diff --git "a/checkpoint-1144800/trainer_state.json" "b/checkpoint-1144800/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1144800/trainer_state.json"
@@ -0,0 +1,13555 @@
+{
+  "best_metric": 3.8402984142303467,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/npi-sim-ques/transformer/1/checkpoints/checkpoint-992160",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 1144800,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.9382,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.826,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1933,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9815,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8161,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7147,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6058,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5348,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4544,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4065,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3447,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3244,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.2734,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2075,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1803,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1381,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1198,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0787,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0487,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0242,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0053,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 4.9826,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 4.9623,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9436,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9222,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.9004,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 4.8719,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8583,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8359,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8294,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 4.8119,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7943,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7855,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7653,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.762,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7567,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7372,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 4.7022,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 4.7115,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 4.6834,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 4.6802,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6767,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6705,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6538,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6545,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6442,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6358,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6269,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.6037,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.5997,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.6056,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.59,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.5756,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5619,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.567,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5488,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5649,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 4.5377,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5431,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5365,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5124,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5203,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5076,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4985,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4938,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5108,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.4864,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4767,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4631,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.4641,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4802,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4789,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.452,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4699,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379669187473754e-05,
+      "loss": 4.4646,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 4.4497,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 4.4352,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 4.4412,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 4.4291,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 4.4237,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 4.4241,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 4.4196,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.4237,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 4.4206,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.4004,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928747290126923e-05,
+      "loss": 4.4059,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927908695375871e-05,
+      "loss": 4.406,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927070100624819e-05,
+      "loss": 4.3959,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926231505873767e-05,
+      "loss": 4.3922,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925392911122715e-05,
+      "loss": 4.3894,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 4.3909,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 4.3782,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 4.3787,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 4.3736,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.3766,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.3702,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.3642,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918687428875045e-05,
+      "loss": 4.3493,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917848834123993e-05,
+      "loss": 4.3549,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917010239372941e-05,
+      "loss": 4.3581,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.363,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91533468775121e-05,
+      "loss": 4.3484,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914497730880531e-05,
+      "loss": 4.3387,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913659136129479e-05,
+      "loss": 4.3349,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912820541378427e-05,
+      "loss": 4.342,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911981946627375e-05,
+      "loss": 4.313,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111449897566966e-05,
+      "loss": 4.3225,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103063950056446e-05,
+      "loss": 4.3247,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094678002545926e-05,
+      "loss": 4.3234,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086292055035406e-05,
+      "loss": 4.315,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9077906107524886e-05,
+      "loss": 4.3224,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9069520160014366e-05,
+      "loss": 4.304,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9061134212503846e-05,
+      "loss": 4.3314,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052764643797055e-05,
+      "loss": 4.3106,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9044378696286535e-05,
+      "loss": 4.3122,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9035992748776015e-05,
+      "loss": 4.3161,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027606801265495e-05,
+      "loss": 4.3239,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901925361136244e-05,
+      "loss": 4.3014,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901086766385192e-05,
+      "loss": 4.2956,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024817163414e-05,
+      "loss": 4.3012,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899409576883088e-05,
+      "loss": 4.2905,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898570982132036e-05,
+      "loss": 4.296,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897732387380984e-05,
+      "loss": 4.2919,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896893792629932e-05,
+      "loss": 4.2813,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89605519787888e-05,
+      "loss": 4.2991,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895216603127827e-05,
+      "loss": 4.2806,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894378008376775e-05,
+      "loss": 4.2753,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893539413625723e-05,
+      "loss": 4.2735,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892702456755045e-05,
+      "loss": 4.2789,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891863862003992e-05,
+      "loss": 4.2657,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891025267252941e-05,
+      "loss": 4.2818,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890186672501889e-05,
+      "loss": 4.2754,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889348077750837e-05,
+      "loss": 4.2619,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888511120880158e-05,
+      "loss": 4.2563,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887672526129106e-05,
+      "loss": 4.2654,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886833931378054e-05,
+      "loss": 4.2615,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885995336627002e-05,
+      "loss": 4.2658,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515674187595e-05,
+      "loss": 4.2614,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2621,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.2574,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8826425955031666e-05,
+      "loss": 4.2517,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8818040007521146e-05,
+      "loss": 4.2434,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8809654060010626e-05,
+      "loss": 4.2421,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8801268112500106e-05,
+      "loss": 4.2489,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879289854379332e-05,
+      "loss": 4.2279,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87845125962828e-05,
+      "loss": 4.2491,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877612664877228e-05,
+      "loss": 4.2438,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876774070126176e-05,
+      "loss": 4.235,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875937113255497e-05,
+      "loss": 4.2363,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875098518504445e-05,
+      "loss": 4.2375,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.197122573852539,
+      "eval_runtime": 304.6949,
+      "eval_samples_per_second": 1252.371,
+      "eval_steps_per_second": 39.138,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874259923753393e-05,
+      "loss": 4.2118,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873421329002341e-05,
+      "loss": 4.2212,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872582734251289e-05,
+      "loss": 4.2319,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871744139500237e-05,
+      "loss": 4.2169,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870907182629558e-05,
+      "loss": 4.2223,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870068587878506e-05,
+      "loss": 4.2114,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869229993127454e-05,
+      "loss": 4.2099,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683913983764027e-05,
+      "loss": 4.2038,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675544415057236e-05,
+      "loss": 4.2136,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667158467546716e-05,
+      "loss": 4.2008,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658772520036196e-05,
+      "loss": 4.2208,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650386572525676e-05,
+      "loss": 4.2138,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642017003818885e-05,
+      "loss": 4.1972,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633631056308365e-05,
+      "loss": 4.1927,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8625245108797845e-05,
+      "loss": 4.1911,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8616859161287324e-05,
+      "loss": 4.1901,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608473213776804e-05,
+      "loss": 4.1857,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860008726626628e-05,
+      "loss": 4.1877,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591701318755764e-05,
+      "loss": 4.1811,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583315371245244e-05,
+      "loss": 4.1961,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857494580253846e-05,
+      "loss": 4.187,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856655985502793e-05,
+      "loss": 4.1907,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855817390751741e-05,
+      "loss": 4.1848,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854978796000689e-05,
+      "loss": 4.193,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854141839130011e-05,
+      "loss": 4.1705,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853303244378958e-05,
+      "loss": 4.1724,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852464649627906e-05,
+      "loss": 4.1686,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851626054876854e-05,
+      "loss": 4.1664,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850787460125802e-05,
+      "loss": 4.1681,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849950503255123e-05,
+      "loss": 4.1642,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849111908504072e-05,
+      "loss": 4.1682,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84827331375302e-05,
+      "loss": 4.1683,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847434719001968e-05,
+      "loss": 4.1558,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846596124250916e-05,
+      "loss": 4.1638,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845759167380237e-05,
+      "loss": 4.1717,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844920572629185e-05,
+      "loss": 4.1582,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844081977878133e-05,
+      "loss": 4.142,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843243383127081e-05,
+      "loss": 4.1544,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842404788376029e-05,
+      "loss": 4.1361,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415678315053496e-05,
+      "loss": 4.1461,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8407292367542976e-05,
+      "loss": 4.1464,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398906420032456e-05,
+      "loss": 4.1525,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8390520472521936e-05,
+      "loss": 4.141,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838215090381515e-05,
+      "loss": 4.1467,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837376495630463e-05,
+      "loss": 4.1494,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836537900879411e-05,
+      "loss": 4.1421,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835699306128359e-05,
+      "loss": 4.1426,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83486234925768e-05,
+      "loss": 4.1252,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834023754506628e-05,
+      "loss": 4.1329,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833185159755576e-05,
+      "loss": 4.1394,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832346565004524e-05,
+      "loss": 4.1323,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831509608133845e-05,
+      "loss": 4.1257,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830671013382793e-05,
+      "loss": 4.1142,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829832418631741e-05,
+      "loss": 4.1276,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828993823880689e-05,
+      "loss": 4.1141,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281568670100105e-05,
+      "loss": 4.1299,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273182722589585e-05,
+      "loss": 4.1168,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264796775079065e-05,
+      "loss": 4.1247,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256410827568545e-05,
+      "loss": 4.1261,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248024880058025e-05,
+      "loss": 4.1024,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239655311351234e-05,
+      "loss": 4.1144,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231269363840714e-05,
+      "loss": 4.1134,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222883416330194e-05,
+      "loss": 4.1049,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214497468819674e-05,
+      "loss": 4.105,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820612790011288e-05,
+      "loss": 4.1212,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819774195260236e-05,
+      "loss": 4.1081,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818935600509184e-05,
+      "loss": 4.0982,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818097005758132e-05,
+      "loss": 4.0898,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817260048887454e-05,
+      "loss": 4.0973,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816421454136402e-05,
+      "loss": 4.1121,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81558285938535e-05,
+      "loss": 4.1155,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814744264634298e-05,
+      "loss": 4.0991,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813907307763619e-05,
+      "loss": 4.1138,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813068713012567e-05,
+      "loss": 4.1089,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812230118261515e-05,
+      "loss": 4.1027,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811391523510463e-05,
+      "loss": 4.0885,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810554566639784e-05,
+      "loss": 4.1034,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809715971888732e-05,
+      "loss": 4.0917,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80887737713768e-05,
+      "loss": 4.0896,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808038782386628e-05,
+      "loss": 4.0942,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807201825515949e-05,
+      "loss": 4.0887,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806363230764897e-05,
+      "loss": 4.1018,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805524636013845e-05,
+      "loss": 4.0919,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804686041262793e-05,
+      "loss": 4.0834,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803849084392114e-05,
+      "loss": 4.0878,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803010489641062e-05,
+      "loss": 4.09,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80217189489001e-05,
+      "loss": 4.0854,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801333300138958e-05,
+      "loss": 4.0848,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800496343268279e-05,
+      "loss": 4.0746,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799657748517227e-05,
+      "loss": 4.0909,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798819153766175e-05,
+      "loss": 4.0738,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797980559015123e-05,
+      "loss": 4.0786,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971436021444446e-05,
+      "loss": 4.0713,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7963050073933926e-05,
+      "loss": 4.0855,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954664126423406e-05,
+      "loss": 4.0821,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946278178912886e-05,
+      "loss": 4.0757,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937908610206095e-05,
+      "loss": 4.0631,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929522662695575e-05,
+      "loss": 4.071,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7921136715185055e-05,
+      "loss": 4.073,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912750767674535e-05,
+      "loss": 4.0808,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7904381198967744e-05,
+      "loss": 4.0726,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895995251457224e-05,
+      "loss": 4.0636,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887609303946704e-05,
+      "loss": 4.0581,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787923973523992e-05,
+      "loss": 4.0709,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78708537877294e-05,
+      "loss": 4.0419,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786248421902261e-05,
+      "loss": 4.0572,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785409827151209e-05,
+      "loss": 4.0599,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784571232400157e-05,
+      "loss": 4.0589,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783732637649105e-05,
+      "loss": 4.0488,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782894042898053e-05,
+      "loss": 4.0587,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782055448147001e-05,
+      "loss": 4.0508,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781216853395948e-05,
+      "loss": 4.0702,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780378258644896e-05,
+      "loss": 4.0558,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779541301774218e-05,
+      "loss": 4.0523,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778702707023166e-05,
+      "loss": 4.0688,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777864112272114e-05,
+      "loss": 4.0718,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770271554014354e-05,
+      "loss": 4.0525,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761885606503834e-05,
+      "loss": 4.0477,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775349965899331e-05,
+      "loss": 4.0572,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774511371148279e-05,
+      "loss": 4.0463,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773672776397227e-05,
+      "loss": 4.0485,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7728341816461747e-05,
+      "loss": 4.0545,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719955868951227e-05,
+      "loss": 4.0415,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7711586300244436e-05,
+      "loss": 4.0611,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703200352733916e-05,
+      "loss": 4.0431,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694814405223396e-05,
+      "loss": 4.0438,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686428457712876e-05,
+      "loss": 4.0374,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767805888900609e-05,
+      "loss": 4.0455,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766967294149557e-05,
+      "loss": 4.0374,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766128699398505e-05,
+      "loss": 4.052,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765290104647453e-05,
+      "loss": 4.0492,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764453147776774e-05,
+      "loss": 4.0394,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763614553025722e-05,
+      "loss": 4.0246,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76277595827467e-05,
+      "loss": 4.0464,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761937363523618e-05,
+      "loss": 4.0367,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761100406652939e-05,
+      "loss": 4.0448,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760261811901887e-05,
+      "loss": 4.0398,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759423217150835e-05,
+      "loss": 4.0421,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758584622399783e-05,
+      "loss": 4.0397,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577476655291045e-05,
+      "loss": 4.0424,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569090707780525e-05,
+      "loss": 4.0271,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560704760270005e-05,
+      "loss": 4.0254,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552318812759485e-05,
+      "loss": 4.038,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543949244052694e-05,
+      "loss": 4.0205,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75355796753459e-05,
+      "loss": 4.0363,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752719372783538e-05,
+      "loss": 4.0414,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751880778032486e-05,
+      "loss": 4.0256,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751042183281434e-05,
+      "loss": 4.025,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750203588530382e-05,
+      "loss": 4.0282,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.028327941894531,
+      "eval_runtime": 305.0036,
+      "eval_samples_per_second": 1251.103,
+      "eval_steps_per_second": 39.098,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74936499377933e-05,
+      "loss": 4.0102,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748526399028278e-05,
+      "loss": 4.019,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747687804277226e-05,
+      "loss": 4.0284,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746849209526175e-05,
+      "loss": 4.0216,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746010614775123e-05,
+      "loss": 4.0264,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745172020024071e-05,
+      "loss": 4.0081,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744333425273019e-05,
+      "loss": 4.0217,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743494830521967e-05,
+      "loss": 4.0055,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742656235770914e-05,
+      "loss": 4.0191,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741817641019862e-05,
+      "loss": 4.0134,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740980684149184e-05,
+      "loss": 4.0227,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740142089398132e-05,
+      "loss": 4.0237,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739303494647079e-05,
+      "loss": 4.0051,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738464899896027e-05,
+      "loss": 4.0063,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737626305144975e-05,
+      "loss": 4.0067,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736787710393923e-05,
+      "loss": 4.0025,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735949115642871e-05,
+      "loss": 3.9991,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73511052089182e-05,
+      "loss": 4.0029,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734273564021141e-05,
