5884,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.526408855560571e-05,
+      "loss": 3.5939,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.525570260809519e-05,
+      "loss": 3.5937,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52473330393884e-05,
+      "loss": 3.599,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523894709187788e-05,
+      "loss": 3.5957,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523056114436736e-05,
+      "loss": 3.5898,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.522217519685684e-05,
+      "loss": 3.5942,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.521380562815005e-05,
+      "loss": 3.6054,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.520541968063954e-05,
+      "loss": 3.5971,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.519703373312902e-05,
+      "loss": 3.5869,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.51886477856185e-05,
+      "loss": 3.5948,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5180278216911706e-05,
+      "loss": 3.5833,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5171892269401186e-05,
+      "loss": 3.6063,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5163506321890666e-05,
+      "loss": 3.5907,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5155120374380146e-05,
+      "loss": 3.6044,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5146750805673355e-05,
+      "loss": 3.5803,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5138364858162835e-05,
+      "loss": 3.5999,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5129978910652315e-05,
+      "loss": 3.5935,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5121592963141795e-05,
+      "loss": 3.5962,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5113223394435004e-05,
+      "loss": 3.5972,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.510483744692449e-05,
+      "loss": 3.5865,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.509645149941397e-05,
+      "loss": 3.6034,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.508806555190345e-05,
+      "loss": 3.616,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.507969598319666e-05,
+      "loss": 3.5829,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.507131003568614e-05,
+      "loss": 3.5932,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506292408817562e-05,
+      "loss": 3.5959,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.505455451946883e-05,
+      "loss": 3.5921,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.504616857195831e-05,
+      "loss": 3.5889,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.503778262444779e-05,
+      "loss": 3.5996,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.502939667693727e-05,
+      "loss": 3.5923,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.502101072942675e-05,
+      "loss": 3.5951,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.501262478191623e-05,
+      "loss": 3.5968,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.840970277786255,
+      "eval_runtime": 306.9945,
+      "eval_samples_per_second": 1242.99,
+      "eval_steps_per_second": 38.844,
+      "step": 915840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.500423883440571e-05,
+      "loss": 3.5891,
+      "step": 915968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.499585288689519e-05,
+      "loss": 3.5789,
+      "step": 916480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4987483318188405e-05,
+      "loss": 3.5946,
+      "step": 916992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4979097370677885e-05,
+      "loss": 3.5922,
+      "step": 917504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4970711423167364e-05,
+      "loss": 3.5961,
+      "step": 918016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4962325475656844e-05,
+      "loss": 3.5838,
+      "step": 918528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4953955906950054e-05,
+      "loss": 3.596,
+      "step": 919040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4945569959439533e-05,
+      "loss": 3.5847,
+      "step": 919552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4937184011929013e-05,
+      "loss": 3.5896,
+      "step": 920064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.492879806441849e-05,
+      "loss": 3.5836,
+      "step": 920576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.49204284957117e-05,
+      "loss": 3.5891,
+      "step": 921088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.491204254820118e-05,
+      "loss": 3.5981,
+      "step": 921600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.49036729794944e-05,
+      "loss": 3.5995,
+      "step": 922112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.489528703198388e-05,
+      "loss": 3.5799,
+      "step": 922624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.488690108447336e-05,
+      "loss": 3.59,
+      "step": 923136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.487851513696284e-05,
+      "loss": 3.5723,
+      "step": 923648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.487012918945232e-05,
+      "loss": 3.5865,
+      "step": 924160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.48617432419418e-05,
+      "loss": 3.5876,
+      "step": 924672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.485335729443127e-05,
+      "loss": 3.5799,
+      "step": 925184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.484497134692075e-05,
+      "loss": 3.5899,
+      "step": 925696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.483660177821397e-05,
+      "loss": 3.5969,
+      "step": 926208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.482821583070345e-05,
+      "loss": 3.5899,
+      "step": 926720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.481982988319292e-05,
+      "loss": 3.6012,
+      "step": 927232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.48114439356824e-05,
+      "loss": 3.5945,
+      "step": 927744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4803074366975616e-05,
+      "loss": 3.584,
+      "step": 928256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4794688419465096e-05,
+      "loss": 3.5911,
+      "step": 928768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4786302471954576e-05,
+      "loss": 3.5829,
+      "step": 929280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4777916524444056e-05,
+      "loss": 3.589,
+      "step": 929792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.476954695573727e-05,
+      "loss": 3.5813,
+      "step": 930304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.476117738703048e-05,
+      "loss": 3.5832,
+      "step": 930816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.475279143951996e-05,
+      "loss": 3.5897,
+      "step": 931328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.474440549200944e-05,
+      "loss": 3.5858,
+      "step": 931840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4736019544498914e-05,
+      "loss": 3.5856,
+      "step": 932352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4727633596988394e-05,
+      "loss": 3.5931,
+      "step": 932864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4719247649477874e-05,
+      "loss": 3.5817,
+      "step": 933376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4710861701967354e-05,
+      "loss": 3.5846,
+      "step": 933888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4702475754456834e-05,
+      "loss": 3.5958,
+      "step": 934400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.469410618575005e-05,
+      "loss": 3.5728,
+      "step": 934912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.468572023823953e-05,
+      "loss": 3.5834,
+      "step": 935424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.467733429072901e-05,
+      "loss": 3.5782,
+      "step": 935936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.466894834321849e-05,
+      "loss": 3.5764,
+      "step": 936448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.46605787745117e-05,
+      "loss": 3.5835,
+      "step": 936960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465219282700118e-05,
+      "loss": 3.5876,
+      "step": 937472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.464380687949066e-05,
+      "loss": 3.5903,
+      "step": 937984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.463542093198014e-05,
+      "loss": 3.5896,
+      "step": 938496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.462705136327335e-05,
+      "loss": 3.591,
+      "step": 939008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.461866541576283e-05,
+      "loss": 3.5875,
+      "step": 939520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.461027946825231e-05,
+      "loss": 3.5826,
+      "step": 940032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.460189352074179e-05,
+      "loss": 3.563,
+      "step": 940544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4593523952035e-05,
+      "loss": 3.5901,
+      "step": 941056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.458515438332822e-05,
+      "loss": 3.5788,
+      "step": 941568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.457676843581769e-05,
+      "loss": 3.5842,
+      "step": 942080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.456838248830717e-05,
+      "loss": 3.5803,
+      "step": 942592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.455999654079665e-05,
+      "loss": 3.5741,
+      "step": 943104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.455161059328613e-05,
+      "loss": 3.5777,
+      "step": 943616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.454322464577561e-05,
+      "loss": 3.5731,
+      "step": 944128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.453483869826509e-05,
+      "loss": 3.5866,
+      "step": 944640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.452645275075457e-05,
+      "loss": 3.5646,
+      "step": 945152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.451808318204778e-05,
+      "loss": 3.5971,
+      "step": 945664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.450969723453726e-05,
+      "loss": 3.5845,
+      "step": 946176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.450131128702674e-05,
+      "loss": 3.5626,
+      "step": 946688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.449292533951622e-05,
+      "loss": 3.5911,
+      "step": 947200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.448455577080944e-05,
+      "loss": 3.5705,
+      "step": 947712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.447616982329892e-05,
+      "loss": 3.574,
+      "step": 948224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.44677838757884e-05,
+      "loss": 3.5781,
+      "step": 948736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.445939792827788e-05,
+      "loss": 3.5899,
+      "step": 949248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4451028359571086e-05,
+      "loss": 3.5702,
+      "step": 949760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4442642412060566e-05,
+      "loss": 3.5762,
+      "step": 950272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4434256464550046e-05,
+      "loss": 3.5586,
+      "step": 950784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4425886895843255e-05,
+      "loss": 3.5795,
+      "step": 951296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4417500948332735e-05,
+      "loss": 3.5853,
+      "step": 951808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4409115000822215e-05,
+      "loss": 3.5853,
+      "step": 952320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4400729053311695e-05,
+      "loss": 3.5836,
+      "step": 952832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.439235948460491e-05,
+      "loss": 3.5824,
+      "step": 953344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.438397353709439e-05,
+      "loss": 3.5933,
+      "step": 953856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.437558758958387e-05,
+      "loss": 3.5794,
+      "step": 954368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.436720164207335e-05,
+      "loss": 3.5856,
+      "step": 954880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.435881569456283e-05,
+      "loss": 3.5736,
+      "step": 955392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.435042974705231e-05,
+      "loss": 3.5778,
+      "step": 955904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.434204379954179e-05,
+      "loss": 3.578,
+      "step": 956416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4333674230835e-05,
+      "loss": 3.5773,
+      "step": 956928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.432528828332448e-05,
+      "loss": 3.5874,
+      "step": 957440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.431690233581396e-05,
+      "loss": 3.5845,
+      "step": 957952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.430851638830344e-05,
+      "loss": 3.5866,
+      "step": 958464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.430013044079292e-05,
+      "loss": 3.5662,
+      "step": 958976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.42917444932824e-05,
+      "loss": 3.5809,
+      "step": 959488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.428335854577188e-05,
+      "loss": 3.5839,
+      "step": 960000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.427497259826136e-05,
+      "loss": 3.5746,
+      "step": 960512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4266603029554575e-05,
+      "loss": 3.5737,
+      "step": 961024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4258217082044055e-05,
+      "loss": 3.5838,
+      "step": 961536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424983113453353e-05,
+      "loss": 3.5834,
+      "step": 962048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4241461565826744e-05,
+      "loss": 3.5755,
+      "step": 962560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4233075618316224e-05,
+      "loss": 3.5725,
+      "step": 963072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.42246896708057e-05,
+      "loss": 3.5831,
+      "step": 963584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.421630372329518e-05,
+      "loss": 3.5901,
+      "step": 964096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.420791777578466e-05,
+      "loss": 3.5852,
+      "step": 964608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.419953182827414e-05,
+      "loss": 3.5675,
+      "step": 965120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.419114588076362e-05,
+      "loss": 3.5774,
+      "step": 965632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.41827599332531e-05,
+      "loss": 3.5729,
+      "step": 966144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.417439036454631e-05,
+      "loss": 3.5819,
+      "step": 966656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.416602079583953e-05,
+      "loss": 3.5911,
+      "step": 967168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4157634848329e-05,
+      "loss": 3.5764,
+      "step": 967680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.414924890081848e-05,
+      "loss": 3.5723,
+      "step": 968192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.414086295330796e-05,
+      "loss": 3.575,
+      "step": 968704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.413247700579744e-05,
+      "loss": 3.5707,
+      "step": 969216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.412409105828692e-05,
+      "loss": 3.5675,
+      "step": 969728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.41157051107764e-05,
+      "loss": 3.5752,
+      "step": 970240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.410731916326588e-05,
+      "loss": 3.5741,
+      "step": 970752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.409894959455909e-05,
+      "loss": 3.5758,
+      "step": 971264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.409056364704857e-05,
+      "loss": 3.5682,
+      "step": 971776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.408217769953805e-05,
+      "loss": 3.571,
+      "step": 972288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.407379175202754e-05,
+      "loss": 3.5687,
+      "step": 972800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4065422183320746e-05,
+      "loss": 3.5889,
+      "step": 973312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4057036235810226e-05,
+      "loss": 3.5711,
+      "step": 973824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4048650288299706e-05,
+      "loss": 3.5859,
+      "step": 974336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4040264340789186e-05,
+      "loss": 3.5836,
+      "step": 974848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4031894772082395e-05,
+      "loss": 3.5945,
+      "step": 975360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4023508824571875e-05,
+      "loss": 3.5752,
+      "step": 975872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4015122877061355e-05,
+      "loss": 3.5746,
+      "step": 976384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4006753308354564e-05,
+      "loss": 3.5772,
+      "step": 976896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3998367360844044e-05,
+      "loss": 3.5799,
+      "step": 977408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3989981413333524e-05,
+      "loss": 3.5755,
+      "step": 977920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3981595465823004e-05,
+      "loss": 3.5791,
+      "step": 978432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3973209518312484e-05,
+      "loss": 3.5748,
+      "step": 978944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.396482357080197e-05,
+      "loss": 3.5884,
+      "step": 979456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.395643762329145e-05,
+      "loss": 3.5759,
+      "step": 979968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.394806805458466e-05,
+      "loss": 3.5725,
+      "step": 980480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393968210707414e-05,
+      "loss": 3.5776,
+      "step": 980992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393129615956362e-05,
+      "loss": 3.5717,
+      "step": 981504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.39229102120531e-05,
+      "loss": 3.5865,
+      "step": 982016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.391454064334631e-05,
+      "loss": 3.5728,
+      "step": 982528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.390615469583579e-05,
+      "loss": 3.5882,
+      "step": 983040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.389776874832527e-05,
+      "loss": 3.5677,
+      "step": 983552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.388938280081475e-05,
+      "loss": 3.5797,
+      "step": 984064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.388101323210796e-05,
+      "loss": 3.5752,
+      "step": 984576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.387262728459744e-05,
+      "loss": 3.5775,
+      "step": 985088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3864241337086925e-05,
+      "loss": 3.5732,
+      "step": 985600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3855855389576405e-05,
+      "loss": 3.5757,
+      "step": 986112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3847485820869614e-05,
+      "loss": 3.5843,
+      "step": 986624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3839099873359094e-05,
+      "loss": 3.5941,
+      "step": 987136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3830713925848574e-05,
+      "loss": 3.5684,
+      "step": 987648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3822327978338054e-05,
+      "loss": 3.5717,
+      "step": 988160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.381395840963126e-05,
+      "loss": 3.5784,
+      "step": 988672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.380558884092447e-05,
+      "loss": 3.5736,
+      "step": 989184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.379720289341395e-05,
+      "loss": 3.5705,
+      "step": 989696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.378881694590343e-05,
+      "loss": 3.5849,
+      "step": 990208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.378043099839291e-05,
+      "loss": 3.5726,
+      "step": 990720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.377204505088239e-05,
+      "loss": 3.5824,
+      "step": 991232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.376365910337188e-05,
+      "loss": 3.5729,
+      "step": 991744
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.842820644378662,
+      "eval_runtime": 306.2269,
+      "eval_samples_per_second": 1246.106,
+      "eval_steps_per_second": 38.942,
+      "step": 992160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.375527315586136e-05,
+      "loss": 3.5756,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.374688720835083e-05,
+      "loss": 3.5603,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.373850126084031e-05,
+      "loss": 3.5762,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.373011531332979e-05,
+      "loss": 3.5781,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372174574462301e-05,
+      "loss": 3.5774,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.371335979711248e-05,