+      "loss": 3.9979,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733434969270089e-05,
+      "loss": 4.0207,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732596374519037e-05,
+      "loss": 4.0019,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731757779767985e-05,
+      "loss": 4.0144,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7309208228973056e-05,
+      "loss": 4.0086,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300838660266265e-05,
+      "loss": 4.015,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292452712755745e-05,
+      "loss": 3.9878,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7284066765245225e-05,
+      "loss": 3.999,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275680817734705e-05,
+      "loss": 3.9896,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267294870224185e-05,
+      "loss": 3.9986,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258908922713665e-05,
+      "loss": 3.9903,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725052297520315e-05,
+      "loss": 3.992,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724215340649636e-05,
+      "loss": 4.0023,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723376745898584e-05,
+      "loss": 3.999,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722538151147532e-05,
+      "loss": 3.9854,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72169955639648e-05,
+      "loss": 3.9989,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720860961645428e-05,
+      "loss": 3.996,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720022366894376e-05,
+      "loss": 3.993,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719183772143324e-05,
+      "loss": 3.9803,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718345177392272e-05,
+      "loss": 3.9898,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717508220521593e-05,
+      "loss": 3.9716,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716669625770541e-05,
+      "loss": 3.9871,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715831031019489e-05,
+      "loss": 3.9811,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714992436268437e-05,
+      "loss": 3.9908,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141554793977585e-05,
+      "loss": 3.9843,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133168846467065e-05,
+      "loss": 3.9859,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124782898956545e-05,
+      "loss": 3.9899,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116396951446025e-05,
+      "loss": 3.9867,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7108027382739234e-05,
+      "loss": 3.985,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099641435228714e-05,
+      "loss": 3.9679,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091255487718194e-05,
+      "loss": 3.9743,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082869540207674e-05,
+      "loss": 3.985,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707449997150088e-05,
+      "loss": 3.9777,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706611402399036e-05,
+      "loss": 3.9702,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705772807647984e-05,
+      "loss": 3.9651,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704934212896932e-05,
+      "loss": 3.9753,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704097256026254e-05,
+      "loss": 3.9587,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703260299155575e-05,
+      "loss": 3.9781,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702421704404523e-05,
+      "loss": 3.9658,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701583109653471e-05,
+      "loss": 3.9777,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700744514902419e-05,
+      "loss": 3.9745,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699905920151367e-05,
+      "loss": 3.9582,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699067325400315e-05,
+      "loss": 3.964,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698228730649263e-05,
+      "loss": 3.9668,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69739013589821e-05,
+      "loss": 3.9572,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696553179027532e-05,
+      "loss": 3.9647,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69571458427648e-05,
+      "loss": 3.9712,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694877627405801e-05,
+      "loss": 3.9668,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694039032654749e-05,
+      "loss": 3.9552,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693200437903697e-05,
+      "loss": 3.946,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692361843152645e-05,
+      "loss": 3.9527,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915232484015926e-05,
+      "loss": 3.9677,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906846536505406e-05,
+      "loss": 3.9744,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898460588994886e-05,
+      "loss": 3.9579,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6890091020288095e-05,
+      "loss": 3.9718,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881705072777575e-05,
+      "loss": 3.9693,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873319125267055e-05,
+      "loss": 3.9678,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864933177756535e-05,
+      "loss": 3.9445,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856547230246015e-05,
+      "loss": 3.9684,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684817766153923e-05,
+      "loss": 3.958,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683979171402871e-05,
+      "loss": 3.9518,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683140576651819e-05,
+      "loss": 3.9595,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682301981900767e-05,
+      "loss": 3.9542,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681463387149715e-05,
+      "loss": 3.9664,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680624792398663e-05,
+      "loss": 3.9585,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679786197647611e-05,
+      "loss": 3.947,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678949240776932e-05,
+      "loss": 3.954,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67811064602588e-05,
+      "loss": 3.9552,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677272051274828e-05,
+      "loss": 3.9531,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676433456523776e-05,
+      "loss": 3.9509,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675594861772724e-05,
+      "loss": 3.9442,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674756267021672e-05,
+      "loss": 3.9564,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739193101509935e-05,
+      "loss": 3.9446,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730807153999415e-05,
+      "loss": 3.9481,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722421206488895e-05,
+      "loss": 3.9417,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6714035258978375e-05,
+      "loss": 3.9578,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705649311467855e-05,
+      "loss": 3.9523,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697263363957335e-05,
+      "loss": 3.9464,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688877416446815e-05,
+      "loss": 3.9388,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680507847740024e-05,
+      "loss": 3.9469,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667213827903323e-05,
+      "loss": 3.9382,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666375233152271e-05,
+      "loss": 3.9592,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665536638401219e-05,
+      "loss": 3.9433,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664698043650167e-05,
+      "loss": 3.9398,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663859448899115e-05,
+      "loss": 3.934,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663020854148063e-05,
+      "loss": 3.9448,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662182259397011e-05,
+      "loss": 3.9211,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661345302526333e-05,
+      "loss": 3.9361,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660506707775281e-05,
+      "loss": 3.9378,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659668113024229e-05,
+      "loss": 3.9368,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658829518273176e-05,
+      "loss": 3.9243,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657992561402498e-05,
+      "loss": 3.9412,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657153966651446e-05,
+      "loss": 3.9262,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656315371900393e-05,
+      "loss": 3.9509,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655476777149341e-05,
+      "loss": 3.9355,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654638182398289e-05,
+      "loss": 3.9303,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653799587647237e-05,
+      "loss": 3.9481,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652960992896185e-05,
+      "loss": 3.9551,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6521240360255066e-05,
+      "loss": 3.9344,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512854412744546e-05,
+      "loss": 3.9279,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6504468465234026e-05,
+      "loss": 3.9368,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6496082517723506e-05,
+      "loss": 3.93,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487696570212986e-05,
+      "loss": 3.9291,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479327001506195e-05,
+      "loss": 3.9421,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470941053995675e-05,
+      "loss": 3.9212,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462555106485155e-05,
+      "loss": 3.944,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454169158974635e-05,
+      "loss": 3.9288,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445783211464115e-05,
+      "loss": 3.9278,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437413642757324e-05,
+      "loss": 3.9212,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429027695246804e-05,
+      "loss": 3.9326,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642064174773629e-05,
+      "loss": 3.9196,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641225580022577e-05,
+      "loss": 3.9409,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640386985271525e-05,
+      "loss": 3.9312,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639548390520473e-05,
+      "loss": 3.9284,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638709795769421e-05,
+      "loss": 3.9111,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637871201018369e-05,
+      "loss": 3.9347,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63703424414769e-05,
+      "loss": 3.928,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636195649396638e-05,
+      "loss": 3.9274,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635357054645586e-05,
+      "loss": 3.924,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634520097774907e-05,
+      "loss": 3.9333,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633681503023855e-05,
+      "loss": 3.9289,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632842908272803e-05,
+      "loss": 3.9343,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632004313521751e-05,
+      "loss": 3.9163,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311673566510724e-05,
+      "loss": 3.9168,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303287619000204e-05,
+      "loss": 3.9261,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294901671489684e-05,
+      "loss": 3.9099,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286515723979164e-05,
+      "loss": 3.9263,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6278129776468644e-05,
+      "loss": 3.9329,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626976020776185e-05,
+      "loss": 3.9182,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626137426025133e-05,
+      "loss": 3.915,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625298831274081e-05,
+      "loss": 3.9236,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.952652931213379,
+      "eval_runtime": 308.9364,
+      "eval_samples_per_second": 1235.177,
+      "eval_steps_per_second": 38.6,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624460236523029e-05,
+      "loss": 3.9113,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.62362327965235e-05,
+      "loss": 3.9117,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622784684901298e-05,
+      "loss": 3.9211,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621946090150246e-05,
+      "loss": 3.9106,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621107495399194e-05,
+      "loss": 3.925,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620268900648142e-05,
+      "loss": 3.9044,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619431943777464e-05,
+      "loss": 3.9183,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618593349026412e-05,
+      "loss": 3.9022,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617754754275359e-05,
+      "loss": 3.9146,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616916159524307e-05,
+      "loss": 3.9073,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616079202653629e-05,
+      "loss": 3.919,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615240607902577e-05,
+      "loss": 3.9187,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614402013151524e-05,
+      "loss": 3.9016,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613563418400472e-05,
+      "loss": 3.906,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61272482364942e-05,
+      "loss": 3.9028,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6118878667787416e-05,
+      "loss": 3.9031,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110492720276896e-05,
+      "loss": 3.8973,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6102106772766376e-05,
+      "loss": 3.9023,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093720825255856e-05,
+      "loss": 3.8976,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085334877745336e-05,
+      "loss": 3.9238,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076965309038545e-05,
+      "loss": 3.902,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068579361528025e-05,
+      "loss": 3.9091,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6060193414017505e-05,
+      "loss": 3.9123,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051807466506985e-05,
+      "loss": 3.9123,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043437897800194e-05,
+      "loss": 3.8905,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6035051950289674e-05,
+      "loss": 3.9011,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026666002779154e-05,
+      "loss": 3.8935,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018280055268634e-05,
+      "loss": 3.8994,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009894107758114e-05,
+      "loss": 3.8909,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600152453905133e-05,
+      "loss": 3.8934,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599313859154081e-05,
+      "loss": 3.9036,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598476902283402e-05,
+      "loss": 3.9022,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59763830753235e-05,
+      "loss": 3.8954,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596799712781298e-05,
+      "loss": 3.9009,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595961118030246e-05,
+      "loss": 3.9015,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595122523279194e-05,
+      "loss": 3.8971,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594283928528142e-05,
+      "loss": 3.8842,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59344533377709e-05,
+      "loss": 3.8968,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592606739026038e-05,
+      "loss": 3.8743,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591769782155359e-05,
+      "loss": 3.8926,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590931187404307e-05,
+      "loss": 3.8869,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5900925926532554e-05,
+      "loss": 3.8998,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5892539979022034e-05,
+      "loss": 3.8959,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5884154031511514e-05,
+      "loss": 3.8889,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587578446280472e-05,
+      "loss": 3.8975,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58673985152942e-05,
+      "loss": 3.8967,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585901256778368e-05,
+      "loss": 3.8926,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585062662027316e-05,
+      "loss": 3.878,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584224067276264e-05,
+      "loss": 3.8812,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583387110405585e-05,
+      "loss": 3.8961,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582548515654533e-05,
+      "loss": 3.8818,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581709920903481e-05,
+      "loss": 3.8781,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580871326152429e-05,
+      "loss": 3.8758,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580034369281751e-05,
+      "loss": 3.888,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579195774530699e-05,
+      "loss": 3.8707,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578357179779647e-05,
+      "loss": 3.8856,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577518585028595e-05,
+      "loss": 3.8792,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576679990277543e-05,
+      "loss": 3.8846,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758430334068637e-05,
+      "loss": 3.8858,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5750044386558117e-05,
+      "loss": 3.8727,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741658439047596e-05,
+      "loss": 3.8717,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733272491537076e-05,
+      "loss": 3.884,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572488654402655e-05,
+      "loss": 3.8676,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716516975319765e-05,
+      "loss": 3.8761,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5708131027809245e-05,
+      "loss": 3.8834,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699745080298725e-05,
+      "loss": 3.8828,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691359132788205e-05,
+      "loss": 3.8656,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5682973185277685e-05,
+      "loss": 3.8616,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674587237767165e-05,
+      "loss": 3.8638,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5666201290256645e-05,
+      "loss": 3.8774,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657815342746125e-05,
+      "loss": 3.8872,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649445774039334e-05,
+      "loss": 3.8725,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5641059826528814e-05,
+      "loss": 3.8856,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632673879018294e-05,
+      "loss": 3.8862,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624287931507774e-05,
+      "loss": 3.8812,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615901983997254e-05,
+      "loss": 3.8628,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560753241529046e-05,
+      "loss": 3.8807,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559914646777994e-05,
+      "loss": 3.8747,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559076052026943e-05,
+      "loss": 3.8673,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558237457275891e-05,
+      "loss": 3.8735,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557398862524839e-05,
+      "loss": 3.8745,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556560267773787e-05,
+      "loss": 3.8784,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555723310903108e-05,
+      "loss": 3.8748,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554884716152056e-05,
+      "loss": 3.8664,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554046121401004e-05,
+      "loss": 3.872,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553207526649952e-05,
+      "loss": 3.8691,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523689318989e-05,
+      "loss": 3.8706,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551531975028221e-05,
+      "loss": 3.8685,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550693380277169e-05,
+      "loss": 3.8609,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549854785526117e-05,
+      "loss": 3.8737,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549016190775065e-05,
+      "loss": 3.868,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548177596024013e-05,
+      "loss": 3.8622,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5473406391533343e-05,
+      "loss": 3.8609,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5465020444022823e-05,
+      "loss": 3.8783,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54566344965123e-05,
+      "loss": 3.8709,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544824854900178e-05,
+      "loss": 3.8645,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543987898029499e-05,
+      "loss": 3.8555,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543149303278447e-05,
+      "loss": 3.8657,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542310708527395e-05,
+      "loss": 3.857,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541472113776343e-05,
+      "loss": 3.8788,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540635156905664e-05,
+      "loss": 3.8641,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539796562154612e-05,
+      "loss": 3.8605,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53895796740356e-05,
+      "loss": 3.8565,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538119372652508e-05,
+      "loss": 3.8658,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53728241578183e-05,
+      "loss": 3.8386,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536443821030778e-05,
+      "loss": 3.8575,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535605226279726e-05,
+      "loss": 3.8575,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534766631528674e-05,
+      "loss": 3.859,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339296746579946e-05,
+      "loss": 3.8455,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5330910799069426e-05,
+      "loss": 3.863,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5322524851558906e-05,
+      "loss": 3.8478,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314138904048386e-05,
+      "loss": 3.8689,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305769335341595e-05,
+      "loss": 3.8598,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297383387831075e-05,
+      "loss": 3.8503,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5288997440320555e-05,
+      "loss": 3.8697,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5280611492810035e-05,
+      "loss": 3.8767,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527224192410325e-05,
+      "loss": 3.8614,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526385597659273e-05,
+      "loss": 3.8505,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525547002908221e-05,
+      "loss": 3.8612,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5247084081571684e-05,
+      "loss": 3.8513,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52387145128649e-05,
+      "loss": 3.8516,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523032856535438e-05,
+      "loss": 3.8658,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522194261784386e-05,
+      "loss": 3.8473,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521355667033333e-05,
+      "loss": 3.8643,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520518710162655e-05,
+      "loss": 3.8539,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519680115411603e-05,
+      "loss": 3.8566,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518841520660551e-05,
+      "loss": 3.8407,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518002925909499e-05,
+      "loss": 3.8593,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517164331158447e-05,
+      "loss": 3.8458,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5163273742877684e-05,
+      "loss": 3.8625,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515488779536716e-05,
+      "loss": 3.8545,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514650184785664e-05,
+      "loss": 3.8579,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513811590034612e-05,
+      "loss": 3.8361,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129746331639333e-05,
+      "loss": 3.8622,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121360384128807e-05,
+      "loss": 3.8512,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112974436618287e-05,
+      "loss": 3.855,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104588489107766e-05,
+      "loss": 3.8504,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096218920400976e-05,
+      "loss": 3.8599,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508783297289046e-05,
+      "loss": 3.8564,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507944702537994e-05,
+      "loss": 3.8583,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507106107786942e-05,
+      "loss": 3.844,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506269150916263e-05,
+      "loss": 3.8471,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505430556165211e-05,
+      "loss": 3.8493,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504591961414159e-05,
+      "loss": 3.8399,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503753366663107e-05,
+      "loss": 3.8533,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502916409792428e-05,
+      "loss": 3.8582,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502077815041376e-05,
+      "loss": 3.8475,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501239220290324e-05,
+      "loss": 3.8404,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500400625539272e-05,
+      "loss": 3.8528,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.911623001098633,
+      "eval_runtime": 305.2751,
+      "eval_samples_per_second": 1249.991,
+      "eval_steps_per_second": 39.063,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49956203078822e-05,
+      "loss": 3.8317,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498723436037168e-05,
+      "loss": 3.8407,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497884841286116e-05,
+      "loss": 3.8462,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497046246535065e-05,
+      "loss": 3.8452,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496207651784013e-05,
+      "loss": 3.8541,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495369057032961e-05,
+      "loss": 3.8328,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945304622819087e-05,
+      "loss": 3.8448,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936918675308567e-05,
+      "loss": 3.8298,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928549106601776e-05,
+      "loss": 3.8476,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4920163159091256e-05,
+      "loss": 3.8387,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911777211580736e-05,
+      "loss": 3.844,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903391264070216e-05,
+      "loss": 3.8525,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4895021695363425e-05,
+      "loss": 3.829,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886635747852905e-05,
+      "loss": 3.8415,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878249800342385e-05,
+      "loss": 3.8336,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869863852831864e-05,
+      "loss": 3.8351,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486149428412508e-05,
+      "loss": 3.8267,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485310833661456e-05,
+      "loss": 3.8344,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484472238910404e-05,
+      "loss": 3.8307,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483633644159352e-05,
+      "loss": 3.8524,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482796687288673e-05,
+      "loss": 3.8319,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481958092537621e-05,
+      "loss": 3.8412,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481119497786569e-05,
+      "loss": 3.8439,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480280903035516e-05,
+      "loss": 3.8448,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479443946164838e-05,
+      "loss": 3.823,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478605351413786e-05,
+      "loss": 3.8372,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477766756662734e-05,
+      "loss": 3.8239,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476928161911682e-05,
+      "loss": 3.8333,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47608956716063e-05,
+      "loss": 3.8213,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752526102899514e-05,
+      "loss": 3.8268,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4744140155388994e-05,
+      "loss": 3.8364,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47357705866822e-05,
+      "loss": 3.8366,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472738463917168e-05,
+      "loss": 3.8265,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471899869166116e-05,
+      "loss": 3.8371,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710612744150636e-05,
+      "loss": 3.8287,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4702226796640116e-05,
+      "loss": 3.8371,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693840849129596e-05,
+      "loss": 3.8176,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685454901619076e-05,
+      "loss": 3.8305,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4677068954108556e-05,
+      "loss": 3.8132,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466869938540177e-05,
+      "loss": 3.8239,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466031343789125e-05,
+      "loss": 3.8203,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465192749038073e-05,
+      "loss": 3.8348,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464354154287021e-05,
+      "loss": 3.8302,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463517197416342e-05,
+      "loss": 3.8277,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46267860266529e-05,
+      "loss": 3.8316,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461840007914238e-05,
+      "loss": 3.8309,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461001413163186e-05,
+      "loss": 3.827,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460164456292507e-05,
+      "loss": 3.8203,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459325861541455e-05,
+      "loss": 3.81,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458487266790403e-05,
+      "loss": 3.8315,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457648672039351e-05,
+      "loss": 3.8207,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4568117151686725e-05,
+      "loss": 3.8141,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559731204176205e-05,
+      "loss": 3.8162,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551345256665685e-05,
+      "loss": 3.8233,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542959309155165e-05,
+      "loss": 3.8078,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534589740448374e-05,
+      "loss": 3.8159,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4526203792937854e-05,
+      "loss": 3.8163,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517817845427334e-05,
+      "loss": 3.8176,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509431897916814e-05,
+      "loss": 3.8277,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4501062329210023e-05,
+      "loss": 3.8085,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44926763816995e-05,
+      "loss": 3.8103,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448429043418898e-05,
+      "loss": 3.822,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447590448667846e-05,
+      "loss": 3.8081,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446753491797168e-05,
+      "loss": 3.8126,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445914897046116e-05,
+      "loss": 3.821,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445076302295064e-05,
+      "loss": 3.8206,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444237707544012e-05,
+      "loss": 3.8046,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44339911279296e-05,
+      "loss": 3.7986,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442562155922281e-05,
+      "loss": 3.7997,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441723561171229e-05,
+      "loss": 3.8135,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440884966420177e-05,
+      "loss": 3.8269,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440046371669125e-05,
+      "loss": 3.8094,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439209414798446e-05,
+      "loss": 3.8224,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438370820047394e-05,
+      "loss": 3.8215,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437532225296342e-05,
+      "loss": 3.8239,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43669363054529e-05,
+      "loss": 3.8032,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435856673674611e-05,
+      "loss": 3.8192,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435018078923559e-05,
+      "loss": 3.8113,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434179484172507e-05,
+      "loss": 3.8045,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433340889421455e-05,
+      "loss": 3.8181,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432503932550776e-05,
+      "loss": 3.8122,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431665337799724e-05,
+      "loss": 3.8126,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430826743048672e-05,
+      "loss": 3.8141,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42998814829762e-05,
+      "loss": 3.8104,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429151191426941e-05,
+      "loss": 3.8128,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428312596675889e-05,
+      "loss": 3.8077,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427474001924837e-05,
+      "loss": 3.8113,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426635407173785e-05,
+      "loss": 3.8049,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257984503031066e-05,
+      "loss": 3.8049,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249598555520546e-05,
+      "loss": 3.8115,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4241212608010026e-05,
+      "loss": 3.8113,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232826660499506e-05,
+      "loss": 3.8014,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224457091792715e-05,
+      "loss": 3.8011,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216071144282195e-05,
+      "loss": 3.8178,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207685196771675e-05,
+      "loss": 3.8123,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4199299249261155e-05,
+      "loss": 3.8073,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190929680554364e-05,
+      "loss": 3.7966,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182543733043844e-05,
+      "loss": 3.8039,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4174157785533324e-05,
+      "loss": 3.8003,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165771838022804e-05,
+      "loss": 3.8144,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415740226931602e-05,
+      "loss": 3.8121,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41490163218055e-05,
+      "loss": 3.802,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414063037429498e-05,
+      "loss": 3.7956,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413224442678446e-05,
+      "loss": 3.809,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412387485807767e-05,
+      "loss": 3.7775,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411548891056715e-05,
+      "loss": 3.8039,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410710296305663e-05,
+      "loss": 3.7948,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409871701554611e-05,
+      "loss": 3.8017,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409034744683932e-05,
+      "loss": 3.7878,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40819614993288e-05,
+      "loss": 3.805,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407357555181828e-05,
+      "loss": 3.7971,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406518960430776e-05,
+      "loss": 3.8031,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056820035600974e-05,
+      "loss": 3.8069,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4048434088090454e-05,
+      "loss": 3.7923,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4040048140579934e-05,
+      "loss": 3.8129,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031662193069414e-05,
+      "loss": 3.8202,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402329262436262e-05,
+      "loss": 3.803,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40149066768521e-05,
+      "loss": 3.7918,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400652072934158e-05,
+      "loss": 3.8061,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399813478183106e-05,
+      "loss": 3.7983,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398976521312427e-05,
+      "loss": 3.7912,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398137926561375e-05,
+      "loss": 3.8105,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397299331810323e-05,
+      "loss": 3.7907,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396460737059271e-05,
+      "loss": 3.8046,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395623780188593e-05,
+      "loss": 3.799,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394785185437541e-05,
+      "loss": 3.8013,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393946590686489e-05,
+      "loss": 3.7919,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393107995935437e-05,
+      "loss": 3.7963,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922710390647576e-05,
+      "loss": 3.795,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3914324443137056e-05,
+      "loss": 3.808,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905938495626536e-05,
+      "loss": 3.7974,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897552548116016e-05,
+      "loss": 3.8012,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3889182979409225e-05,
+      "loss": 3.7805,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880797031898705e-05,
+      "loss": 3.8071,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872411084388185e-05,
+      "loss": 3.7944,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3864025136877665e-05,
+      "loss": 3.803,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385565556817088e-05,
+      "loss": 3.7935,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384726962066036e-05,
+      "loss": 3.8048,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383888367314984e-05,
+      "loss": 3.7995,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830497725639314e-05,
+      "loss": 3.808,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382212815693253e-05,
+      "loss": 3.7859,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381374220942201e-05,
+      "loss": 3.7956,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380535626191148e-05,
+      "loss": 3.7932,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379697031440096e-05,
+      "loss": 3.7872,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378860074569418e-05,
+      "loss": 3.8014,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378021479818366e-05,
+      "loss": 3.8009,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377182885067313e-05,
+      "loss": 3.799,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376344290316262e-05,
+      "loss": 3.7796,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3755073334455835e-05,
+      "loss": 3.7989,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.886021614074707,
+      "eval_runtime": 314.3717,
+      "eval_samples_per_second": 1213.821,
+      "eval_steps_per_second": 37.933,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3746687386945315e-05,
+      "loss": 3.7813,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373830143943479e-05,
+      "loss": 3.7831,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372991549192427e-05,
+      "loss": 3.7951,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372152954441375e-05,
+      "loss": 3.7933,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371314359690323e-05,
+      "loss": 3.795,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370475764939271e-05,
+      "loss": 3.7829,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369637170188219e-05,
+      "loss": 3.7882,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368798575437167e-05,
+      "loss": 3.7789,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367961618566488e-05,
+      "loss": 3.7913,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367123023815436e-05,
+      "loss": 3.7865,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366284429064384e-05,
+      "loss": 3.7889,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365445834313332e-05,
+      "loss": 3.798,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364608877442653e-05,
+      "loss": 3.7787,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363770282691601e-05,
+      "loss": 3.7897,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362931687940549e-05,
+      "loss": 3.7834,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362093093189497e-05,
+      "loss": 3.7817,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361256136318818e-05,
+      "loss": 3.7742,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360417541567766e-05,
+      "loss": 3.7843,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359578946816714e-05,
+      "loss": 3.7758,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358740352065662e-05,
+      "loss": 3.7977,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357903395194983e-05,
+      "loss": 3.785,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357064800443931e-05,
+      "loss": 3.7875,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356226205692879e-05,
+      "loss": 3.7931,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355387610941827e-05,
+      "loss": 3.7924,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3545506540711486e-05,
+      "loss": 3.7697,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3537120593200966e-05,
+      "loss": 3.784,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3528734645690446e-05,
+      "loss": 3.7774,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3520348698179926e-05,
+      "loss": 3.7802,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3511979129473135e-05,
+      "loss": 3.7707,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3503593181962615e-05,
+      "loss": 3.7743,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3495207234452095e-05,
+      "loss": 3.7862,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3486837665745304e-05,
+      "loss": 3.7829,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3478451718234784e-05,
+      "loss": 3.7792,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3470065770724264e-05,
+      "loss": 3.7829,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3461679823213744e-05,
+      "loss": 3.7809,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3453293875703224e-05,
+      "loss": 3.7829,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344490792819271e-05,
+      "loss": 3.7713,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343652198068219e-05,
+      "loss": 3.7717,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342813603317167e-05,
+      "loss": 3.7684,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341976646446488e-05,
+      "loss": 3.7718,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341138051695436e-05,
+      "loss": 3.7702,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.340299456944384e-05,
+      "loss": 3.7826,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339460862193332e-05,
+      "loss": 3.7809,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338623905322653e-05,
+      "loss": 3.7771,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337785310571601e-05,
+      "loss": 3.7856,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336946715820549e-05,
+      "loss": 3.7772,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336108121069497e-05,
+      "loss": 3.7794,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335271164198818e-05,
+      "loss": 3.7689,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3344325694477664e-05,
+      "loss": 3.7575,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3335939746967144e-05,
+      "loss": 3.7828,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3327553799456624e-05,
+      "loss": 3.7716,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331918423074983e-05,
+      "loss": 3.7688,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331079828323931e-05,
+      "loss": 3.7669,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330241233572879e-05,
+      "loss": 3.7711,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3294026388218266e-05,
+      "loss": 3.7587,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328565681951148e-05,
+      "loss": 3.7676,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327727087200096e-05,
+      "loss": 3.7693,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326888492449044e-05,
+      "loss": 3.7662,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3260498976979915e-05,
+      "loss": 3.7763,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325212940827313e-05,
+      "loss": 3.7585,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324374346076261e-05,
+      "loss": 3.7631,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32353575132521e-05,
+      "loss": 3.7738,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322697156574157e-05,
+      "loss": 3.7606,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321860199703479e-05,
+      "loss": 3.763,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321021604952427e-05,
+      "loss": 3.7686,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320183010201374e-05,
+      "loss": 3.7728,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3193460533306956e-05,
+      "loss": 3.7564,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3185074585796436e-05,
+      "loss": 3.757,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3176688638285916e-05,
+      "loss": 3.7483,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316830269077539e-05,
+      "loss": 3.7668,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3159933122068605e-05,
+      "loss": 3.7793,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3151547174558085e-05,
+      "loss": 3.7633,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3143161227047565e-05,
+      "loss": 3.7704,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3134775279537045e-05,
+      "loss": 3.7734,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312640571083026e-05,
+      "loss": 3.78,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311801976331974e-05,
+      "loss": 3.7556,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109633815809214e-05,
+      "loss": 3.7708,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3101247868298694e-05,
+      "loss": 3.7666,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309287829959191e-05,
+      "loss": 3.7545,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308449235208139e-05,
+      "loss": 3.7715,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307610640457086e-05,
+      "loss": 3.7673,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.306772045706034e-05,
+      "loss": 3.7646,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305935088835356e-05,
+      "loss": 3.7688,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305096494084304e-05,
+      "loss": 3.7636,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304257899333252e-05,
+      "loss": 3.758,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3034193045822e-05,
+      "loss": 3.7632,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025823477115214e-05,
+      "loss": 3.7653,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.301743752960469e-05,
+      "loss": 3.7539,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300905158209417e-05,
+      "loss": 3.7586,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300066563458365e-05,
+      "loss": 3.7671,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299229606587686e-05,
+      "loss": 3.7654,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2983910118366337e-05,
+      "loss": 3.7538,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2975524170855817e-05,
+      "loss": 3.7564,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2967138223345296e-05,
+      "loss": 3.7737,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295876865463851e-05,
+      "loss": 3.766,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295038270712799e-05,
+      "loss": 3.7634,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294199675961747e-05,
+      "loss": 3.7475,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293361081210695e-05,
+      "loss": 3.7606,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292524124340016e-05,
+      "loss": 3.7568,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291685529588964e-05,
+      "loss": 3.7663,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290846934837912e-05,
+      "loss": 3.7678,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29000834008686e-05,
+      "loss": 3.757,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289171383216181e-05,
+      "loss": 3.749,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288332788465129e-05,
+      "loss": 3.7622,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287494193714077e-05,
+      "loss": 3.7349,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286655598963025e-05,
+      "loss": 3.7571,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2858186420923466e-05,
+      "loss": 3.7464,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2849800473412946e-05,
+      "loss": 3.7608,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2841414525902426e-05,
+      "loss": 3.7432,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2833028578391906e-05,
+      "loss": 3.7602,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2824659009685115e-05,
+      "loss": 3.7508,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2816273062174595e-05,
+      "loss": 3.759,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2807887114664075e-05,
+      "loss": 3.7622,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2799501167153555e-05,
+      "loss": 3.7517,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2791131598446764e-05,
+      "loss": 3.7638,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2782745650936244e-05,
+      "loss": 3.7726,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2774359703425724e-05,
+      "loss": 3.7581,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2765973755915204e-05,
+      "loss": 3.7521,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275760418720842e-05,
+      "loss": 3.7565,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27492182396979e-05,
+      "loss": 3.7573,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274083229218738e-05,
+      "loss": 3.7433,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273244634467686e-05,
+      "loss": 3.7707,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272407677597007e-05,
+      "loss": 3.7435,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271569082845955e-05,
+      "loss": 3.7635,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270730488094903e-05,
+      "loss": 3.7547,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269891893343851e-05,
+      "loss": 3.7566,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269054936473172e-05,
+      "loss": 3.7496,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26821634172212e-05,
+      "loss": 3.7553,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267377746971068e-05,
+      "loss": 3.7495,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266539152220016e-05,
+      "loss": 3.7648,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265702195349337e-05,
+      "loss": 3.7522,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264863600598285e-05,
+      "loss": 3.7593,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264025005847233e-05,
+      "loss": 3.7392,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263186411096181e-05,
+      "loss": 3.7636,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262349454225502e-05,
+      "loss": 3.7498,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26151085947445e-05,
+      "loss": 3.7586,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260672264723398e-05,
+      "loss": 3.7527,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.259833669972346e-05,
+      "loss": 3.7572,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258996713101667e-05,
+      "loss": 3.7542,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258158118350615e-05,
+      "loss": 3.7686,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257319523599563e-05,
+      "loss": 3.7407,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256480928848511e-05,
+      "loss": 3.7534,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255643971977833e-05,
+      "loss": 3.7483,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254805377226781e-05,
+      "loss": 3.7482,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253966782475729e-05,