+      "loss": 3.5708,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.370497384960196e-05,
+      "loss": 3.573,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.369658790209144e-05,
+      "loss": 3.5707,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3688218333384656e-05,
+      "loss": 3.5689,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367983238587413e-05,
+      "loss": 3.5716,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3671446438363616e-05,
+      "loss": 3.5686,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3663060490853096e-05,
+      "loss": 3.581,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.365470730095004e-05,
+      "loss": 3.5888,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.364632135343952e-05,
+      "loss": 3.5604,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3637935405929e-05,
+      "loss": 3.5671,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.362954945841848e-05,
+      "loss": 3.5597,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3621163510907954e-05,
+      "loss": 3.5678,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3612777563397434e-05,
+      "loss": 3.572,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3604391615886914e-05,
+      "loss": 3.5633,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3596005668376394e-05,
+      "loss": 3.5686,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.35876360996696e-05,
+      "loss": 3.585,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357925015215908e-05,
+      "loss": 3.5685,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357086420464857e-05,
+      "loss": 3.58,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.356247825713805e-05,
+      "loss": 3.5785,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.355410868843126e-05,
+      "loss": 3.5694,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.354572274092074e-05,
+      "loss": 3.5741,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.353733679341022e-05,
+      "loss": 3.5639,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.35289508458997e-05,
+      "loss": 3.5717,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.352058127719291e-05,
+      "loss": 3.5638,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351219532968239e-05,
+      "loss": 3.5635,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.350380938217187e-05,
+      "loss": 3.5751,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.349543981346508e-05,
+      "loss": 3.5664,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.348705386595456e-05,
+      "loss": 3.5734,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347866791844404e-05,
+      "loss": 3.5699,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3470281970933523e-05,
+      "loss": 3.5655,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3461896023423003e-05,
+      "loss": 3.5728,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.345351007591248e-05,
+      "loss": 3.5786,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.344512412840196e-05,
+      "loss": 3.5511,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.343673818089144e-05,
+      "loss": 3.5694,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.342836861218465e-05,
+      "loss": 3.5646,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341998266467413e-05,
+      "loss": 3.5574,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341159671716361e-05,
+      "loss": 3.5643,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.340321076965309e-05,
+      "loss": 3.5762,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.33948412009463e-05,
+      "loss": 3.5727,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.338645525343578e-05,
+      "loss": 3.5721,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.337806930592526e-05,
+      "loss": 3.5723,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336968335841474e-05,
+      "loss": 3.5686,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336131378970796e-05,
+      "loss": 3.5634,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.335292784219744e-05,
+      "loss": 3.547,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.334454189468692e-05,
+      "loss": 3.5735,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3336172325980126e-05,
+      "loss": 3.5655,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3327786378469606e-05,
+      "loss": 3.5693,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3319400430959086e-05,
+      "loss": 3.5601,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3311014483448566e-05,
+      "loss": 3.5576,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3302628535938046e-05,
+      "loss": 3.5573,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3294242588427526e-05,
+      "loss": 3.5538,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3285856640917006e-05,
+      "loss": 3.572,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3277470693406486e-05,
+      "loss": 3.5532,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3269101124699695e-05,
+      "loss": 3.5741,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3260715177189175e-05,
+      "loss": 3.5717,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3252329229678655e-05,
+      "loss": 3.5487,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.324394328216814e-05,
+      "loss": 3.5713,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.323557371346135e-05,
+      "loss": 3.5555,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.322718776595083e-05,
+      "loss": 3.5578,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321880181844031e-05,
+      "loss": 3.5573,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321041587092979e-05,
+      "loss": 3.5753,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3202046302223e-05,
+      "loss": 3.5566,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.319366035471248e-05,
+      "loss": 3.556,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.318527440720196e-05,
+      "loss": 3.5419,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.317688845969143e-05,
+      "loss": 3.5599,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.316851889098465e-05,
+      "loss": 3.5702,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.316013294347413e-05,
+      "loss": 3.5718,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315174699596361e-05,
+      "loss": 3.5645,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3143377427256824e-05,
+      "loss": 3.5677,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3134991479746304e-05,
+      "loss": 3.5743,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3126605532235784e-05,
+      "loss": 3.5597,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3118219584725264e-05,
+      "loss": 3.5666,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310985001601847e-05,
+      "loss": 3.5637,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310146406850795e-05,
+      "loss": 3.5596,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309307812099743e-05,
+      "loss": 3.5601,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.308470855229064e-05,
+      "loss": 3.5618,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.307632260478012e-05,
+      "loss": 3.5711,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30679366572696e-05,
+      "loss": 3.5696,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.305955070975908e-05,
+      "loss": 3.569,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.305116476224856e-05,
+      "loss": 3.5485,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304277881473804e-05,
+      "loss": 3.5666,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.303439286722752e-05,
+      "loss": 3.5651,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3026006919717e-05,
+      "loss": 3.5596,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.301763735101021e-05,
+      "loss": 3.5574,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300925140349969e-05,
+      "loss": 3.5661,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.300086545598917e-05,
+      "loss": 3.5641,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299247950847865e-05,
+      "loss": 3.5591,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.298410993977186e-05,
+      "loss": 3.5556,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.297572399226134e-05,
+      "loss": 3.5689,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.296733804475082e-05,
+      "loss": 3.5688,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.29589520972403e-05,
+      "loss": 3.5694,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.295056614972979e-05,
+      "loss": 3.5479,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2942180202219267e-05,
+      "loss": 3.5631,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2933794254708747e-05,
+      "loss": 3.5596,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2925424686001956e-05,
+      "loss": 3.5648,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2917038738491436e-05,
+      "loss": 3.574,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2908669169784645e-05,
+      "loss": 3.5568,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2900283222274125e-05,
+      "loss": 3.5566,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2891897274763605e-05,
+      "loss": 3.5536,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2883511327253085e-05,
+      "loss": 3.5603,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2875125379742565e-05,
+      "loss": 3.5496,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2866739432232044e-05,
+      "loss": 3.5569,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2858353484721524e-05,
+      "loss": 3.5587,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284998391601474e-05,
+      "loss": 3.5576,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284159796850422e-05,
+      "loss": 3.5514,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.28332120209937e-05,
+      "loss": 3.555,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.282482607348318e-05,
+      "loss": 3.5514,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.281645650477639e-05,
+      "loss": 3.5721,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.280807055726587e-05,
+      "loss": 3.5565,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279968460975535e-05,
+      "loss": 3.5662,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279129866224483e-05,
+      "loss": 3.5685,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.278292909353804e-05,
+      "loss": 3.5788,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.277454314602752e-05,