+      "loss": 3.7515,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253128187724677e-05,
+      "loss": 3.7593,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2522912308539976e-05,
+      "loss": 3.7585,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2514526361029456e-05,
+      "loss": 3.7374,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2506140413518936e-05,
+      "loss": 3.7596,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8695147037506104,
+      "eval_runtime": 304.879,
+      "eval_samples_per_second": 1251.615,
+      "eval_steps_per_second": 39.114,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2497770844812145e-05,
+      "loss": 3.7464,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2489384897301625e-05,
+      "loss": 3.7382,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2480998949791105e-05,
+      "loss": 3.7548,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2472613002280585e-05,
+      "loss": 3.7496,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2464243433573794e-05,
+      "loss": 3.7527,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245585748606328e-05,
+      "loss": 3.7392,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244747153855276e-05,
+      "loss": 3.7428,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243908559104224e-05,
+      "loss": 3.7387,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243071602233545e-05,
+      "loss": 3.7482,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.242233007482493e-05,
+      "loss": 3.7459,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.241394412731441e-05,
+      "loss": 3.7489,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.240555817980389e-05,
+      "loss": 3.7545,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23971886110971e-05,
+      "loss": 3.7394,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238880266358658e-05,
+      "loss": 3.7429,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238041671607606e-05,
+      "loss": 3.7452,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237203076856554e-05,
+      "loss": 3.7363,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.236364482105502e-05,
+      "loss": 3.7324,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23552588735445e-05,
+      "loss": 3.7397,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.234687292603398e-05,
+      "loss": 3.7364,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233848697852346e-05,
+      "loss": 3.752,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2330117409816674e-05,
+      "loss": 3.744,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2321731462306154e-05,
+      "loss": 3.7485,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2313345514795634e-05,
+      "loss": 3.7492,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2304959567285114e-05,
+      "loss": 3.7497,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.229658999857832e-05,
+      "loss": 3.7319,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.22882040510678e-05,
+      "loss": 3.7436,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227981810355728e-05,
+      "loss": 3.7318,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227143215604676e-05,
+      "loss": 3.7391,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.226306258733997e-05,
+      "loss": 3.733,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.225467663982945e-05,
+      "loss": 3.7316,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.224629069231893e-05,
+      "loss": 3.7469,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.223792112361215e-05,
+      "loss": 3.7424,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222953517610163e-05,
+      "loss": 3.7391,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222114922859111e-05,
+      "loss": 3.7417,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.221276328108059e-05,
+      "loss": 3.7383,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.220437733357006e-05,
+      "loss": 3.7457,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.219599138605954e-05,
+      "loss": 3.7329,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.218760543854902e-05,
+      "loss": 3.7324,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.21792194910385e-05,
+      "loss": 3.7266,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217084992233171e-05,
+      "loss": 3.7322,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.216246397482119e-05,
+      "loss": 3.7329,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.215407802731067e-05,
+      "loss": 3.7412,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2145708458603886e-05,
+      "loss": 3.7397,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2137322511093366e-05,
+      "loss": 3.7369,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2128936563582846e-05,
+      "loss": 3.745,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2120550616072326e-05,
+      "loss": 3.7386,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2112181047365535e-05,
+      "loss": 3.7407,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2103795099855015e-05,
+      "loss": 3.7276,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2095409152344495e-05,
+      "loss": 3.717,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2087023204833975e-05,
+      "loss": 3.745,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2078653636127184e-05,
+      "loss": 3.7321,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2070267688616664e-05,
+      "loss": 3.7262,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2061881741106144e-05,
+      "loss": 3.7311,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2053495793595623e-05,
+      "loss": 3.7274,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.204512622488884e-05,
+      "loss": 3.7235,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.203674027737832e-05,
+      "loss": 3.7248,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.20283543298678e-05,
+      "loss": 3.7279,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.201996838235728e-05,
+      "loss": 3.7279,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.201159881365049e-05,
+      "loss": 3.7392,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.200321286613997e-05,
+      "loss": 3.7201,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199482691862945e-05,
+      "loss": 3.7222,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.198644097111893e-05,
+      "loss": 3.7427,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197807140241214e-05,
+      "loss": 3.7152,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196968545490162e-05,
+      "loss": 3.7223,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.19612995073911e-05,
+      "loss": 3.7319,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195291355988058e-05,
+      "loss": 3.7381,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.194454399117379e-05,
+      "loss": 3.7176,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.193615804366327e-05,
+      "loss": 3.7181,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.192777209615275e-05,
+      "loss": 3.7083,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191938614864223e-05,
+      "loss": 3.7251,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191101657993544e-05,
+      "loss": 3.7409,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.190263063242492e-05,
+      "loss": 3.7255,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18942446849144e-05,
+      "loss": 3.7291,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.188585873740388e-05,
+      "loss": 3.7351,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.187748916869709e-05,
+      "loss": 3.7415,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186910322118657e-05,
+      "loss": 3.7192,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186071727367605e-05,
+      "loss": 3.732,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.185233132616553e-05,
+      "loss": 3.7285,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.184396175745875e-05,
+      "loss": 3.7161,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.183557580994823e-05,
+      "loss": 3.7314,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.182718986243771e-05,
+      "loss": 3.7307,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1818803914927187e-05,
+      "loss": 3.7303,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1810434346220396e-05,
+      "loss": 3.7269,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1802048398709876e-05,
+      "loss": 3.7275,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1793662451199356e-05,
+      "loss": 3.7169,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1785276503688836e-05,
+      "loss": 3.7271,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1776906934982045e-05,
+      "loss": 3.7264,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1768520987471525e-05,
+      "loss": 3.7199,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1760135039961005e-05,
+      "loss": 3.7221,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1751749092450484e-05,
+      "loss": 3.7216,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.17433795237437e-05,
+      "loss": 3.731,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.173499357623318e-05,
+      "loss": 3.7193,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.172660762872266e-05,
+      "loss": 3.7151,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.171822168121214e-05,
+      "loss": 3.736,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170985211250535e-05,
+      "loss": 3.7259,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170146616499483e-05,
+      "loss": 3.7278,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.169308021748431e-05,
+      "loss": 3.7094,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.168471064877752e-05,
+      "loss": 3.7245,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1676324701267e-05,
+      "loss": 3.7163,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.166793875375648e-05,
+      "loss": 3.7261,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165955280624596e-05,
+      "loss": 3.7337,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1651183237539174e-05,
+      "loss": 3.7172,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1642797290028654e-05,
+      "loss": 3.7162,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1634411342518134e-05,
+      "loss": 3.7213,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1626025395007614e-05,
+      "loss": 3.6995,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.161765582630082e-05,
+      "loss": 3.7243,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.16092698787903e-05,
+      "loss": 3.7059,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160088393127978e-05,
+      "loss": 3.7238,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.159249798376926e-05,
+      "loss": 3.7075,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.158412841506247e-05,
+      "loss": 3.7208,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.157574246755195e-05,
+      "loss": 3.7157,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.156735652004143e-05,
+      "loss": 3.7171,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155897057253091e-05,
+      "loss": 3.7296,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155060100382413e-05,
+      "loss": 3.7127,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154221505631361e-05,
+      "loss": 3.728,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.153382910880309e-05,
+      "loss": 3.7352,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.152544316129257e-05,
+      "loss": 3.7218,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151707359258578e-05,
+      "loss": 3.7191,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150868764507526e-05,
+      "loss": 3.715,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.150030169756474e-05,
+      "loss": 3.7198,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1491915750054217e-05,
+      "loss": 3.7136,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1483546181347426e-05,
+      "loss": 3.7311,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1475160233836906e-05,
+      "loss": 3.7039,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1466774286326386e-05,
+      "loss": 3.7249,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1458388338815866e-05,
+      "loss": 3.72,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.145001877010908e-05,
+      "loss": 3.7185,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144163282259856e-05,
+      "loss": 3.7092,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.143324687508804e-05,
+      "loss": 3.7209,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.142486092757752e-05,
+      "loss": 3.709,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.141649135887073e-05,
+      "loss": 3.7323,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.140810541136021e-05,
+      "loss": 3.7129,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139971946384969e-05,
+      "loss": 3.7235,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139133351633917e-05,
+      "loss": 3.7014,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.138296394763238e-05,
+      "loss": 3.729,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.137457800012186e-05,
+      "loss": 3.7163,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.136619205261134e-05,
+      "loss": 3.7225,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.135780610510082e-05,
+      "loss": 3.7161,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1349436536394035e-05,
+      "loss": 3.7164,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1341050588883515e-05,
+      "loss": 3.7219,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1332664641372995e-05,
+      "loss": 3.7346,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1324278693862475e-05,
+      "loss": 3.7024,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1315909125155684e-05,
+      "loss": 3.7163,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1307523177645164e-05,
+      "loss": 3.713,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1299137230134644e-05,
+      "loss": 3.7178,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1290751282624124e-05,
+      "loss": 3.7115,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128238171391733e-05,
+      "loss": 3.7208,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.127399576640681e-05,
+      "loss": 3.7216,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.126560981889629e-05,
+      "loss": 3.7054,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.12572402501895e-05,
+      "loss": 3.725,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.858961343765259,
+      "eval_runtime": 304.9762,
+      "eval_samples_per_second": 1251.216,
+      "eval_steps_per_second": 39.101,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124885430267898e-05,
+      "loss": 3.7083,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124046835516847e-05,
+      "loss": 3.7038,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123208240765795e-05,
+      "loss": 3.7189,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.122369646014743e-05,
+      "loss": 3.7107,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.121531051263691e-05,
+      "loss": 3.7169,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.120692456512638e-05,
+      "loss": 3.7038,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119853861761586e-05,
+      "loss": 3.7104,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119015267010534e-05,
+      "loss": 3.7032,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118178310139856e-05,
+      "loss": 3.7164,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117339715388803e-05,
+      "loss": 3.7051,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.116501120637751e-05,
+      "loss": 3.7131,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115662525886699e-05,
+      "loss": 3.7184,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1148272068963936e-05,
+      "loss": 3.7092,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113988612145342e-05,
+      "loss": 3.708,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11315001739429e-05,
+      "loss": 3.7072,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.112311422643238e-05,
+      "loss": 3.6987,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1114728278921855e-05,
+      "loss": 3.6992,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1106342331411335e-05,
+      "loss": 3.7055,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1097956383900815e-05,
+      "loss": 3.7041,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1089570436390295e-05,
+      "loss": 3.7115,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1081200867683504e-05,
+      "loss": 3.7152,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1072814920172984e-05,
+      "loss": 3.7103,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1064428972662464e-05,
+      "loss": 3.7132,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1056043025151944e-05,
+      "loss": 3.7157,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.104767345644516e-05,
+      "loss": 3.695,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103928750893464e-05,
+      "loss": 3.7141,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103090156142412e-05,
+      "loss": 3.6988,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10225156139136e-05,
+      "loss": 3.7015,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101414604520681e-05,
+      "loss": 3.6978,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100576009769629e-05,
+      "loss": 3.7019,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099737415018577e-05,
+      "loss": 3.707,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098900458147898e-05,
+      "loss": 3.713,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098061863396846e-05,
+      "loss": 3.7051,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097223268645794e-05,
+      "loss": 3.7104,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.096384673894742e-05,
+      "loss": 3.6982,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.09554607914369e-05,
+      "loss": 3.7115,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.094707484392638e-05,
+      "loss": 3.7058,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093868889641586e-05,
+      "loss": 3.6916,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0930302948905345e-05,
+      "loss": 3.6936,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0921933380198554e-05,
+      "loss": 3.6973,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0913547432688034e-05,
+      "loss": 3.7018,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0905161485177514e-05,
+      "loss": 3.7033,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0896775537666994e-05,
+      "loss": 3.7059,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08884059689602e-05,
+      "loss": 3.7014,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.088002002144968e-05,
+      "loss": 3.7132,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087163407393916e-05,
+      "loss": 3.7013,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.086326450523237e-05,
+      "loss": 3.7094,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085487855772185e-05,
+      "loss": 3.6967,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084649261021133e-05,
+      "loss": 3.6788,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083810666270081e-05,
+      "loss": 3.7134,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082973709399403e-05,
+      "loss": 3.6992,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082135114648351e-05,
+      "loss": 3.6983,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081296519897299e-05,
+      "loss": 3.6957,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080457925146247e-05,
+      "loss": 3.6949,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0796209682755676e-05,
+      "loss": 3.6918,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0787823735245156e-05,
+      "loss": 3.6897,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0779437787734636e-05,
+      "loss": 3.6963,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0771051840224116e-05,
+      "loss": 3.6913,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0762682271517325e-05,
+      "loss": 3.7093,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0754296324006805e-05,
+      "loss": 3.6895,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0745910376496285e-05,
+      "loss": 3.6884,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0737524428985765e-05,
+      "loss": 3.7074,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072915486027898e-05,
+      "loss": 3.6824,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072076891276846e-05,
+      "loss": 3.6875,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071238296525794e-05,
+      "loss": 3.7001,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.070399701774742e-05,
+      "loss": 3.7052,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069562744904063e-05,
+      "loss": 3.685,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068724150153011e-05,
+      "loss": 3.685,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067885555401959e-05,
+      "loss": 3.6783,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067046960650907e-05,
+      "loss": 3.6917,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066210003780228e-05,
+      "loss": 3.7072,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065371409029176e-05,
+      "loss": 3.6931,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064532814278124e-05,
+      "loss": 3.6949,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063694219527072e-05,
+      "loss": 3.7032,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0628572626563935e-05,
+      "loss": 3.7081,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0620186679053415e-05,
+      "loss": 3.6845,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0611817110346624e-05,
+      "loss": 3.702,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0603431162836104e-05,
+      "loss": 3.69,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0595045215325584e-05,
+      "loss": 3.6903,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0586659267815064e-05,
+      "loss": 3.6938,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0578273320304544e-05,
+      "loss": 3.7022,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0569887372794024e-05,
+      "loss": 3.6953,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0561501425283504e-05,
+      "loss": 3.7003,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0553115477772983e-05,
+      "loss": 3.6923,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054474590906619e-05,
+      "loss": 3.685,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053635996155567e-05,
+      "loss": 3.6959,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.052797401404515e-05,
+      "loss": 3.6938,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051958806653464e-05,
+      "loss": 3.6882,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051121849782785e-05,
+      "loss": 3.6899,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.050283255031733e-05,
+      "loss": 3.6913,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049444660280681e-05,
+      "loss": 3.6976,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.048606065529629e-05,
+      "loss": 3.6867,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04776910865895e-05,
+      "loss": 3.682,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046930513907898e-05,
+      "loss": 3.7015,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046091919156846e-05,
+      "loss": 3.698,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0452549622861666e-05,
+      "loss": 3.6949,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0444163675351146e-05,
+      "loss": 3.6813,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0435777727840626e-05,
+      "loss": 3.6895,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0427391780330106e-05,
+      "loss": 3.6845,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041902221162332e-05,
+      "loss": 3.695,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04106362641128e-05,
+      "loss": 3.6998,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040225031660228e-05,
+      "loss": 3.691,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039386436909176e-05,
+      "loss": 3.6857,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038549480038497e-05,
+      "loss": 3.6827,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037710885287445e-05,
+      "loss": 3.6737,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036873928416766e-05,
+      "loss": 3.6887,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036035333665714e-05,
+      "loss": 3.6772,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035196738914662e-05,
+      "loss": 3.694,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.03435814416361e-05,
+      "loss": 3.6764,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033519549412558e-05,
+      "loss": 3.6883,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032680954661506e-05,
+      "loss": 3.6825,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031842359910454e-05,
+      "loss": 3.6903,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031003765159402e-05,
+      "loss": 3.6942,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0301668082887236e-05,
+      "loss": 3.6839,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.029328213537671e-05,
+      "loss": 3.6979,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028489618786619e-05,
+      "loss": 3.7016,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027651024035567e-05,
+      "loss": 3.6913,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026814067164888e-05,
+      "loss": 3.6923,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025975472413836e-05,
+      "loss": 3.6827,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025136877662784e-05,
+      "loss": 3.6876,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.024298282911732e-05,
+      "loss": 3.6863,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0234613260410534e-05,
+      "loss": 3.6933,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0226227312900013e-05,
+      "loss": 3.6761,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0217841365389493e-05,
+      "loss": 3.6945,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0209455417878973e-05,
+      "loss": 3.6908,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020108584917218e-05,
+      "loss": 3.687,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.019269990166166e-05,
+      "loss": 3.6777,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.018431395415114e-05,
+      "loss": 3.6907,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.017592800664062e-05,
+      "loss": 3.6804,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.016755843793383e-05,
+      "loss": 3.6979,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015917249042331e-05,