+      "loss": 3.5589,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2766157198517e-05,
+      "loss": 3.5581,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.275778762981021e-05,
+      "loss": 3.5626,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2749401682299694e-05,
+      "loss": 3.5607,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2741015734789174e-05,
+      "loss": 3.557,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2732629787278654e-05,
+      "loss": 3.5617,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2724243839768134e-05,
+      "loss": 3.5591,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2715857892257614e-05,
+      "loss": 3.5681,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2707471944747094e-05,
+      "loss": 3.5601,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.26991023760403e-05,
+      "loss": 3.5551,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.269071642852978e-05,
+      "loss": 3.5593,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268233048101926e-05,
+      "loss": 3.5565,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.267394453350874e-05,
+      "loss": 3.5658,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.266557496480195e-05,
+      "loss": 3.56,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.265718901729143e-05,
+      "loss": 3.5677,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264880306978091e-05,
+      "loss": 3.5539,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264041712227039e-05,
+      "loss": 3.5588,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263204755356361e-05,
+      "loss": 3.5617,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.262366160605309e-05,
+      "loss": 3.5588,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.261527565854257e-05,
+      "loss": 3.5604,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.260688971103204e-05,
+      "loss": 3.5576,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2598520142325257e-05,
+      "loss": 3.5684,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2590134194814736e-05,
+      "loss": 3.5792,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2581748247304216e-05,
+      "loss": 3.5534,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.257336229979369e-05,
+      "loss": 3.5537,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2565009109890635e-05,
+      "loss": 3.5596,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2556623162380115e-05,
+      "loss": 3.5599,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.25482372148696e-05,
+      "loss": 3.5497,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253985126735908e-05,
+      "loss": 3.5689,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253146531984856e-05,
+      "loss": 3.5579,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.252307937233804e-05,
+      "loss": 3.5654,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2514693424827514e-05,
+      "loss": 3.5536,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.843473434448242,
+      "eval_runtime": 308.1556,
+      "eval_samples_per_second": 1238.306,
+      "eval_steps_per_second": 38.698,
+      "step": 1068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2506307477316994e-05,
+      "loss": 3.5609,
+      "step": 1068544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.249793790861021e-05,
+      "loss": 3.5465,
+      "step": 1069056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.248955196109969e-05,
+      "loss": 3.5594,
+      "step": 1069568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.248116601358916e-05,
+      "loss": 3.5613,
+      "step": 1070080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.247278006607864e-05,
+      "loss": 3.5608,
+      "step": 1070592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.246441049737186e-05,
+      "loss": 3.5524,
+      "step": 1071104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.245602454986134e-05,
+      "loss": 3.5522,
+      "step": 1071616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.244763860235082e-05,
+      "loss": 3.5527,
+      "step": 1072128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.24392526548403e-05,
+      "loss": 3.5533,
+      "step": 1072640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2430883086133515e-05,
+      "loss": 3.5555,
+      "step": 1073152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.242249713862299e-05,
+      "loss": 3.5529,
+      "step": 1073664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.241411119111247e-05,
+      "loss": 3.5633,
+      "step": 1074176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2405741622405684e-05,
+      "loss": 3.5721,
+      "step": 1074688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2397355674895164e-05,
+      "loss": 3.5474,
+      "step": 1075200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.238896972738464e-05,
+      "loss": 3.5535,
+      "step": 1075712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.238058377987412e-05,
+      "loss": 3.5389,
+      "step": 1076224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.23721978323636e-05,
+      "loss": 3.5542,
+      "step": 1076736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.236381188485308e-05,
+      "loss": 3.5494,
+      "step": 1077248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.235542593734256e-05,
+      "loss": 3.5529,
+      "step": 1077760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.234703998983204e-05,
+      "loss": 3.5465,
+      "step": 1078272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.233867042112525e-05,
+      "loss": 3.5697,
+      "step": 1078784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.233028447361473e-05,
+      "loss": 3.5542,
+      "step": 1079296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.232189852610421e-05,
+      "loss": 3.5623,
+      "step": 1079808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.231351257859369e-05,
+      "loss": 3.5594,
+      "step": 1080320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.23051430098869e-05,
+      "loss": 3.5588,
+      "step": 1080832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.229675706237638e-05,
+      "loss": 3.555,
+      "step": 1081344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.228838749366959e-05,
+      "loss": 3.5495,
+      "step": 1081856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.228000154615907e-05,
+      "loss": 3.5545,
+      "step": 1082368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.227161559864855e-05,
+      "loss": 3.5493,
+      "step": 1082880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.226324602994176e-05,
+      "loss": 3.5482,
+      "step": 1083392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.225486008243124e-05,
+      "loss": 3.5561,
+      "step": 1083904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2246474134920726e-05,
+      "loss": 3.5481,
+      "step": 1084416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2238088187410206e-05,
+      "loss": 3.5604,
+      "step": 1084928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2229702239899686e-05,
+      "loss": 3.5518,
+      "step": 1085440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2221316292389166e-05,
+      "loss": 3.5513,
+      "step": 1085952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2212930344878646e-05,
+      "loss": 3.5601,
+      "step": 1086464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2204544397368126e-05,
+      "loss": 3.5554,
+      "step": 1086976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2196174828661335e-05,
+      "loss": 3.5357,
+      "step": 1087488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2187788881150815e-05,
+      "loss": 3.5509,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2179402933640295e-05,
+      "loss": 3.5481,
+      "step": 1088512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2171016986129775e-05,
+      "loss": 3.5389,
+      "step": 1089024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2162647417422984e-05,
+      "loss": 3.5523,
+      "step": 1089536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2154261469912464e-05,
+      "loss": 3.5602,
+      "step": 1090048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.214589190120568e-05,
+      "loss": 3.55,
+      "step": 1090560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.213750595369516e-05,
+      "loss": 3.5573,
+      "step": 1091072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.212912000618464e-05,
+      "loss": 3.5588,
+      "step": 1091584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.212073405867412e-05,
+      "loss": 3.5523,
+      "step": 1092096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211236448996733e-05,
+      "loss": 3.5495,
+      "step": 1092608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.210397854245681e-05,
+      "loss": 3.5328,
+      "step": 1093120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.209559259494629e-05,
+      "loss": 3.5529,
+      "step": 1093632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.208720664743577e-05,
+      "loss": 3.5508,
+      "step": 1094144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.207882069992525e-05,
+      "loss": 3.5517,
+      "step": 1094656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.207043475241473e-05,
+      "loss": 3.5459,
+      "step": 1095168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.206204880490421e-05,
+      "loss": 3.5436,
+      "step": 1095680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.205366285739369e-05,
+      "loss": 3.5376,
+      "step": 1096192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20452932886869e-05,
+      "loss": 3.5403,
+      "step": 1096704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.203690734117638e-05,
+      "loss": 3.5588,
+      "step": 1097216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2028521393665865e-05,
+      "loss": 3.5348,
+      "step": 1097728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2020135446155344e-05,
+      "loss": 3.5562,
+      "step": 1098240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2011765877448554e-05,
+      "loss": 3.5525,
+      "step": 1098752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2003379929938034e-05,
+      "loss": 3.5355,
+      "step": 1099264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.199501036123124e-05,
+      "loss": 3.5507,
+      "step": 1099776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.198662441372072e-05,
+      "loss": 3.5436,
+      "step": 1100288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.19782384662102e-05,