+      "loss": 3.6871,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015078654291279e-05,
+      "loss": 3.6917,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014241697420601e-05,
+      "loss": 3.6702,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013403102669549e-05,
+      "loss": 3.7,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.012564507918497e-05,
+      "loss": 3.6826,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011725913167445e-05,
+      "loss": 3.6895,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0108889562967656e-05,
+      "loss": 3.6857,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0100503615457136e-05,
+      "loss": 3.6844,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0092117667946616e-05,
+      "loss": 3.6915,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0083731720436096e-05,
+      "loss": 3.7015,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0075362151729305e-05,
+      "loss": 3.6745,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0066976204218785e-05,
+      "loss": 3.6882,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0058590256708265e-05,
+      "loss": 3.6834,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0050204309197745e-05,
+      "loss": 3.6839,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004183474049096e-05,
+      "loss": 3.6804,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.003346517178418e-05,
+      "loss": 3.692,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.002507922427366e-05,
+      "loss": 3.6914,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.001669327676313e-05,
+      "loss": 3.6733,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000830732925261e-05,
+      "loss": 3.6914,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8515453338623047,
+      "eval_runtime": 305.233,
+      "eval_samples_per_second": 1250.163,
+      "eval_steps_per_second": 39.069,
+      "step": 610560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999992138174209e-05,
+      "loss": 3.6797,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999153543423157e-05,
+      "loss": 3.6767,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.998314948672105e-05,
+      "loss": 3.6833,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.997477991801426e-05,
+      "loss": 3.6786,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.996639397050374e-05,
+      "loss": 3.6875,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.995800802299322e-05,
+      "loss": 3.6701,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994963845428643e-05,
+      "loss": 3.686,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9941252506775915e-05,
+      "loss": 3.6731,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9932866559265395e-05,
+      "loss": 3.6834,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9924480611754874e-05,
+      "loss": 3.6778,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9916094664244354e-05,
+      "loss": 3.6792,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9907708716733834e-05,
+      "loss": 3.6893,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9899339148027044e-05,
+      "loss": 3.6779,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9890953200516523e-05,
+      "loss": 3.6801,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9882567253006003e-05,
+      "loss": 3.6793,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.987418130549548e-05,
+      "loss": 3.6678,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.986579535798496e-05,
+      "loss": 3.6679,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.985740941047444e-05,
+      "loss": 3.6778,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984902346296392e-05,
+      "loss": 3.6741,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98406375154534e-05,
+      "loss": 3.677,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983226794674661e-05,
+      "loss": 3.6865,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98238819992361e-05,
+      "loss": 3.6822,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.981549605172558e-05,
+      "loss": 3.6845,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.980711010421506e-05,
+      "loss": 3.6839,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.979874053550827e-05,
+      "loss": 3.6697,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.979035458799775e-05,
+      "loss": 3.681,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.978196864048723e-05,
+      "loss": 3.6695,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.977358269297671e-05,
+      "loss": 3.6753,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.976521312426992e-05,
+      "loss": 3.6716,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.97568271767594e-05,
+      "loss": 3.6696,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.974844122924888e-05,
+      "loss": 3.6763,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.974005528173836e-05,
+      "loss": 3.6813,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9731685713031566e-05,
+      "loss": 3.6758,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9723299765521046e-05,
+      "loss": 3.6785,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.971491381801053e-05,
+      "loss": 3.6692,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.970654424930374e-05,
+      "loss": 3.6839,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.969815830179322e-05,
+      "loss": 3.6758,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96897723542827e-05,
+      "loss": 3.6633,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.968138640677218e-05,
+      "loss": 3.6658,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.967301683806539e-05,
+      "loss": 3.6647,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.966463089055487e-05,
+      "loss": 3.6735,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.965624494304435e-05,
+      "loss": 3.6752,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.964785899553383e-05,
+      "loss": 3.6754,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963948942682704e-05,
+      "loss": 3.6746,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963110347931652e-05,
+      "loss": 3.6811,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9622717531806e-05,
+      "loss": 3.6746,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9614331584295486e-05,
+      "loss": 3.6787,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9605978394392425e-05,
+      "loss": 3.6704,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9597592446881905e-05,
+      "loss": 3.6444,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9589206499371384e-05,
+      "loss": 3.6898,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9580820551860864e-05,
+      "loss": 3.6677,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9572434604350344e-05,
+      "loss": 3.6703,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9564048656839824e-05,
+      "loss": 3.6677,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9555662709329304e-05,
+      "loss": 3.6638,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.954729314062251e-05,
+      "loss": 3.659,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953890719311199e-05,
+      "loss": 3.6604,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953052124560147e-05,
+      "loss": 3.6699,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952213529809095e-05,
+      "loss": 3.6618,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.951374935058043e-05,
+      "loss": 3.6812,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.950536340306991e-05,
+      "loss": 3.6634,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.949697745555939e-05,
+      "loss": 3.6578,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948859150804887e-05,
+      "loss": 3.6778,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948022193934208e-05,
+      "loss": 3.6565,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947183599183156e-05,
+      "loss": 3.659,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.946345004432104e-05,
+      "loss": 3.6723,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.945506409681052e-05,
+      "loss": 3.6736,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.944669452810373e-05,
+      "loss": 3.6584,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.943830858059321e-05,
+      "loss": 3.6575,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942992263308269e-05,
+      "loss": 3.6501,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942155306437591e-05,
+      "loss": 3.6632,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.941316711686539e-05,
+      "loss": 3.6788,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.940478116935487e-05,
+      "loss": 3.6644,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.939639522184435e-05,
+      "loss": 3.6696,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9388025653137556e-05,
+      "loss": 3.6726,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9379639705627036e-05,
+      "loss": 3.6787,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9371253758116516e-05,
+      "loss": 3.6592,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.936288418940973e-05,
+      "loss": 3.6742,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9354498241899205e-05,
+      "loss": 3.6612,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9346112294388685e-05,
+      "loss": 3.6637,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9337726346878165e-05,
+      "loss": 3.6641,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9329340399367645e-05,
+      "loss": 3.6732,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932095445185713e-05,
+      "loss": 3.667,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.931256850434661e-05,
+      "loss": 3.6745,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.930418255683609e-05,
+      "loss": 3.66,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.92958129881293e-05,
+      "loss": 3.6597,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.928742704061878e-05,
+      "loss": 3.6686,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.927904109310826e-05,
+      "loss": 3.6643,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.927067152440147e-05,
+      "loss": 3.6556,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926228557689095e-05,
+      "loss": 3.6649,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.925389962938043e-05,
+      "loss": 3.6633,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.924551368186991e-05,
+      "loss": 3.6722,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.923714411316312e-05,
+      "loss": 3.6584,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.92287581656526e-05,
+      "loss": 3.6545,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9220372218142085e-05,
+      "loss": 3.6757,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9211986270631565e-05,
+      "loss": 3.6667,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9203616701924774e-05,
+      "loss": 3.6662,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9195230754414254e-05,
+      "loss": 3.6557,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9186844806903734e-05,
+      "loss": 3.6593,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9178458859393214e-05,
+      "loss": 3.6578,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.917008929068642e-05,
+      "loss": 3.6674,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.91617033431759e-05,
+      "loss": 3.6715,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.915331739566538e-05,
+      "loss": 3.6639,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.914493144815486e-05,
+      "loss": 3.6576,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.913656187944807e-05,
+      "loss": 3.6503,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.912817593193755e-05,
+      "loss": 3.6537,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911980636323077e-05,
+      "loss": 3.6575,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911142041572025e-05,
+      "loss": 3.65,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.910303446820973e-05,
+      "loss": 3.6674,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.909464852069921e-05,
+      "loss": 3.6516,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.908626257318869e-05,
+      "loss": 3.6572,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.90778930044819e-05,
+      "loss": 3.6526,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906950705697138e-05,
+      "loss": 3.6641,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906112110946086e-05,
+      "loss": 3.6655,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.905273516195034e-05,
+      "loss": 3.6546,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.904434921443982e-05,
+      "loss": 3.6709,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9035963266929297e-05,
+      "loss": 3.676,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9027577319418777e-05,
+      "loss": 3.6634,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.901920775071199e-05,
+      "loss": 3.6634,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.901082180320147e-05,
+      "loss": 3.6565,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.900243585569095e-05,
+      "loss": 3.6555,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.899406628698416e-05,
+      "loss": 3.6645,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.898568033947364e-05,
+      "loss": 3.663,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.897729439196312e-05,
+      "loss": 3.6493,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.89689084444526e-05,
+      "loss": 3.6687,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.896052249694208e-05,
+      "loss": 3.6663,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.895213654943156e-05,
+      "loss": 3.6601,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8943750601921034e-05,
+      "loss": 3.6529,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8935364654410514e-05,
+      "loss": 3.659,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.892699508570373e-05,
+      "loss": 3.6518,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.891860913819321e-05,
+      "loss": 3.6692,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.891022319068269e-05,
+      "loss": 3.6585,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.890183724317217e-05,
+      "loss": 3.6679,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8893467674465386e-05,
+      "loss": 3.6435,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8885081726954866e-05,
+      "loss": 3.6717,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.887669577944434e-05,
+      "loss": 3.6536,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8868326210737555e-05,
+      "loss": 3.6662,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8859940263227035e-05,
+      "loss": 3.6614,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885155431571651e-05,
+      "loss": 3.6531,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.884316836820599e-05,
+      "loss": 3.6637,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8834798799499204e-05,
+      "loss": 3.6758,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8826412851988684e-05,
+      "loss": 3.6464,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8818026904478164e-05,
+      "loss": 3.6625,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8809640956967644e-05,
+      "loss": 3.6557,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880127138826086e-05,
+      "loss": 3.6563,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.879288544075034e-05,
+      "loss": 3.657,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.878451587204355e-05,
+      "loss": 3.6627,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.877612992453303e-05,
+      "loss": 3.6658,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.876774397702251e-05,
+      "loss": 3.6536,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.875935802951198e-05,
+      "loss": 3.663,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8471899032592773,
+      "eval_runtime": 305.6564,
+      "eval_samples_per_second": 1248.431,
+      "eval_steps_per_second": 39.014,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.875097208200146e-05,
+      "loss": 3.6452,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.874258613449094e-05,
+      "loss": 3.651,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.873420018698042e-05,
+      "loss": 3.6584,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.87258142394699e-05,
+      "loss": 3.6543,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.871742829195938e-05,
+      "loss": 3.6603,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.87090587232526e-05,
+      "loss": 3.6456,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870067277574208e-05,
+      "loss": 3.6595,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869228682823156e-05,
+      "loss": 3.647,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.868390088072104e-05,
+      "loss": 3.6516,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.867551493321052e-05,
+      "loss": 3.6542,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8667145364503726e-05,
+      "loss": 3.6511,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8658759416993206e-05,
+      "loss": 3.6591,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8650373469482686e-05,
+      "loss": 3.6636,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8641987521972166e-05,
+      "loss": 3.6477,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8633601574461646e-05,
+      "loss": 3.6532,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8625215626951126e-05,
+      "loss": 3.6432,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8616846058244335e-05,
+      "loss": 3.6406,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8608460110733815e-05,
+      "loss": 3.654,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.86000741632233e-05,
+      "loss": 3.645,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.859168821571278e-05,
+      "loss": 3.6511,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.858330226820226e-05,
+      "loss": 3.6621,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.857491632069174e-05,
+      "loss": 3.6589,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.856654675198495e-05,
+      "loss": 3.6593,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.855816080447443e-05,
+      "loss": 3.6569,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854977485696391e-05,
+      "loss": 3.644,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854138890945339e-05,
+      "loss": 3.6533,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.853300296194287e-05,
+      "loss": 3.6455,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8524617014432344e-05,
+      "loss": 3.6516,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8516231066921824e-05,
+      "loss": 3.6412,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8507845119411304e-05,
+      "loss": 3.646,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849947555070452e-05,
+      "loss": 3.6487,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8491105981997736e-05,
+      "loss": 3.6546,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8482720034487216e-05,
+      "loss": 3.6528,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8474334086976695e-05,
+      "loss": 3.651,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.846594813946617e-05,
+      "loss": 3.645,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.845756219195565e-05,
+      "loss": 3.6545,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.844917624444513e-05,
+      "loss": 3.6537,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8440806675738344e-05,
+      "loss": 3.6405,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843242072822782e-05,
+      "loss": 3.6421,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84240347807173e-05,
+      "loss": 3.6389,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.841564883320678e-05,
+      "loss": 3.6422,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.840726288569626e-05,
+      "loss": 3.6483,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839887693818574e-05,
+      "loss": 3.6527,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8390490990675224e-05,
+      "loss": 3.6494,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8382105043164704e-05,
+      "loss": 3.6526,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837373547445791e-05,
+      "loss": 3.6528,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836534952694739e-05,
+      "loss": 3.6525,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835696357943687e-05,
+      "loss": 3.6433,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834859401073008e-05,
+      "loss": 3.6224,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834020806321956e-05,
+      "loss": 3.6617,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833182211570904e-05,
+      "loss": 3.6418,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.832343616819852e-05,
+      "loss": 3.6456,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8315050220688e-05,
+      "loss": 3.6454,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.830666427317748e-05,
+      "loss": 3.6386,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.829827832566696e-05,
+      "loss": 3.6341,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828990875696017e-05,
+      "loss": 3.6355,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828152280944966e-05,
+      "loss": 3.6449,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.827313686193914e-05,
+      "loss": 3.6365,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.826475091442862e-05,
+      "loss": 3.6555,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.825638134572183e-05,
+      "loss": 3.638,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.824799539821131e-05,
+      "loss": 3.628,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823960945070079e-05,
+      "loss": 3.6545,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823122350319027e-05,
+      "loss": 3.6363,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8222853934483476e-05,
+      "loss": 3.6325,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8214467986972956e-05,
+      "loss": 3.6471,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8206082039462436e-05,
+      "loss": 3.6507,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8197696091951916e-05,
+      "loss": 3.6319,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8189326523245125e-05,
+      "loss": 3.6304,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818094057573461e-05,
+      "loss": 3.6263,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817255462822409e-05,
+      "loss": 3.6402,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.816416868071357e-05,
+      "loss": 3.6508,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.815579911200678e-05,
+      "loss": 3.6408,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.814741316449626e-05,
+      "loss": 3.6452,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813904359578947e-05,
+      "loss": 3.6435,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813065764827895e-05,
+      "loss": 3.6516,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812227170076843e-05,
+      "loss": 3.6378,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.811388575325791e-05,
+      "loss": 3.6511,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810549980574739e-05,
+      "loss": 3.6315,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.809711385823687e-05,
+      "loss": 3.6395,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.808872791072635e-05,
+      "loss": 3.6392,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.808034196321583e-05,
+      "loss": 3.6449,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8071988773312774e-05,
+      "loss": 3.6453,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8063602825802254e-05,
+      "loss": 3.6486,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8055216878291734e-05,
+      "loss": 3.6363,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8046830930781214e-05,
+      "loss": 3.6351,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8038444983270694e-05,
+      "loss": 3.6416,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8030059035760174e-05,
+      "loss": 3.647,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8021673088249654e-05,
+      "loss": 3.6309,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.801328714073913e-05,
+      "loss": 3.6373,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.800491757203234e-05,
+      "loss": 3.641,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799653162452182e-05,
+      "loss": 3.6436,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79881456770113e-05,
+      "loss": 3.638,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797975972950078e-05,
+      "loss": 3.6301,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7971390160794e-05,
+      "loss": 3.6468,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796300421328348e-05,
+      "loss": 3.6447,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.795461826577295e-05,
+      "loss": 3.6435,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794623231826243e-05,
+      "loss": 3.6309,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793784637075191e-05,
+      "loss": 3.6364,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792946042324139e-05,
+      "loss": 3.6324,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792107447573087e-05,
+      "loss": 3.6455,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.791270490702408e-05,
+      "loss": 3.6452,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.790431895951356e-05,
+      "loss": 3.6357,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789593301200304e-05,
+      "loss": 3.6381,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.788754706449252e-05,
+      "loss": 3.625,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7879177495785737e-05,
+      "loss": 3.6326,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787080792707895e-05,
+      "loss": 3.6295,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7862421979568426e-05,
+      "loss": 3.6282,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7854036032057906e-05,
+      "loss": 3.6388,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7845650084547386e-05,
+      "loss": 3.6291,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7837264137036865e-05,
+      "loss": 3.6307,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7828878189526345e-05,
+      "loss": 3.6323,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7820492242015825e-05,
+      "loss": 3.6323,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7812106294505305e-05,
+      "loss": 3.6433,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7803736725798514e-05,
+      "loss": 3.6323,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7795350778287994e-05,
+      "loss": 3.6479,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778698120958121e-05,
+      "loss": 3.6506,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777859526207069e-05,
+      "loss": 3.6414,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777020931456017e-05,
+      "loss": 3.6345,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776182336704965e-05,
+      "loss": 3.6354,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.775343741953913e-05,
+      "loss": 3.6342,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.774505147202861e-05,
+      "loss": 3.6398,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.773666552451809e-05,
+      "loss": 3.6366,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.772827957700757e-05,
+      "loss": 3.6266,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771991000830078e-05,
+      "loss": 3.6458,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771152406079026e-05,
+      "loss": 3.6446,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.770313811327974e-05,
+      "loss": 3.6303,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.769475216576922e-05,
+      "loss": 3.6295,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768638259706243e-05,
+      "loss": 3.639,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7678013028355644e-05,
+      "loss": 3.6265,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7669627080845124e-05,
+      "loss": 3.6489,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7661241133334604e-05,
+      "loss": 3.6328,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7652855185824084e-05,
+      "loss": 3.6455,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7644469238313564e-05,
+      "loss": 3.6201,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7636083290803044e-05,
+      "loss": 3.6431,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7627697343292524e-05,
+      "loss": 3.6361,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761932777458573e-05,
+      "loss": 3.6358,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761094182707521e-05,
+      "loss": 3.6395,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760255587956469e-05,
+      "loss": 3.6268,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.759416993205417e-05,
+      "loss": 3.6382,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758578398454365e-05,
+      "loss": 3.6564,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757739803703313e-05,
+      "loss": 3.6218,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756902846832634e-05,
+      "loss": 3.6383,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756064252081583e-05,
+      "loss": 3.6325,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755225657330531e-05,
+      "loss": 3.6279,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754387062579479e-05,
+      "loss": 3.6335,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753548467828426e-05,
+      "loss": 3.6392,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.752709873077374e-05,
+      "loss": 3.6404,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751871278326322e-05,
+      "loss": 3.6322,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751034321455644e-05,
+      "loss": 3.6364,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8442981243133545,
+      "eval_runtime": 304.1689,
+      "eval_samples_per_second": 1254.537,
+      "eval_steps_per_second": 39.205,
+      "step": 763200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.750195726704591e-05,
+      "loss": 3.6256,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.749357131953539e-05,
+      "loss": 3.6278,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.748518537202487e-05,
+      "loss": 3.6351,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.747679942451435e-05,
+      "loss": 3.6317,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.746841347700383e-05,
+      "loss": 3.6399,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7460043908297046e-05,
+      "loss": 3.6233,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7451657960786526e-05,
+      "loss": 3.6354,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7443272013276006e-05,
+      "loss": 3.6221,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7434886065765486e-05,
+      "loss": 3.633,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7426500118254966e-05,
+      "loss": 3.6307,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7418130549548175e-05,
+      "loss": 3.626,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7409744602037655e-05,
+      "loss": 3.6386,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7401375033330864e-05,
+      "loss": 3.6406,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7392989085820344e-05,
+      "loss": 3.6223,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7384603138309824e-05,
+      "loss": 3.6322,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7376217190799304e-05,
+      "loss": 3.6161,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7367831243288784e-05,
+      "loss": 3.6178,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7359445295778264e-05,
+      "loss": 3.6291,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735105934826775e-05,
+      "loss": 3.6208,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.734267340075723e-05,
+      "loss": 3.6295,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.733428745324671e-05,
+      "loss": 3.6389,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.732590150573619e-05,
+      "loss": 3.6323,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.731751555822567e-05,
+      "loss": 3.635,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730912961071515e-05,
+      "loss": 3.6323,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730076004200836e-05,
+      "loss": 3.6228,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729237409449784e-05,
+      "loss": 3.6286,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.728398814698732e-05,
+      "loss": 3.6239,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.727560219947679e-05,
+      "loss": 3.6312,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.726723263077001e-05,
+      "loss": 3.6164,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.725884668325949e-05,
+      "loss": 3.6247,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.725046073574897e-05,
+      "loss": 3.6245,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7242091167042184e-05,
+      "loss": 3.6315,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7233705219531664e-05,
+      "loss": 3.6337,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7225319272021144e-05,
+      "loss": 3.6245,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7216933324510624e-05,
+      "loss": 3.625,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.72085473770001e-05,
+      "loss": 3.6278,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.720017780829331e-05,
+      "loss": 3.6342,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719179186078279e-05,
+      "loss": 3.6165,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7183405913272266e-05,
+      "loss": 3.6188,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7175019965761746e-05,
+      "loss": 3.6184,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7166634018251226e-05,
+      "loss": 3.6205,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7158248070740706e-05,
+      "loss": 3.6232,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7149862123230186e-05,
+      "loss": 3.6314,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714147617571967e-05,
+      "loss": 3.628,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.713310660701288e-05,
+      "loss": 3.6286,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.712472065950236e-05,
+      "loss": 3.6289,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.711635109079557e-05,
+      "loss": 3.6294,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.710796514328505e-05,
+      "loss": 3.6225,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709957919577453e-05,
+      "loss": 3.599,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709119324826401e-05,
+      "loss": 3.6377,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.708282367955722e-05,
+      "loss": 3.6181,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.70744377320467e-05,
+      "loss": 3.6229,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.706605178453618e-05,
+      "loss": 3.6223,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.705766583702566e-05,
+      "loss": 3.6138,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704927988951514e-05,
+      "loss": 3.6154,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704089394200462e-05,
+      "loss": 3.6115,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7032524373297836e-05,
+      "loss": 3.6263,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7024138425787316e-05,
+      "loss": 3.6111,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7015752478276795e-05,
+      "loss": 3.6315,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7007366530766275e-05,
+      "loss": 3.6191,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6998996962059485e-05,
+      "loss": 3.6032,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6990611014548964e-05,
+      "loss": 3.6337,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6982225067038444e-05,
+      "loss": 3.6118,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6973839119527924e-05,
+      "loss": 3.6076,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6965453172017404e-05,
+      "loss": 3.6238,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6957067224506884e-05,
+      "loss": 3.629,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6948697655800093e-05,
+      "loss": 3.6072,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694031170828957e-05,
+      "loss": 3.6093,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693192576077906e-05,
+      "loss": 3.6047,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.692353981326854e-05,
+      "loss": 3.6146,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.691515386575802e-05,
+      "loss": 3.6263,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69067679182475e-05,
+      "loss": 3.6216,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.689838197073698e-05,
+      "loss": 3.6222,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.689001240203019e-05,
+      "loss": 3.6203,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688162645451967e-05,
+      "loss": 3.63,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.687324050700915e-05,
+      "loss": 3.6189,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.686485455949863e-05,
+      "loss": 3.6293,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68564686119881e-05,
+      "loss": 3.6087,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.684808266447758e-05,
+      "loss": 3.6159,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683969671696706e-05,
+      "loss": 3.6165,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683132714826028e-05,
+      "loss": 3.6202,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.682294120074976e-05,
+      "loss": 3.6251,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.681455525323924e-05,
+      "loss": 3.6262,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.680616930572872e-05,
+      "loss": 3.6168,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.67977833582182e-05,
+      "loss": 3.6121,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678941378951141e-05,
+      "loss": 3.6179,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678102784200089e-05,
+      "loss": 3.6245,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677264189449037e-05,
+      "loss": 3.6084,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.676425594697985e-05,
+      "loss": 3.6143,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6755869999469327e-05,
+      "loss": 3.6234,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6747484051958807e-05,
+      "loss": 3.6176,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6739098104448286e-05,
+      "loss": 3.6148,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6730728535741496e-05,
+      "loss": 3.6097,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672234258823098e-05,
+      "loss": 3.6219,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.671395664072046e-05,
+      "loss": 3.6254,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.670557069320994e-05,
+      "loss": 3.6193,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.669718474569942e-05,
+      "loss": 3.6106,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.66887987981889e-05,
+      "loss": 3.6129,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668041285067838e-05,
+      "loss": 3.6122,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667202690316786e-05,
+      "loss": 3.6226,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.666365733446107e-05,
+      "loss": 3.6247,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.665528776575428e-05,
+      "loss": 3.6128,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664690181824376e-05,
+      "loss": 3.6138,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663851587073324e-05,
+      "loss": 3.6084,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663012992322272e-05,
+      "loss": 3.6085,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6621760354515936e-05,
+      "loss": 3.6112,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6613374407005416e-05,
+      "loss": 3.6102,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6604988459494896e-05,
+      "loss": 3.6114,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6596602511984376e-05,
+      "loss": 3.6107,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6588216564473856e-05,
+      "loss": 3.6065,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6579846995767065e-05,
+      "loss": 3.6152,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6571461048256545e-05,
+      "loss": 3.6086,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6563075100746025e-05,
+      "loss": 3.6229,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6554689153235505e-05,
+      "loss": 3.608,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6546303205724985e-05,
+      "loss": 3.6277,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6537933637018194e-05,
+      "loss": 3.6286,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6529547689507674e-05,
+      "loss": 3.6222,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6521161741997154e-05,
+      "loss": 3.6126,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6512775794486634e-05,
+      "loss": 3.6148,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6504389846976114e-05,
+      "loss": 3.6128,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6496003899465594e-05,
+      "loss": 3.619,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6487617951955074e-05,
+      "loss": 3.614,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.647924838324829e-05,
+      "loss": 3.6056,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.647086243573776e-05,
+      "loss": 3.6211,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.646247648822724e-05,
+      "loss": 3.6242,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.645409054071672e-05,
+      "loss": 3.608,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.64457045932062e-05,
+      "loss": 3.6102,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.643733502449941e-05,
+      "loss": 3.616,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642894907698889e-05,
+      "loss": 3.604,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642056312947837e-05,
+      "loss": 3.6307,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641217718196785e-05,
+      "loss": 3.6092,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.640379123445734e-05,
+      "loss": 3.6253,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.639540528694682e-05,
+      "loss": 3.5981,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63870193394363e-05,
+      "loss": 3.622,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637864977072951e-05,
+      "loss": 3.616,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637026382321899e-05,
+      "loss": 3.6158,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636187787570847e-05,
+      "loss": 3.6159,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6353508307001676e-05,
+      "loss": 3.6081,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6345122359491156e-05,
+      "loss": 3.6182,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6336736411980636e-05,
+      "loss": 3.6345,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6328350464470116e-05,
+      "loss": 3.6036,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6319964516959596e-05,
+      "loss": 3.6137,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6311594948252805e-05,
+      "loss": 3.6142,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.630320900074229e-05,
+      "loss": 3.6097,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.629482305323177e-05,
+      "loss": 3.613,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.628643710572125e-05,
+      "loss": 3.6166,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.627806753701446e-05,
+      "loss": 3.6162,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626968158950394e-05,
+      "loss": 3.6137,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626129564199342e-05,
+      "loss": 3.6133,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8424363136291504,
+      "eval_runtime": 304.9822,
+      "eval_samples_per_second": 1251.191,
+      "eval_steps_per_second": 39.101,
+      "step": 839520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.62529096944829e-05,
+      "loss": 3.5942,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.624452374697238e-05,
+      "loss": 3.6039,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.623613779946186e-05,
+      "loss": 3.6164,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.622775185195134e-05,
+      "loss": 3.6112,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621936590444082e-05,
+      "loss": 3.619,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621099633573403e-05,
+      "loss": 3.6043,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.620261038822351e-05,
+      "loss": 3.6115,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.619422444071299e-05,
+      "loss": 3.6052,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6185838493202476e-05,
+      "loss": 3.6109,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6177468924495685e-05,
+      "loss": 3.6033,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6169082976985165e-05,
+      "loss": 3.6054,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6160697029474645e-05,
+      "loss": 3.6214,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6152327460767854e-05,
+      "loss": 3.6142,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6143941513257334e-05,
+      "loss": 3.6026,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6135555565746814e-05,
+      "loss": 3.6096,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6127169618236294e-05,
+      "loss": 3.5942,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6118783670725774e-05,
+      "loss": 3.599,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611039772321525e-05,
+      "loss": 3.6091,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.610201177570473e-05,
+      "loss": 3.5972,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6093625828194214e-05,
+      "loss": 3.607,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.608525625948742e-05,
+      "loss": 3.615,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.60768703119769e-05,
+      "loss": 3.6129,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606848436446638e-05,
+      "loss": 3.6179,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606009841695586e-05,
+      "loss": 3.6122,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605171246944534e-05,
+      "loss": 3.6028,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.604332652193482e-05,
+      "loss": 3.6057,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.603495695322803e-05,
+      "loss": 3.6013,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.602657100571751e-05,
+      "loss": 3.6105,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.601818505820699e-05,
+      "loss": 3.5961,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.600979911069647e-05,
+      "loss": 3.6051,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.600142954198968e-05,
+      "loss": 3.6064,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.599304359447917e-05,
+      "loss": 3.6064,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.598465764696865e-05,
+      "loss": 3.6084,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.597627169945813e-05,
+      "loss": 3.6084,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.596788575194761e-05,
+      "loss": 3.6011,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595951618324082e-05,
+      "loss": 3.6094,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59511302357303e-05,
+      "loss": 3.6163,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5942744288219777e-05,
+      "loss": 3.5956,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5934358340709257e-05,
+      "loss": 3.5972,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5925988772002466e-05,
+      "loss": 3.5986,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5917602824491946e-05,
+      "loss": 3.5967,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5909216876981426e-05,
+      "loss": 3.6018,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5900830929470906e-05,
+      "loss": 3.6078,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5892444981960385e-05,
+      "loss": 3.6107,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58840754132536e-05,
+      "loss": 3.6053,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.587568946574308e-05,
+      "loss": 3.6099,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.586730351823256e-05,
+      "loss": 3.6104,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585891757072204e-05,
+      "loss": 3.6048,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585054800201525e-05,
+      "loss": 3.5797,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.584216205450473e-05,
+      "loss": 3.6147,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.583377610699421e-05,
+      "loss": 3.5967,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.582539015948369e-05,
+      "loss": 3.6038,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58170205907769e-05,
+      "loss": 3.5969,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580863464326638e-05,
+      "loss": 3.5926,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580024869575586e-05,
+      "loss": 3.5967,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579186274824534e-05,
+      "loss": 3.5939,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.578347680073482e-05,
+      "loss": 3.6047,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5775107232028035e-05,
+      "loss": 3.5879,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5766721284517515e-05,
+      "loss": 3.612,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5758335337006995e-05,
+      "loss": 3.6011,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5749949389496475e-05,
+      "loss": 3.5794,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5741563441985955e-05,
+      "loss": 3.6132,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5733193873279164e-05,
+      "loss": 3.5917,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5724807925768644e-05,
+      "loss": 3.5916,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5716421978258124e-05,
+      "loss": 3.6002,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5708036030747604e-05,
+      "loss": 3.6088,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569966646204081e-05,
+      "loss": 3.588,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.569128051453029e-05,
+      "loss": 3.5956,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.568289456701977e-05,
+      "loss": 3.5816,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.567450861950925e-05,
+      "loss": 3.593,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.566612267199873e-05,
+      "loss": 3.6087,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.565773672448821e-05,
+      "loss": 3.5977,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564936715578143e-05,
+      "loss": 3.6016,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564098120827091e-05,
+      "loss": 3.6002,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563259526076038e-05,
+      "loss": 3.6102,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.562420931324986e-05,
+      "loss": 3.6001,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.561583974454308e-05,
+      "loss": 3.6064,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.560745379703255e-05,
+      "loss": 3.5912,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559906784952203e-05,
+      "loss": 3.596,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559068190201151e-05,
+      "loss": 3.5953,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5582312333304726e-05,
+      "loss": 3.6002,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5573926385794206e-05,
+      "loss": 3.6035,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5565540438283686e-05,
+      "loss": 3.6064,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5557154490773166e-05,
+      "loss": 3.6031,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5548768543262646e-05,
+      "loss": 3.5874,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5540398974555855e-05,
+      "loss": 3.5991,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5532013027045335e-05,
+      "loss": 3.6033,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5523627079534815e-05,
+      "loss": 3.5885,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5515241132024295e-05,
+      "loss": 3.5954,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5506855184513775e-05,
+      "loss": 3.6041,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5498485615806984e-05,
+      "loss": 3.595,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5490099668296464e-05,
+      "loss": 3.5993,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5481713720785944e-05,
+      "loss": 3.5872,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.547332777327543e-05,
+      "loss": 3.6041,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.546494182576491e-05,
+      "loss": 3.6025,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.545655587825439e-05,
+      "loss": 3.6049,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.544816993074387e-05,
+      "loss": 3.5834,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543978398323335e-05,
+      "loss": 3.5969,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543141441452656e-05,
+      "loss": 3.5903,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.542304484581977e-05,
+      "loss": 3.602,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.541465889830925e-05,