+      "loss": 3.5451,
+      "step": 1100800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.196985251869968e-05,
+      "loss": 3.5372,
+      "step": 1101312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.196146657118916e-05,
+      "loss": 3.5579,
+      "step": 1101824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.195308062367864e-05,
+      "loss": 3.5425,
+      "step": 1102336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1944694676168116e-05,
+      "loss": 3.5357,
+      "step": 1102848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.19363087286576e-05,
+      "loss": 3.5292,
+      "step": 1103360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.192795553875455e-05,
+      "loss": 3.544,
+      "step": 1103872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.191956959124403e-05,
+      "loss": 3.5491,
+      "step": 1104384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.191118364373351e-05,
+      "loss": 3.56,
+      "step": 1104896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.190279769622299e-05,
+      "loss": 3.5456,
+      "step": 1105408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.189441174871247e-05,
+      "loss": 3.5566,
+      "step": 1105920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.188602580120195e-05,
+      "loss": 3.5563,
+      "step": 1106432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.187763985369142e-05,
+      "loss": 3.5469,
+      "step": 1106944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.18692539061809e-05,
+      "loss": 3.5463,
+      "step": 1107456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1860884337474116e-05,
+      "loss": 3.5474,
+      "step": 1107968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185249838996359e-05,
+      "loss": 3.545,
+      "step": 1108480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.184411244245307e-05,
+      "loss": 3.5432,
+      "step": 1108992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1835742873746285e-05,
+      "loss": 3.5457,
+      "step": 1109504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.182735692623577e-05,
+      "loss": 3.5551,
+      "step": 1110016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1818970978725245e-05,
+      "loss": 3.5541,
+      "step": 1110528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1810585031214725e-05,
+      "loss": 3.5499,
+      "step": 1111040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1802199083704205e-05,
+      "loss": 3.5328,
+      "step": 1111552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1793813136193685e-05,
+      "loss": 3.5468,
+      "step": 1112064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1785427188683165e-05,
+      "loss": 3.5488,
+      "step": 1112576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1777041241172645e-05,
+      "loss": 3.5471,
+      "step": 1113088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1768671672465854e-05,
+      "loss": 3.5407,
+      "step": 1113600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1760285724955334e-05,
+      "loss": 3.5472,
+      "step": 1114112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1751899777444814e-05,
+      "loss": 3.5506,
+      "step": 1114624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1743513829934294e-05,
+      "loss": 3.5399,
+      "step": 1115136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.173514426122751e-05,
+      "loss": 3.5431,
+      "step": 1115648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.172675831371699e-05,
+      "loss": 3.5507,
+      "step": 1116160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.171837236620647e-05,
+      "loss": 3.5501,
+      "step": 1116672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170998641869595e-05,
+      "loss": 3.5567,
+      "step": 1117184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170160047118543e-05,
+      "loss": 3.533,
+      "step": 1117696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.169321452367491e-05,
+      "loss": 3.5472,
+      "step": 1118208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.168482857616439e-05,
+      "loss": 3.5478,
+      "step": 1118720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.16764590074576e-05,
+      "loss": 3.5439,
+      "step": 1119232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.166807305994708e-05,
+      "loss": 3.5614,
+      "step": 1119744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.165970349124029e-05,
+      "loss": 3.5432,
+      "step": 1120256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.165131754372977e-05,
+      "loss": 3.5369,
+      "step": 1120768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.164293159621925e-05,
+      "loss": 3.5437,
+      "step": 1121280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.163454564870873e-05,
+      "loss": 3.5442,
+      "step": 1121792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.162615970119821e-05,
+      "loss": 3.5299,
+      "step": 1122304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.161777375368769e-05,
+      "loss": 3.5418,
+      "step": 1122816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1609387806177174e-05,
+      "loss": 3.544,
+      "step": 1123328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.160101823747038e-05,
+      "loss": 3.5395,
+      "step": 1123840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.159263228995986e-05,
+      "loss": 3.5363,
+      "step": 1124352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.158424634244934e-05,
+      "loss": 3.5438,
+      "step": 1124864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.157586039493882e-05,
+      "loss": 3.5344,
+      "step": 1125376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.156749082623203e-05,
+      "loss": 3.5553,
+      "step": 1125888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.155910487872151e-05,
+      "loss": 3.541,
+      "step": 1126400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.155071893121099e-05,
+      "loss": 3.5479,
+      "step": 1126912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154233298370047e-05,
+      "loss": 3.5526,
+      "step": 1127424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.153396341499368e-05,
+      "loss": 3.5653,
+      "step": 1127936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.152557746748316e-05,
+      "loss": 3.541,
+      "step": 1128448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.151719151997264e-05,
+      "loss": 3.5411,
+      "step": 1128960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.150882195126586e-05,
+      "loss": 3.5484,
+      "step": 1129472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.150043600375534e-05,
+      "loss": 3.5398,
+      "step": 1129984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149205005624482e-05,
+      "loss": 3.5439,
+      "step": 1130496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.14836641087343e-05,
+      "loss": 3.5483,
+      "step": 1131008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.147527816122378e-05,
+      "loss": 3.544,
+      "step": 1131520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.146689221371325e-05,
+      "loss": 3.5524,
+      "step": 1132032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.145850626620273e-05,
+      "loss": 3.5449,
+      "step": 1132544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1450136697495946e-05,
+      "loss": 3.5404,
+      "step": 1133056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1441750749985426e-05,
+      "loss": 3.5428,
+      "step": 1133568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.14333648024749e-05,
+      "loss": 3.5475,
+      "step": 1134080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.142497885496438e-05,
+      "loss": 3.542,
+      "step": 1134592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1416609286257595e-05,
+      "loss": 3.546,
+      "step": 1135104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.140822333874708e-05,
+      "loss": 3.5516,
+      "step": 1135616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1399837391236555e-05,
+      "loss": 3.538,
+      "step": 1136128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1391451443726034e-05,
+      "loss": 3.5451,
+      "step": 1136640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.138308187501925e-05,
+      "loss": 3.5467,
+      "step": 1137152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1374695927508724e-05,
+      "loss": 3.5494,
+      "step": 1137664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1366309979998204e-05,
+      "loss": 3.5409,
+      "step": 1138176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1357924032487683e-05,
+      "loss": 3.5448,
+      "step": 1138688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.13495544637809e-05,
+      "loss": 3.5537,
+      "step": 1139200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.134116851627037e-05,
+      "loss": 3.5547,
+      "step": 1139712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133278256875985e-05,
+      "loss": 3.5413,
+      "step": 1140224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.132439662124933e-05,
+      "loss": 3.5429,
+      "step": 1140736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.131602705254255e-05,
+      "loss": 3.5389,
+      "step": 1141248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.130764110503203e-05,
+      "loss": 3.5458,
+      "step": 1141760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.129925515752151e-05,
+      "loss": 3.5322,
+      "step": 1142272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.129086921001099e-05,
+      "loss": 3.5567,
+      "step": 1142784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128248326250047e-05,
+      "loss": 3.5414,
+      "step": 1143296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.127409731498995e-05,
+      "loss": 3.5475,
+      "step": 1143808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.126571136747943e-05,
+      "loss": 3.5392,
+      "step": 1144320
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.842696189880371,
+      "eval_runtime": 312.1121,
+      "eval_samples_per_second": 1222.609,
+      "eval_steps_per_second": 38.207,
+      "step": 1144800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.125732541996891e-05,
+      "loss": 3.5034,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.124893947245839e-05,
+      "loss": 3.5333,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.124055352494787e-05,
+      "loss": 3.545,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123216757743735e-05,