+      "loss": 3.6061,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.540627295079873e-05,
+      "loss": 3.5931,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.539788700328821e-05,
+      "loss": 3.5941,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538950105577769e-05,
+      "loss": 3.5866,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53811314870709e-05,
+      "loss": 3.5877,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5372745539560385e-05,
+      "loss": 3.5927,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5364359592049865e-05,
+      "loss": 3.5899,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5355973644539344e-05,
+      "loss": 3.593,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5347587697028824e-05,
+      "loss": 3.5906,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5339201749518304e-05,
+      "loss": 3.5873,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5330832180811513e-05,
+      "loss": 3.5897,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5322446233300993e-05,
+      "loss": 3.5883,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5314060285790473e-05,
+      "loss": 3.6064,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.530567433827995e-05,
+      "loss": 3.5902,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.529728839076943e-05,
+      "loss": 3.6036,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528891882206264e-05,
+      "loss": 3.6053,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528053287455212e-05,
+      "loss": 3.6084,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52721469270416e-05,
+      "loss": 3.5897,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.526376097953108e-05,
+      "loss": 3.5954,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52553914108243e-05,
+      "loss": 3.5942,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.524700546331378e-05,
+      "loss": 3.5976,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523861951580326e-05,
+      "loss": 3.598,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523023356829274e-05,
+      "loss": 3.5882,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.522184762078222e-05,
+      "loss": 3.5957,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.521346167327169e-05,
+      "loss": 3.6056,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.520507572576117e-05,
+      "loss": 3.5955,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.519670615705439e-05,
+      "loss": 3.5873,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.518832020954386e-05,
+      "loss": 3.5961,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517993426203334e-05,
+      "loss": 3.5869,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517154831452282e-05,
+      "loss": 3.608,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.51631623670123e-05,
+      "loss": 3.5908,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5154792798305516e-05,
+      "loss": 3.6047,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5146406850794996e-05,
+      "loss": 3.5794,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5138020903284476e-05,
+      "loss": 3.6004,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512965133457769e-05,
+      "loss": 3.5951,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5121265387067165e-05,
+      "loss": 3.5936,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5112879439556645e-05,
+      "loss": 3.5989,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5104493492046125e-05,
+      "loss": 3.5863,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5096107544535605e-05,
+      "loss": 3.6014,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5087737975828814e-05,
+      "loss": 3.6118,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5079352028318294e-05,
+      "loss": 3.5842,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5070966080807774e-05,
+      "loss": 3.5913,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5062580133297254e-05,
+      "loss": 3.5954,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.505421056459047e-05,
+      "loss": 3.592,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.504582461707995e-05,
+      "loss": 3.5905,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.503743866956943e-05,
+      "loss": 3.6004,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.502905272205891e-05,
+      "loss": 3.5933,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.502066677454839e-05,
+      "loss": 3.5943,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.50122972058416e-05,
+      "loss": 3.5958,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.841254711151123,
+      "eval_runtime": 304.0455,
+      "eval_samples_per_second": 1255.046,
+      "eval_steps_per_second": 39.221,
+      "step": 915840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.500391125833108e-05,
+      "loss": 3.5895,
+      "step": 915968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.499552531082056e-05,
+      "loss": 3.5794,
+      "step": 916480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.498713936331004e-05,
+      "loss": 3.5983,
+      "step": 916992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.497876979460325e-05,
+      "loss": 3.5928,
+      "step": 917504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.497038384709273e-05,
+      "loss": 3.5947,
+      "step": 918016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.496199789958221e-05,
+      "loss": 3.5843,
+      "step": 918528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4953611952071694e-05,
+      "loss": 3.5961,
+      "step": 919040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4945226004561174e-05,
+      "loss": 3.5843,
+      "step": 919552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4936840057050654e-05,
+      "loss": 3.5909,
+      "step": 920064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.492847048834386e-05,
+      "loss": 3.5859,
+      "step": 920576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.492008454083334e-05,
+      "loss": 3.5908,
+      "step": 921088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.491169859332282e-05,
+      "loss": 3.5982,
+      "step": 921600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.490332902461603e-05,
+      "loss": 3.5993,
+      "step": 922112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.489494307710551e-05,
+      "loss": 3.5822,
+      "step": 922624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.488655712959499e-05,
+      "loss": 3.5905,
+      "step": 923136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.487817118208447e-05,
+      "loss": 3.5737,
+      "step": 923648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.486978523457395e-05,
+      "loss": 3.584,
+      "step": 924160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.486139928706343e-05,
+      "loss": 3.5873,
+      "step": 924672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.485301333955291e-05,
+      "loss": 3.5801,
+      "step": 925184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.484462739204239e-05,
+      "loss": 3.5875,
+      "step": 925696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.483625782333561e-05,
+      "loss": 3.5954,
+      "step": 926208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.482787187582509e-05,
+      "loss": 3.5893,
+      "step": 926720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.481948592831457e-05,
+      "loss": 3.6011,
+      "step": 927232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.481109998080405e-05,
+      "loss": 3.5942,
+      "step": 927744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.480271403329352e-05,
+      "loss": 3.5802,
+      "step": 928256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4794328085783e-05,
+      "loss": 3.5911,
+      "step": 928768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4785958517076217e-05,
+      "loss": 3.5823,
+      "step": 929280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4777572569565697e-05,
+      "loss": 3.5888,
+      "step": 929792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.476918662205517e-05,
+      "loss": 3.58,
+      "step": 930304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.476080067454465e-05,
+      "loss": 3.5821,
+      "step": 930816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4752431105837866e-05,
+      "loss": 3.5885,
+      "step": 931328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4744045158327345e-05,
+      "loss": 3.5866,
+      "step": 931840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4735659210816825e-05,
+      "loss": 3.587,
+      "step": 932352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4727273263306305e-05,
+      "loss": 3.5923,
+      "step": 932864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4718887315795785e-05,
+      "loss": 3.5845,
+      "step": 933376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4710517747088994e-05,
+      "loss": 3.5871,
+      "step": 933888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4702131799578474e-05,
+      "loss": 3.5969,
+      "step": 934400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4693745852067954e-05,
+      "loss": 3.5722,
+      "step": 934912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4685359904557434e-05,
+      "loss": 3.5845,
+      "step": 935424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4676990335850643e-05,
+      "loss": 3.581,
+      "step": 935936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.466860438834012e-05,
+      "loss": 3.577,
+      "step": 936448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.46602184408296e-05,
+      "loss": 3.5825,
+      "step": 936960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465183249331908e-05,
+      "loss": 3.5872,
+      "step": 937472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.46434629246123e-05,
+      "loss": 3.5922,
+      "step": 937984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.463507697710178e-05,
+      "loss": 3.5882,
+      "step": 938496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.462669102959126e-05,
+      "loss": 3.5908,
+      "step": 939008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.461830508208074e-05,
+      "loss": 3.5927,
+      "step": 939520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.460991913457022e-05,
+      "loss": 3.5825,
+      "step": 940032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.460154956586343e-05,
+      "loss": 3.5649,
+      "step": 940544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.459316361835291e-05,
+      "loss": 3.5919,
+      "step": 941056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.458477767084239e-05,
+      "loss": 3.5779,
+      "step": 941568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.457639172333187e-05,
+      "loss": 3.5824,
+      "step": 942080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.456802215462508e-05,
+      "loss": 3.5806,
+      "step": 942592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.455963620711456e-05,
+      "loss": 3.5767,
+      "step": 943104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.455125025960404e-05,
+      "loss": 3.5765,
+      "step": 943616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.454286431209352e-05,
+      "loss": 3.5741,
+      "step": 944128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4534478364583004e-05,
+      "loss": 3.5867,
+      "step": 944640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.452610879587621e-05,
+      "loss": 3.5679,
+      "step": 945152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.451772284836569e-05,
+      "loss": 3.596,
+      "step": 945664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.450933690085517e-05,
+      "loss": 3.5826,
+      "step": 946176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.450095095334465e-05,
+      "loss": 3.5645,
+      "step": 946688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.449258138463786e-05,
+      "loss": 3.5912,
+      "step": 947200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.448419543712734e-05,
+      "loss": 3.5702,
+      "step": 947712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.447580948961682e-05,
+      "loss": 3.5766,
+      "step": 948224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.44674235421063e-05,
+      "loss": 3.5786,
+      "step": 948736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.445903759459578e-05,
+      "loss": 3.5914,
+      "step": 949248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.445066802588899e-05,
+      "loss": 3.5711,
+      "step": 949760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.444228207837847e-05,
+      "loss": 3.5746,
+      "step": 950272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.443389613086796e-05,
+      "loss": 3.5599,
+      "step": 950784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.442551018335744e-05,
+      "loss": 3.5787,
+      "step": 951296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4417140614650646e-05,
+      "loss": 3.5845,
+      "step": 951808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4408754667140126e-05,
+      "loss": 3.584,
+      "step": 952320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4400368719629606e-05,
+      "loss": 3.5831,
+      "step": 952832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4391982772119086e-05,
+      "loss": 3.5834,
+      "step": 953344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4383613203412295e-05,
+      "loss": 3.5948,
+      "step": 953856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4375227255901775e-05,
+      "loss": 3.579,
+      "step": 954368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4366841308391255e-05,
+      "loss": 3.5866,
+      "step": 954880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4358455360880735e-05,
+      "loss": 3.5745,
+      "step": 955392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4350069413370215e-05,
+      "loss": 3.5793,
+      "step": 955904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4341699844663424e-05,
+      "loss": 3.5755,
+      "step": 956416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.433331389715291e-05,
+      "loss": 3.578,
+      "step": 956928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.432492794964239e-05,
+      "loss": 3.5877,
+      "step": 957440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.431654200213187e-05,
+      "loss": 3.5841,
+      "step": 957952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.430817243342508e-05,
+      "loss": 3.5888,
+      "step": 958464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.429978648591456e-05,
+      "loss": 3.5677,
+      "step": 958976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.429140053840404e-05,
+      "loss": 3.5804,
+      "step": 959488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.428301459089352e-05,
+      "loss": 3.5858,
+      "step": 960000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4274628643383e-05,
+      "loss": 3.5722,
+      "step": 960512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.426625907467621e-05,
+      "loss": 3.5731,
+      "step": 961024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.425787312716569e-05,
+      "loss": 3.5846,
+      "step": 961536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424948717965517e-05,
+      "loss": 3.5811,
+      "step": 962048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424110123214465e-05,
+      "loss": 3.5774,
+      "step": 962560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4232731663437865e-05,
+      "loss": 3.5716,
+      "step": 963072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4224345715927345e-05,
+      "loss": 3.5809,
+      "step": 963584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4215959768416825e-05,
+      "loss": 3.5894,
+      "step": 964096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4207573820906304e-05,
+      "loss": 3.5833,
+      "step": 964608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4199204252199514e-05,
+      "loss": 3.5665,
+      "step": 965120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4190818304688994e-05,
+      "loss": 3.5794,
+      "step": 965632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4182432357178473e-05,
+      "loss": 3.5717,
+      "step": 966144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4174046409667953e-05,
+      "loss": 3.5833,
+      "step": 966656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.416567684096116e-05,
+      "loss": 3.5896,
+      "step": 967168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.415729089345064e-05,
+      "loss": 3.5747,
+      "step": 967680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.414890494594012e-05,
+      "loss": 3.5732,
+      "step": 968192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.41405189984296e-05,
+      "loss": 3.5718,
+      "step": 968704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.413214942972282e-05,
+      "loss": 3.5705,
+      "step": 969216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.41237634822123e-05,
+      "loss": 3.5674,
+      "step": 969728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.411537753470178e-05,
+      "loss": 3.5765,
+      "step": 970240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.410699158719125e-05,
+      "loss": 3.5738,
+      "step": 970752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.409862201848447e-05,
+      "loss": 3.5741,
+      "step": 971264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.409023607097395e-05,
+      "loss": 3.5671,
+      "step": 971776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.408185012346343e-05,
+      "loss": 3.575,
+      "step": 972288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.40734641759529e-05,
+      "loss": 3.5682,
+      "step": 972800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4065094607246116e-05,
+      "loss": 3.5867,
+      "step": 973312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4056708659735596e-05,
+      "loss": 3.5708,
+      "step": 973824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.404832271222507e-05,
+      "loss": 3.5855,
+      "step": 974336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4039936764714556e-05,
+      "loss": 3.5838,
+      "step": 974848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.403156719600777e-05,
+      "loss": 3.5952,
+      "step": 975360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.402318124849725e-05,
+      "loss": 3.5754,
+      "step": 975872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4014795300986725e-05,
+      "loss": 3.575,
+      "step": 976384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4006409353476205e-05,
+      "loss": 3.5771,
+      "step": 976896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3998023405965685e-05,
+      "loss": 3.5769,
+      "step": 977408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.39896538372589e-05,
+      "loss": 3.5745,
+      "step": 977920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3981267889748374e-05,
+      "loss": 3.5774,
+      "step": 978432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3972881942237854e-05,
+      "loss": 3.5747,
+      "step": 978944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.396451237353107e-05,
+      "loss": 3.5871,
+      "step": 979456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.395612642602054e-05,
+      "loss": 3.5756,
+      "step": 979968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.394774047851002e-05,
+      "loss": 3.5725,
+      "step": 980480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393935453099951e-05,
+      "loss": 3.5788,
+      "step": 980992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393096858348899e-05,
+      "loss": 3.572,
+      "step": 981504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.39225990147822e-05,
+      "loss": 3.5882,
+      "step": 982016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.391421306727168e-05,
+      "loss": 3.5717,
+      "step": 982528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.390582711976116e-05,
+      "loss": 3.5874,
+      "step": 983040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.389744117225064e-05,
+      "loss": 3.5674,
+      "step": 983552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.388907160354385e-05,
+      "loss": 3.5806,
+      "step": 984064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.388068565603333e-05,
+      "loss": 3.5769,
+      "step": 984576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.387229970852281e-05,
+      "loss": 3.5775,
+      "step": 985088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.386391376101229e-05,
+      "loss": 3.5736,
+      "step": 985600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.385552781350177e-05,
+      "loss": 3.5761,
+      "step": 986112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.384715824479498e-05,
+      "loss": 3.5839,
+      "step": 986624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3838772297284463e-05,
+      "loss": 3.5902,
+      "step": 987136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.383038634977394e-05,
+      "loss": 3.57,
+      "step": 987648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.382200040226342e-05,
+      "loss": 3.5729,
+      "step": 988160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.381363083355663e-05,
+      "loss": 3.5783,
+      "step": 988672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.380524488604611e-05,
+      "loss": 3.5759,
+      "step": 989184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.379685893853559e-05,
+      "loss": 3.57,
+      "step": 989696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.378847299102507e-05,
+      "loss": 3.5855,
+      "step": 990208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.378010342231828e-05,
+      "loss": 3.5738,
+      "step": 990720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.377171747480776e-05,
+      "loss": 3.5829,
+      "step": 991232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.376333152729724e-05,
+      "loss": 3.5717,
+      "step": 991744
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8402984142303467,
+      "eval_runtime": 304.6228,
+      "eval_samples_per_second": 1252.667,
+      "eval_steps_per_second": 39.147,
+      "step": 992160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.375494557978672e-05,
+      "loss": 3.5755,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.37465596322762e-05,
+      "loss": 3.5602,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.373817368476568e-05,
+      "loss": 3.5799,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372978773725516e-05,
+      "loss": 3.5767,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372141816854838e-05,
+      "loss": 3.5757,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.371303222103786e-05,
+      "loss": 3.5715,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.370464627352734e-05,
+      "loss": 3.5728,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.369626032601682e-05,
+      "loss": 3.5703,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3687890757310026e-05,
+      "loss": 3.568,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3679504809799506e-05,
+      "loss": 3.5727,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3671118862288986e-05,
+      "loss": 3.5692,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3662732914778466e-05,
+      "loss": 3.5798,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3654363346071675e-05,
+      "loss": 3.5878,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3645977398561155e-05,
+      "loss": 3.5628,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3637591451050635e-05,
+      "loss": 3.5667,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3629205503540115e-05,
+      "loss": 3.5611,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3620819556029595e-05,
+      "loss": 3.5659,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.361243360851908e-05,
+      "loss": 3.5709,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.360404766100856e-05,
+      "loss": 3.5625,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3595661713498035e-05,
+      "loss": 3.5647,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.358729214479125e-05,
+      "loss": 3.5832,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357890619728073e-05,
+      "loss": 3.5711,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3570520249770204e-05,
+      "loss": 3.581,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3562134302259684e-05,
+      "loss": 3.5784,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.35537647335529e-05,
+      "loss": 3.5654,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.354537878604238e-05,
+      "loss": 3.5728,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.353699283853185e-05,
+      "loss": 3.5641,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.352860689102133e-05,
+      "loss": 3.5708,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.352023732231455e-05,
+      "loss": 3.5628,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3511851374804035e-05,
+      "loss": 3.5634,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.350346542729351e-05,
+      "loss": 3.574,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.349507947978299e-05,
+      "loss": 3.567,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.348669353227247e-05,
+      "loss": 3.5738,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347832396356568e-05,
+      "loss": 3.5687,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346993801605516e-05,
+      "loss": 3.5669,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346155206854464e-05,
+      "loss": 3.5714,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.345316612103412e-05,
+      "loss": 3.5814,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3444796552327326e-05,
+      "loss": 3.5508,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3436410604816806e-05,
+      "loss": 3.5692,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3428024657306286e-05,
+      "loss": 3.5654,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341963870979577e-05,
+      "loss": 3.5576,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341126914108898e-05,
+      "loss": 3.5624,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.340288319357846e-05,
+      "loss": 3.5766,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.339449724606794e-05,
+      "loss": 3.5719,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.338611129855742e-05,
+      "loss": 3.5725,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.337774172985063e-05,
+      "loss": 3.5718,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336935578234011e-05,
+      "loss": 3.5735,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336096983482959e-05,
+      "loss": 3.5643,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.335258388731907e-05,
+      "loss": 3.5498,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.334421431861228e-05,
+      "loss": 3.5764,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.333582837110176e-05,
+      "loss": 3.5636,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.332744242359124e-05,