+      "loss": 3.5471,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.122379800873056e-05,
+      "loss": 3.5416,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.121541206122004e-05,
+      "loss": 3.5435,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.120702611370952e-05,
+      "loss": 3.5356,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1198640166199004e-05,
+      "loss": 3.5366,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1190254218688484e-05,
+      "loss": 3.5408,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1181868271177964e-05,
+      "loss": 3.54,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1173482323667443e-05,
+      "loss": 3.537,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.116509637615692e-05,
+      "loss": 3.5483,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.115674318625386e-05,
+      "loss": 3.5548,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.114835723874334e-05,
+      "loss": 3.5338,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.113997129123282e-05,
+      "loss": 3.5372,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11315853437223e-05,
+      "loss": 3.5251,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112319939621178e-05,
+      "loss": 3.5415,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.111481344870126e-05,
+      "loss": 3.5307,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.110642750119074e-05,
+      "loss": 3.5382,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.109804155368022e-05,
+      "loss": 3.5377,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.108967198497344e-05,
+      "loss": 3.5509,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.108128603746292e-05,
+      "loss": 3.5416,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107290008995239e-05,
+      "loss": 3.544,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.106451414244187e-05,
+      "loss": 3.5454,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1056144573735086e-05,
+      "loss": 3.5458,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.104775862622456e-05,
+      "loss": 3.5365,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1039389057517775e-05,
+      "loss": 3.5331,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1031003110007255e-05,
+      "loss": 3.5384,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1022617162496735e-05,
+      "loss": 3.532,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.101423121498621e-05,
+      "loss": 3.5348,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.100584526747569e-05,
+      "loss": 3.5394,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0997459319965175e-05,
+      "loss": 3.5352,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0989073372454655e-05,
+      "loss": 3.544,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0980703803747864e-05,
+      "loss": 3.5369,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0972317856237344e-05,
+      "loss": 3.537,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0963931908726824e-05,
+      "loss": 3.543,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0955545961216304e-05,
+      "loss": 3.5408,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.094717639250951e-05,
+      "loss": 3.5196,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.093879044499899e-05,
+      "loss": 3.5453,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.093040449748847e-05,
+      "loss": 3.5281,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092201854997795e-05,
+      "loss": 3.5216,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.091364898127116e-05,
+      "loss": 3.5387,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.090526303376064e-05,
+      "loss": 3.5484,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.089687708625013e-05,
+      "loss": 3.5344,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.088849113873961e-05,
+      "loss": 3.5372,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.088012157003282e-05,
+      "loss": 3.5456,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.08717356225223e-05,
+      "loss": 3.538,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086336605381551e-05,
+      "loss": 3.5339,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.085498010630499e-05,
+      "loss": 3.5164,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.084659415879447e-05,
+      "loss": 3.5381,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.083820821128395e-05,
+      "loss": 3.5406,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0829822263773427e-05,
+      "loss": 3.535,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0821436316262907e-05,
+      "loss": 3.5294,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0813050368752387e-05,
+      "loss": 3.5264,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0804664421241866e-05,
+      "loss": 3.53,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.079629485253508e-05,
+      "loss": 3.522,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.078790890502456e-05,
+      "loss": 3.5443,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.077952295751404e-05,
+      "loss": 3.5203,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.077113701000352e-05,
+      "loss": 3.539,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076276744129673e-05,
+      "loss": 3.5395,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.075438149378621e-05,
+      "loss": 3.5214,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.074601192507942e-05,
+      "loss": 3.5358,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.07376259775689e-05,
+      "loss": 3.531,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072924003005838e-05,
+      "loss": 3.5274,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072085408254786e-05,
+      "loss": 3.5207,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071246813503734e-05,
+      "loss": 3.5494,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.070408218752682e-05,
+      "loss": 3.5293,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0695712618820036e-05,
+      "loss": 3.5229,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0687326671309516e-05,
+      "loss": 3.5135,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0678940723798996e-05,
+      "loss": 3.528,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0670554776288476e-05,
+      "loss": 3.5346,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0662168828777956e-05,
+      "loss": 3.5433,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0653782881267436e-05,
+      "loss": 3.5306,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0645396933756916e-05,
+      "loss": 3.5401,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0637010986246396e-05,
+      "loss": 3.5418,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0628641417539605e-05,
+      "loss": 3.5325,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0620271848832814e-05,
+      "loss": 3.5308,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0611885901322294e-05,
+      "loss": 3.5323,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0603499953811774e-05,
+      "loss": 3.529,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0595114006301254e-05,
+      "loss": 3.5298,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0586728058790734e-05,
+      "loss": 3.5291,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.057835849008395e-05,
+      "loss": 3.5394,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056997254257343e-05,
+      "loss": 3.5376,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056158659506291e-05,
+      "loss": 3.5349,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055320064755239e-05,
+      "loss": 3.5201,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.05448310788456e-05,
+      "loss": 3.5328,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.053644513133508e-05,
+      "loss": 3.5379,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.052805918382456e-05,
+      "loss": 3.5347,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051967323631404e-05,
+      "loss": 3.5263,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051128728880352e-05,
+      "loss": 3.5317,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0502901341292995e-05,
+      "loss": 3.5389,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0494515393782475e-05,
+      "loss": 3.5207,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.048614582507569e-05,
+      "loss": 3.5303,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0477759877565164e-05,
+      "loss": 3.5347,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0469373930054644e-05,
+      "loss": 3.5353,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0460987982544127e-05,
+      "loss": 3.5407,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0452602035033607e-05,
+      "loss": 3.5264,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0444216087523087e-05,
+      "loss": 3.527,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0435830140012567e-05,
+      "loss": 3.5322,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0427444192502047e-05,
+      "loss": 3.5343,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0419074623795256e-05,
+      "loss": 3.541,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.041070505508847e-05,
+      "loss": 3.5353,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.040231910757795e-05,
+      "loss": 3.521,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.039393316006743e-05,
+      "loss": 3.5259,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.038554721255691e-05,
+      "loss": 3.5361,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.037716126504639e-05,
+      "loss": 3.5075,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.036877531753587e-05,
+      "loss": 3.5275,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.036038937002535e-05,
+      "loss": 3.5329,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.035201980131856e-05,
+      "loss": 3.5225,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.034363385380804e-05,
+      "loss": 3.5201,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.033524790629752e-05,
+      "loss": 3.5323,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0326861958787e-05,
+      "loss": 3.5172,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.031849239008021e-05,
+      "loss": 3.5401,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.031010644256969e-05,
+      "loss": 3.5267,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0301720495059173e-05,
+      "loss": 3.5317,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0293334547548653e-05,
+      "loss": 3.5371,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0284964978841862e-05,
+      "loss": 3.5518,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0276579031331342e-05,
+      "loss": 3.5296,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0268193083820822e-05,
+      "loss": 3.524,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0259823515114035e-05,
+      "loss": 3.5351,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0251437567603515e-05,
+      "loss": 3.5258,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0243051620092994e-05,
+      "loss": 3.5296,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0234665672582474e-05,
+      "loss": 3.5335,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0226279725071954e-05,
+      "loss": 3.5316,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0217893777561434e-05,
+      "loss": 3.5404,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0209507830050914e-05,
+      "loss": 3.5321,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0201121882540394e-05,
+      "loss": 3.5253,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0192752313833607e-05,
+      "loss": 3.5289,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0184366366323087e-05,
+      "loss": 3.5306,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0175980418812567e-05,
+      "loss": 3.5259,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0167594471302047e-05,
+      "loss": 3.5324,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0159224902595256e-05,
+      "loss": 3.5401,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0150838955084736e-05,
+      "loss": 3.5279,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0142453007574216e-05,
+      "loss": 3.5222,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0134099817671157e-05,
+      "loss": 3.5321,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0125713870160637e-05,
+      "loss": 3.5318,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0117327922650117e-05,
+      "loss": 3.5322,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0108941975139597e-05,
+      "loss": 3.5313,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.010055602762908e-05,
+      "loss": 3.5405,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.009217008011856e-05,
+      "loss": 3.5374,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.008378413260804e-05,
+      "loss": 3.5291,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.007539818509752e-05,
+      "loss": 3.5244,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.006702861639073e-05,
+      "loss": 3.5277,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0058659047683942e-05,
+      "loss": 3.5319,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0050273100173422e-05,
+      "loss": 3.5161,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0041887152662902e-05,
+      "loss": 3.5399,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0033501205152382e-05,
+      "loss": 3.533,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0025115257641862e-05,
+      "loss": 3.531,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.001672931013134e-05,
+      "loss": 3.5233,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.000835974142455e-05,
+      "loss": 3.5331,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.843080759048462,
+      "eval_runtime": 304.1752,
+      "eval_samples_per_second": 1254.511,
+      "eval_steps_per_second": 39.204,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.999997379391403e-05,
+      "loss": 3.5168,
+      "step": 1221632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9991587846403514e-05,
+      "loss": 3.5272,
+      "step": 1222144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9983201898892994e-05,
+      "loss": 3.5329,
+      "step": 1222656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9974815951382474e-05,
+      "loss": 3.5308,
+      "step": 1223168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9966430003871947e-05,
+      "loss": 3.5321,
+      "step": 1223680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9958044056361427e-05,
+      "loss": 3.5188,
+      "step": 1224192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9949658108850907e-05,
+      "loss": 3.5235,
+      "step": 1224704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9941288540144123e-05,
+      "loss": 3.5207,
+      "step": 1225216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.99329025926336e-05,
+      "loss": 3.5317,
+      "step": 1225728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.992451664512308e-05,
+      "loss": 3.5176,
+      "step": 1226240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.991613069761256e-05,
+      "loss": 3.5377,
+      "step": 1226752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9907777507709504e-05,
+      "loss": 3.5424,
+      "step": 1227264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9899391560198984e-05,
+      "loss": 3.5181,
+      "step": 1227776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9891005612688468e-05,
+      "loss": 3.5214,
+      "step": 1228288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9882619665177948e-05,
+      "loss": 3.5121,
+      "step": 1228800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.987423371766742e-05,
+      "loss": 3.5254,
+      "step": 1229312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.98658477701569e-05,
+      "loss": 3.5212,
+      "step": 1229824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.985746182264638e-05,
+      "loss": 3.5238,
+      "step": 1230336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.984907587513586e-05,
+      "loss": 3.5232,
+      "step": 1230848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9840706306429073e-05,
+      "loss": 3.5344,
+      "step": 1231360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9832320358918553e-05,
+      "loss": 3.526,
+      "step": 1231872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9823934411408033e-05,
+      "loss": 3.5297,
+      "step": 1232384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9815548463897513e-05,
+      "loss": 3.5321,
+      "step": 1232896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9807162516386993e-05,
+      "loss": 3.5355,
+      "step": 1233408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9798776568876473e-05,
+      "loss": 3.5226,
+      "step": 1233920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9790407000169686e-05,
+      "loss": 3.5201,
+      "step": 1234432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9782021052659165e-05,
+      "loss": 3.5257,
+      "step": 1234944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9773635105148645e-05,
+      "loss": 3.5179,
+      "step": 1235456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9765249157638125e-05,
+      "loss": 3.5199,
+      "step": 1235968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9756863210127605e-05,
+      "loss": 3.5207,
+      "step": 1236480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9748477262617085e-05,
+      "loss": 3.5266,
+      "step": 1236992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9740091315106565e-05,
+      "loss": 3.5288,
+      "step": 1237504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9731721746399778e-05,
+      "loss": 3.5193,
+      "step": 1238016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9723335798889258e-05,
+      "loss": 3.5243,
+      "step": 1238528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9714949851378738e-05,
+      "loss": 3.5317,
+      "step": 1239040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9706563903868218e-05,
+      "loss": 3.5233,
+      "step": 1239552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9698194335161427e-05,
+      "loss": 3.5105,
+      "step": 1240064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.968982476645464e-05,
+      "loss": 3.5282,
+      "step": 1240576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.968143881894412e-05,
+      "loss": 3.5099,
+      "step": 1241088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.96730528714336e-05,
+      "loss": 3.515,
+      "step": 1241600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.966466692392308e-05,
+      "loss": 3.521,
+      "step": 1242112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.965628097641256e-05,
+      "loss": 3.5311,
+      "step": 1242624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.964789502890204e-05,
+      "loss": 3.5236,
+      "step": 1243136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.963950908139152e-05,
+      "loss": 3.5246,
+      "step": 1243648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9631123133881e-05,
+      "loss": 3.5311,
+      "step": 1244160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.962275356517421e-05,
+      "loss": 3.5227,
+      "step": 1244672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.961436761766369e-05,
+      "loss": 3.5207,
+      "step": 1245184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.96059980489569e-05,
+      "loss": 3.504,
+      "step": 1245696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.959761210144638e-05,
+      "loss": 3.5205,
+      "step": 1246208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.958922615393586e-05,
+      "loss": 3.5277,
+      "step": 1246720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9580840206425344e-05,
+      "loss": 3.5189,
+      "step": 1247232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9572454258914824e-05,
+      "loss": 3.513,
+      "step": 1247744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9564068311404304e-05,
+      "loss": 3.5093,
+      "step": 1248256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9555682363893784e-05,
+      "loss": 3.517,
+      "step": 1248768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9547312795186993e-05,
+      "loss": 3.506,
+      "step": 1249280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9538926847676473e-05,
+      "loss": 3.5284,
+      "step": 1249792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9530540900165953e-05,
+      "loss": 3.5104,
+      "step": 1250304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.952215495265543e-05,
+      "loss": 3.5266,
+      "step": 1250816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9513785383948645e-05,
+      "loss": 3.5251,
+      "step": 1251328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9505399436438125e-05,
+      "loss": 3.5064,
+      "step": 1251840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9497029867731334e-05,
+      "loss": 3.5224,
+      "step": 1252352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9488643920220814e-05,