+      "loss": 3.5681,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3319056476080727e-05,
+      "loss": 3.5617,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3310686907373936e-05,
+      "loss": 3.5582,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3302300959863416e-05,
+      "loss": 3.5575,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3293915012352896e-05,
+      "loss": 3.5548,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3285529064842376e-05,
+      "loss": 3.5713,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3277159496135585e-05,
+      "loss": 3.5555,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3268773548625065e-05,
+      "loss": 3.5731,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3260387601114545e-05,
+      "loss": 3.567,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3252001653604025e-05,
+      "loss": 3.5509,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3243632084897234e-05,
+      "loss": 3.5717,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3235246137386714e-05,
+      "loss": 3.5547,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3226860189876194e-05,
+      "loss": 3.5599,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321847424236568e-05,
+      "loss": 3.5569,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321010467365889e-05,
+      "loss": 3.5751,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320171872614837e-05,
+      "loss": 3.5573,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.319333277863785e-05,
+      "loss": 3.5553,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.318494683112733e-05,
+      "loss": 3.5433,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.317657726242054e-05,
+      "loss": 3.5598,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.316819131491002e-05,
+      "loss": 3.5696,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.31598053673995e-05,
+      "loss": 3.5706,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315141941988898e-05,
+      "loss": 3.5622,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.314304985118219e-05,
+      "loss": 3.5669,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.313466390367167e-05,
+      "loss": 3.5757,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.312627795616115e-05,
+      "loss": 3.5593,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3117892008650634e-05,
+      "loss": 3.5669,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3109506061140114e-05,
+      "loss": 3.5635,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310113649243332e-05,
+      "loss": 3.5599,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30927505449228e-05,
+      "loss": 3.5578,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.308436459741228e-05,
+      "loss": 3.5623,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.307597864990176e-05,
+      "loss": 3.5709,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.306760908119497e-05,
+      "loss": 3.5699,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.305922313368445e-05,
+      "loss": 3.5712,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.305083718617393e-05,
+      "loss": 3.5514,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304245123866341e-05,
+      "loss": 3.5645,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.303408166995662e-05,
+      "loss": 3.5661,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30256957224461e-05,
+      "loss": 3.5553,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.301730977493558e-05,
+      "loss": 3.5573,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300892382742507e-05,
+      "loss": 3.566,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.300055425871828e-05,
+      "loss": 3.5609,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2992168311207757e-05,
+      "loss": 3.5629,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2983782363697237e-05,
+      "loss": 3.5569,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2975396416186717e-05,
+      "loss": 3.5675,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2967026847479926e-05,
+      "loss": 3.5715,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2958640899969406e-05,
+      "loss": 3.567,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2950254952458886e-05,
+      "loss": 3.5495,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2941869004948365e-05,
+      "loss": 3.5617,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2933499436241575e-05,
+      "loss": 3.5583,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2925113488731055e-05,
+      "loss": 3.5627,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2916727541220534e-05,
+      "loss": 3.5746,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.290834159371002e-05,
+      "loss": 3.5562,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289997202500323e-05,
+      "loss": 3.5558,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289158607749271e-05,
+      "loss": 3.5528,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.288320012998219e-05,
+      "loss": 3.5609,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.287481418247167e-05,
+      "loss": 3.5525,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.286644461376488e-05,
+      "loss": 3.5562,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.285805866625436e-05,
+      "loss": 3.5579,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284967271874384e-05,
+      "loss": 3.5574,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.284128677123332e-05,
+      "loss": 3.5513,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.283291720252653e-05,
+      "loss": 3.5562,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.282453125501601e-05,
+      "loss": 3.5492,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.281614530750549e-05,
+      "loss": 3.5707,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2807759359994975e-05,
+      "loss": 3.5563,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2799373412484455e-05,
+      "loss": 3.5672,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2791003843777664e-05,
+      "loss": 3.5671,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2782617896267144e-05,
+      "loss": 3.5808,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2774231948756624e-05,
+      "loss": 3.5571,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.276586238004983e-05,
+      "loss": 3.5614,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.275747643253931e-05,
+      "loss": 3.5618,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.274909048502879e-05,
+      "loss": 3.5576,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.274070453751827e-05,
+      "loss": 3.5576,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.273231859000775e-05,
+      "loss": 3.56,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.272394902130096e-05,
+      "loss": 3.5585,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.271556307379044e-05,
+      "loss": 3.5665,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.270717712627993e-05,
+      "loss": 3.5582,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.269879117876941e-05,
+      "loss": 3.5541,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.269042161006262e-05,
+      "loss": 3.5605,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.26820356625521e-05,
+      "loss": 3.5572,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.267364971504158e-05,
+      "loss": 3.5646,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.266526376753106e-05,
+      "loss": 3.5614,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2656894198824267e-05,
+      "loss": 3.5644,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2648508251313747e-05,
+      "loss": 3.5532,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2640122303803226e-05,
+      "loss": 3.5594,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.26317363562927e-05,
+      "loss": 3.5643,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2623366787585916e-05,
+      "loss": 3.5584,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2614980840075395e-05,
+      "loss": 3.5587,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.260659489256488e-05,
+      "loss": 3.5588,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2598208945054355e-05,
+      "loss": 3.5667,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258983937634757e-05,
+      "loss": 3.5753,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258145342883705e-05,
+      "loss": 3.5556,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.257306748132653e-05,
+      "loss": 3.5541,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2564681533816004e-05,
+      "loss": 3.5576,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.255631196510922e-05,
+      "loss": 3.5612,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.25479260175987e-05,
+      "loss": 3.5508,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253954007008817e-05,
+      "loss": 3.5705,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253115412257765e-05,
+      "loss": 3.5595,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.252278455387087e-05,
+      "loss": 3.5667,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.251439860636035e-05,
+      "loss": 3.5508,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.841140031814575,
+      "eval_runtime": 305.9348,
+      "eval_samples_per_second": 1247.295,
+      "eval_steps_per_second": 38.979,
+      "step": 1068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.250601265884983e-05,
+      "loss": 3.557,
+      "step": 1068544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.249762671133931e-05,
+      "loss": 3.5497,
+      "step": 1069056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2489257142632525e-05,
+      "loss": 3.5609,
+      "step": 1069568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2480871195122e-05,
+      "loss": 3.5604,
+      "step": 1070080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.247248524761148e-05,
+      "loss": 3.5612,
+      "step": 1070592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.246409930010096e-05,
+      "loss": 3.5552,
+      "step": 1071104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2455729731394174e-05,
+      "loss": 3.5538,
+      "step": 1071616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.244734378388365e-05,
+      "loss": 3.554,
+      "step": 1072128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.243895783637313e-05,
+      "loss": 3.5516,
+      "step": 1072640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.243057188886261e-05,
+      "loss": 3.5564,
+      "step": 1073152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.242220232015582e-05,
+      "loss": 3.5527,
+      "step": 1073664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.24138163726453e-05,
+      "loss": 3.566,
+      "step": 1074176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.240543042513478e-05,
+      "loss": 3.5689,
+      "step": 1074688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.239704447762426e-05,
+      "loss": 3.5476,
+      "step": 1075200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.238867490891747e-05,
+      "loss": 3.5546,
+      "step": 1075712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.238028896140695e-05,
+      "loss": 3.542,
+      "step": 1076224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.237190301389643e-05,
+      "loss": 3.5517,
+      "step": 1076736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.236351706638591e-05,
+      "loss": 3.5492,
+      "step": 1077248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.235514749767912e-05,
+      "loss": 3.551,
+      "step": 1077760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.23467615501686e-05,
+      "loss": 3.5446,
+      "step": 1078272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.233837560265808e-05,
+      "loss": 3.5679,
+      "step": 1078784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.232998965514756e-05,
+      "loss": 3.5548,
+      "step": 1079296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2321620086440777e-05,
+      "loss": 3.5628,
+      "step": 1079808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2313234138930256e-05,
+      "loss": 3.5595,
+      "step": 1080320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2304848191419736e-05,
+      "loss": 3.5555,
+      "step": 1080832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2296462243909216e-05,
+      "loss": 3.5522,
+      "step": 1081344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2288092675202425e-05,
+      "loss": 3.549,
+      "step": 1081856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2279706727691905e-05,
+      "loss": 3.5546,
+      "step": 1082368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2271320780181385e-05,
+      "loss": 3.547,
+      "step": 1082880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2262934832670865e-05,
+      "loss": 3.548,
+      "step": 1083392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2254565263964074e-05,
+      "loss": 3.5556,
+      "step": 1083904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2246179316453554e-05,
+      "loss": 3.5493,
+      "step": 1084416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2237793368943034e-05,
+      "loss": 3.5613,
+      "step": 1084928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2229407421432514e-05,
+      "loss": 3.5517,
+      "step": 1085440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.222103785272573e-05,
+      "loss": 3.5525,
+      "step": 1085952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.221265190521521e-05,
+      "loss": 3.5612,
+      "step": 1086464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.220426595770469e-05,
+      "loss": 3.561,
+      "step": 1086976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.219588001019417e-05,
+      "loss": 3.5358,
+      "step": 1087488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.218751044148738e-05,
+      "loss": 3.5516,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.217912449397686e-05,
+      "loss": 3.5501,
+      "step": 1088512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.217073854646634e-05,
+      "loss": 3.54,
+      "step": 1089024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.216235259895582e-05,
+      "loss": 3.5524,
+      "step": 1089536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.215398303024903e-05,
+      "loss": 3.5625,
+      "step": 1090048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.214559708273851e-05,
+      "loss": 3.5528,
+      "step": 1090560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.213721113522799e-05,
+      "loss": 3.5564,
+      "step": 1091072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.212882518771747e-05,
+      "loss": 3.5584,
+      "step": 1091584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2120455619010684e-05,
+      "loss": 3.5558,
+      "step": 1092096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2112069671500164e-05,
+      "loss": 3.5499,
+      "step": 1092608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2103683723989644e-05,
+      "loss": 3.5359,
+      "step": 1093120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2095297776479124e-05,
+      "loss": 3.5567,
+      "step": 1093632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.208692820777233e-05,
+      "loss": 3.5493,
+      "step": 1094144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.207854226026181e-05,
+      "loss": 3.5505,
+      "step": 1094656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.207015631275129e-05,
+      "loss": 3.5465,
+      "step": 1095168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.206177036524077e-05,
+      "loss": 3.5428,
+      "step": 1095680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.205340079653398e-05,
+      "loss": 3.5383,
+      "step": 1096192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.204501484902346e-05,
+      "loss": 3.5401,
+      "step": 1096704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.203662890151294e-05,
+      "loss": 3.5584,
+      "step": 1097216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.202824295400242e-05,
+      "loss": 3.536,
+      "step": 1097728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201987338529564e-05,
+      "loss": 3.5548,
+      "step": 1098240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.201148743778512e-05,
+      "loss": 3.5505,
+      "step": 1098752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20031014902746e-05,
+      "loss": 3.5386,
+      "step": 1099264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.199471554276408e-05,
+      "loss": 3.551,
+      "step": 1099776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1986345974057287e-05,
+      "loss": 3.5411,
+      "step": 1100288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1977960026546766e-05,
+      "loss": 3.5456,
+      "step": 1100800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1969574079036246e-05,
+      "loss": 3.5365,
+      "step": 1101312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1961188131525726e-05,
+      "loss": 3.56,
+      "step": 1101824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1952802184015206e-05,
+      "loss": 3.546,
+      "step": 1102336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1944432615308415e-05,
+      "loss": 3.5353,
+      "step": 1102848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1936046667797895e-05,
+      "loss": 3.5301,
+      "step": 1103360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1927660720287375e-05,
+      "loss": 3.5428,
+      "step": 1103872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1919274772776855e-05,
+      "loss": 3.5488,
+      "step": 1104384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.191090520407007e-05,
+      "loss": 3.5597,
+      "step": 1104896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.190251925655955e-05,
+      "loss": 3.5436,
+      "step": 1105408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.189413330904903e-05,
+      "loss": 3.5572,
+      "step": 1105920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.188574736153851e-05,
+      "loss": 3.5571,
+      "step": 1106432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.187737779283172e-05,
+      "loss": 3.5475,
+      "step": 1106944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.18689918453212e-05,
+      "loss": 3.5465,
+      "step": 1107456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.186060589781068e-05,
+      "loss": 3.5488,
+      "step": 1107968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185221995030016e-05,
+      "loss": 3.5476,
+      "step": 1108480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.184385038159337e-05,
+      "loss": 3.5405,
+      "step": 1108992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.183546443408285e-05,
+      "loss": 3.5462,
+      "step": 1109504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.182707848657233e-05,
+      "loss": 3.5547,
+      "step": 1110016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.181869253906181e-05,
+      "loss": 3.5573,
+      "step": 1110528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1810322970355025e-05,
+      "loss": 3.5524,
+      "step": 1111040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1801937022844505e-05,
+      "loss": 3.536,
+      "step": 1111552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1793551075333985e-05,
+      "loss": 3.547,
+      "step": 1112064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1785165127823465e-05,
+      "loss": 3.5506,
+      "step": 1112576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1776795559116674e-05,
+      "loss": 3.5443,
+      "step": 1113088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1768409611606154e-05,
+      "loss": 3.5425,
+      "step": 1113600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1760023664095634e-05,
+      "loss": 3.5465,
+      "step": 1114112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1751637716585114e-05,
+      "loss": 3.5486,
+      "step": 1114624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.174326814787832e-05,
+      "loss": 3.5406,
+      "step": 1115136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.17348822003678e-05,
+      "loss": 3.5457,
+      "step": 1115648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.172649625285728e-05,
+      "loss": 3.5479,
+      "step": 1116160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.171811030534676e-05,
+      "loss": 3.551,
+      "step": 1116672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170974073663998e-05,
+      "loss": 3.5552,
+      "step": 1117184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170135478912946e-05,
+      "loss": 3.5326,
+      "step": 1117696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.169296884161894e-05,
+      "loss": 3.547,
+      "step": 1118208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.168458289410842e-05,
+      "loss": 3.5472,
+      "step": 1118720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.167621332540163e-05,
+      "loss": 3.5429,
+      "step": 1119232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.166782737789111e-05,
+      "loss": 3.5611,
+      "step": 1119744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.165944143038059e-05,
+      "loss": 3.5437,
+      "step": 1120256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.165105548287007e-05,
+      "loss": 3.5345,
+      "step": 1120768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1642685914163276e-05,
+      "loss": 3.5422,
+      "step": 1121280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1634299966652756e-05,
+      "loss": 3.5422,
+      "step": 1121792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1625914019142236e-05,
+      "loss": 3.532,
+      "step": 1122304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1617528071631716e-05,
+      "loss": 3.5433,
+      "step": 1122816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.160915850292493e-05,
+      "loss": 3.5417,
+      "step": 1123328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.160077255541441e-05,
+      "loss": 3.539,
+      "step": 1123840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.159238660790389e-05,
+      "loss": 3.5333,
+      "step": 1124352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.158400066039337e-05,
+      "loss": 3.546,
+      "step": 1124864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.157563109168658e-05,
+      "loss": 3.532,
+      "step": 1125376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.156724514417606e-05,
+      "loss": 3.5554,
+      "step": 1125888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.155885919666554e-05,
+      "loss": 3.5408,
+      "step": 1126400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.155047324915502e-05,
+      "loss": 3.549,
+      "step": 1126912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154210368044823e-05,
+      "loss": 3.5512,
+      "step": 1127424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.153371773293771e-05,
+      "loss": 3.567,
+      "step": 1127936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.152533178542719e-05,
+      "loss": 3.5419,
+      "step": 1128448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.151694583791667e-05,
+      "loss": 3.5422,
+      "step": 1128960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1508576269209886e-05,
+      "loss": 3.5487,
+      "step": 1129472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1500190321699366e-05,
+      "loss": 3.5379,
+      "step": 1129984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1491804374188846e-05,
+      "loss": 3.5442,
+      "step": 1130496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1483418426678326e-05,
+      "loss": 3.5457,
+      "step": 1131008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1475048857971535e-05,
+      "loss": 3.5436,
+      "step": 1131520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1466662910461015e-05,
+      "loss": 3.5531,
+      "step": 1132032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1458276962950495e-05,
+      "loss": 3.542,
+      "step": 1132544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.144989101543997e-05,
+      "loss": 3.5386,
+      "step": 1133056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1441521446733184e-05,
+      "loss": 3.5441,
+      "step": 1133568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1433135499222664e-05,
+      "loss": 3.5497,
+      "step": 1134080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1424749551712144e-05,
+      "loss": 3.5428,
+      "step": 1134592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1416363604201624e-05,
+      "loss": 3.5477,
+      "step": 1135104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.140799403549484e-05,
+      "loss": 3.5491,
+      "step": 1135616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.139960808798432e-05,
+      "loss": 3.5367,
+      "step": 1136128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.13912221404738e-05,
+      "loss": 3.5423,
+      "step": 1136640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.138283619296327e-05,
+      "loss": 3.5484,
+      "step": 1137152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.137446662425649e-05,
+      "loss": 3.5465,
+      "step": 1137664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.136608067674597e-05,
+      "loss": 3.5403,
+      "step": 1138176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.135769472923544e-05,
+      "loss": 3.5459,
+      "step": 1138688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.134930878172492e-05,
+      "loss": 3.5496,
+      "step": 1139200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.134093921301814e-05,
+      "loss": 3.5525,
+      "step": 1139712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133255326550762e-05,
+      "loss": 3.5444,
+      "step": 1140224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.13241673179971e-05,
+      "loss": 3.5415,
+      "step": 1140736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.131578137048658e-05,
+      "loss": 3.5393,
+      "step": 1141248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.130741180177979e-05,
+      "loss": 3.5454,
+      "step": 1141760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.129902585426927e-05,
+      "loss": 3.5347,
+      "step": 1142272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1290639906758746e-05,
+      "loss": 3.5566,
+      "step": 1142784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1282253959248226e-05,
+      "loss": 3.5433,
+      "step": 1143296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.127388439054144e-05,
+      "loss": 3.5494,
+      "step": 1143808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1265498443030915e-05,
+      "loss": 3.5374,
+      "step": 1144320
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.840376377105713,
+      "eval_runtime": 305.3163,
+      "eval_samples_per_second": 1249.822,
+      "eval_steps_per_second": 39.058,
+      "step": 1144800
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 8.023106617176269e+17,
+  "trial_name": null,
+  "trial_params": null
+}