+      "loss": 3.5179,
+      "step": 1252864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9480257972710297e-05,
+      "loss": 3.5115,
+      "step": 1253376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9471872025199777e-05,
+      "loss": 3.5061,
+      "step": 1253888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9463486077689257e-05,
+      "loss": 3.5307,
+      "step": 1254400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.945510013017873e-05,
+      "loss": 3.5163,
+      "step": 1254912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9446730561471946e-05,
+      "loss": 3.5093,
+      "step": 1255424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9438344613961426e-05,
+      "loss": 3.503,
+      "step": 1255936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9429958666450903e-05,
+      "loss": 3.5116,
+      "step": 1256448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9421572718940383e-05,
+      "loss": 3.5198,
+      "step": 1256960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9413186771429863e-05,
+      "loss": 3.529,
+      "step": 1257472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9404800823919343e-05,
+      "loss": 3.5174,
+      "step": 1257984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9396431255212552e-05,
+      "loss": 3.5286,
+      "step": 1258496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9388045307702035e-05,
+      "loss": 3.5262,
+      "step": 1259008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9379659360191515e-05,
+      "loss": 3.5211,
+      "step": 1259520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9371273412680995e-05,
+      "loss": 3.5157,
+      "step": 1260032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9362903843974204e-05,
+      "loss": 3.521,
+      "step": 1260544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9354517896463684e-05,
+      "loss": 3.5183,
+      "step": 1261056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9346131948953164e-05,
+      "loss": 3.5159,
+      "step": 1261568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9337746001442644e-05,
+      "loss": 3.5175,
+      "step": 1262080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9329360053932124e-05,
+      "loss": 3.5213,
+      "step": 1262592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9320990485225336e-05,
+      "loss": 3.5252,
+      "step": 1263104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9312604537714816e-05,
+      "loss": 3.5208,
+      "step": 1263616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9304218590204296e-05,
+      "loss": 3.51,
+      "step": 1264128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9295832642693776e-05,
+      "loss": 3.5146,
+      "step": 1264640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9287446695183256e-05,
+      "loss": 3.5239,
+      "step": 1265152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.927907712647647e-05,
+      "loss": 3.518,
+      "step": 1265664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.927069117896595e-05,
+      "loss": 3.5168,
+      "step": 1266176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.926230523145543e-05,
+      "loss": 3.5151,
+      "step": 1266688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.925391928394491e-05,
+      "loss": 3.5263,
+      "step": 1267200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.924553333643439e-05,
+      "loss": 3.5087,
+      "step": 1267712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9237163767727598e-05,
+      "loss": 3.5147,
+      "step": 1268224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9228777820217078e-05,
+      "loss": 3.5158,
+      "step": 1268736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.922039187270656e-05,
+      "loss": 3.5262,
+      "step": 1269248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.921200592519604e-05,
+      "loss": 3.5224,
+      "step": 1269760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.920361997768552e-05,
+      "loss": 3.5217,
+      "step": 1270272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9195234030175e-05,
+      "loss": 3.5075,
+      "step": 1270784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.918684808266448e-05,
+      "loss": 3.5183,
+      "step": 1271296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.917846213515396e-05,
+      "loss": 3.5195,
+      "step": 1271808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.917009256644717e-05,
+      "loss": 3.5261,
+      "step": 1272320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9161722997740382e-05,
+      "loss": 3.5209,
+      "step": 1272832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9153337050229862e-05,
+      "loss": 3.5098,
+      "step": 1273344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9144951102719342e-05,
+      "loss": 3.5088,
+      "step": 1273856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9136565155208822e-05,
+      "loss": 3.5228,
+      "step": 1274368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9128179207698302e-05,
+      "loss": 3.4955,
+      "step": 1274880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9119793260187782e-05,
+      "loss": 3.5102,
+      "step": 1275392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9111407312677262e-05,
+      "loss": 3.5158,
+      "step": 1275904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.910302136516674e-05,
+      "loss": 3.5149,
+      "step": 1276416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9094651796459954e-05,
+      "loss": 3.5054,
+      "step": 1276928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9086265848949434e-05,
+      "loss": 3.5207,
+      "step": 1277440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9077879901438914e-05,
+      "loss": 3.5018,
+      "step": 1277952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9069510332732123e-05,
+      "loss": 3.5272,
+      "step": 1278464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9061124385221607e-05,
+      "loss": 3.513,
+      "step": 1278976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9052738437711087e-05,
+      "loss": 3.521,
+      "step": 1279488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.904435249020056e-05,
+      "loss": 3.5222,
+      "step": 1280000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9035982921493776e-05,
+      "loss": 3.5409,
+      "step": 1280512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9027596973983256e-05,
+      "loss": 3.5109,
+      "step": 1281024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9019211026472736e-05,
+      "loss": 3.5151,
+      "step": 1281536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9010841457765948e-05,
+      "loss": 3.5211,
+      "step": 1282048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9002455510255428e-05,
+      "loss": 3.5071,
+      "step": 1282560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8994069562744908e-05,
+      "loss": 3.5169,
+      "step": 1283072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8985683615234388e-05,
+      "loss": 3.5192,
+      "step": 1283584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.897729766772386e-05,
+      "loss": 3.5123,
+      "step": 1284096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8968911720213345e-05,
+      "loss": 3.5306,
+      "step": 1284608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8960525772702825e-05,
+      "loss": 3.5156,
+      "step": 1285120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8952139825192305e-05,
+      "loss": 3.5106,
+      "step": 1285632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8943770256485514e-05,
+      "loss": 3.514,
+      "step": 1286144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8935384308974994e-05,
+      "loss": 3.5178,
+      "step": 1286656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8926998361464474e-05,
+      "loss": 3.5114,
+      "step": 1287168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8918612413953954e-05,
+      "loss": 3.5231,
+      "step": 1287680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8910242845247166e-05,
+      "loss": 3.5265,
+      "step": 1288192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8901856897736646e-05,
+      "loss": 3.5127,
+      "step": 1288704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8893470950226126e-05,
+      "loss": 3.5061,
+      "step": 1289216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8885101381519335e-05,
+      "loss": 3.5176,
+      "step": 1289728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8876715434008815e-05,
+      "loss": 3.5163,
+      "step": 1290240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8868329486498298e-05,
+      "loss": 3.52,
+      "step": 1290752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8859943538987778e-05,
+      "loss": 3.5153,
+      "step": 1291264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8851573970280987e-05,
+      "loss": 3.5282,
+      "step": 1291776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8843188022770467e-05,
+      "loss": 3.5226,
+      "step": 1292288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8834802075259947e-05,
+      "loss": 3.5172,
+      "step": 1292800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8826416127749427e-05,
+      "loss": 3.5141,
+      "step": 1293312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8818062937846376e-05,
+      "loss": 3.5147,
+      "step": 1293824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8809676990335856e-05,
+      "loss": 3.5212,
+      "step": 1294336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8801291042825336e-05,
+      "loss": 3.5017,
+      "step": 1294848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.879290509531481e-05,
+      "loss": 3.527,
+      "step": 1295360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.878451914780429e-05,
+      "loss": 3.5232,
+      "step": 1295872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.877613320029377e-05,
+      "loss": 3.5187,
+      "step": 1296384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8767747252783252e-05,
+      "loss": 3.5091,
+      "step": 1296896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.875937768407646e-05,
+      "loss": 3.519,
+      "step": 1297408
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8450539112091064,
+      "eval_runtime": 304.1324,
+      "eval_samples_per_second": 1254.687,
+      "eval_steps_per_second": 39.21,
+      "step": 1297440
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 9.092854166133105e+17,
+  "trial_name": null,
+  "trial_params": null
+}