diff --git "a/checkpoint-457914/trainer_state.json" "b/checkpoint-457914/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-457914/trainer_state.json"
@@ -0,0 +1,5437 @@
+{
+  "best_metric": 3.868739366531372,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full/transformer/4/checkpoints/checkpoint-457914",
+  "epoch": 0.025000278439663435,
+  "eval_steps": 10,
+  "global_step": 457914,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 11.0625,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8438,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.19,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9707,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8179,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.6964,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.5965,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5336,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4804,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4012,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3622,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3225,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.2804,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2196,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1926,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1604,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1211,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0868,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0669,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.039,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0085,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9812,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9682,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9393,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.929,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.9023,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8901,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8712,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8371,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756873037409844e-05,
+      "loss": 4.8388,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.8094,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7989,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7874,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7796,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971497605746471e-05,
+      "loss": 4.7549,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970659010995419e-05,
+      "loss": 4.7462,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969820416244367e-05,
+      "loss": 4.7438,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.7264,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7184,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967306269871584e-05,
+      "loss": 4.6998,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966467675120532e-05,
+      "loss": 4.6839,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96562908036948e-05,
+      "loss": 4.6761,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964790485618428e-05,
+      "loss": 4.6541,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963951890867376e-05,
+      "loss": 4.6712,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 4.6455,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6454,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6347,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9605991497435414e-05,
+      "loss": 4.6319,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9597605549924894e-05,
+      "loss": 4.622,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.5979,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.5904,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957248046500079e-05,
+      "loss": 4.583,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956409451749027e-05,
+      "loss": 4.5902,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955570856997975e-05,
+      "loss": 4.5683,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 4.579,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 4.5542,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 4.5466,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5431,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5379,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5345,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949702331620985e-05,
+      "loss": 4.5171,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948863736869932e-05,
+      "loss": 4.5127,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5203,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.5014,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946349590497149e-05,
+      "loss": 4.4957,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945510995746097e-05,
+      "loss": 4.4955,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.4943,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4729,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9429968493733145e-05,
+      "loss": 4.4844,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9421582546222625e-05,
+      "loss": 4.4737,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4645,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4718,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396441082494794e-05,
+      "loss": 4.4555,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938807151378801e-05,
+      "loss": 4.4685,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937968556627749e-05,
+      "loss": 4.469,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937129961876696e-05,
+      "loss": 4.4589,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 4.4418,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 4.4312,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934615815503914e-05,
+      "loss": 4.4218,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933777220752861e-05,
+      "loss": 4.4346,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93293862600181e-05,
+      "loss": 4.4345,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932100031250758e-05,
+      "loss": 4.4327,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931261436499706e-05,
+      "loss": 4.405,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930422841748654e-05,
+      "loss": 4.405,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.4099,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928747290126923e-05,
+      "loss": 4.3979,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927908695375871e-05,
+      "loss": 4.4033,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927070100624819e-05,
+      "loss": 4.3995,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926231505873767e-05,
+      "loss": 4.4,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925394549003088e-05,
+      "loss": 4.3893,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924555954252036e-05,
+      "loss": 4.3855,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923717359500984e-05,
+      "loss": 4.3747,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922878764749932e-05,
+      "loss": 4.3895,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922041807879253e-05,
+      "loss": 4.3722,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921203213128201e-05,
+      "loss": 4.3577,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920364618377149e-05,
+      "loss": 4.3631,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919526023626097e-05,
+      "loss": 4.3764,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918687428875045e-05,
+      "loss": 4.3688,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917850472004366e-05,
+      "loss": 4.3671,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917011877253314e-05,
+      "loss": 4.3471,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916173282502262e-05,
+      "loss": 4.3629,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91533468775121e-05,
+      "loss": 4.3361,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914496093000158e-05,
+      "loss": 4.3532,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913657498249106e-05,
+      "loss": 4.3554,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912820541378427e-05,
+      "loss": 4.3411,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911981946627375e-05,
+      "loss": 4.3353,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911143351876323e-05,
+      "loss": 4.3109,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.3252,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3236,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.3224,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9077906107524886e-05,
+      "loss": 4.3276,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9069520160014366e-05,
+      "loss": 4.3088,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9061134212503846e-05,
+      "loss": 4.3118,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052748264993326e-05,
+      "loss": 4.3183,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90443623174828e-05,
+      "loss": 4.3196,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.3069,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027606801265495e-05,
+      "loss": 4.3064,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.3005,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.3018,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.305,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899407939002715e-05,
+      "loss": 4.2918,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985693442516624e-05,
+      "loss": 4.2879,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977307495006104e-05,
+      "loss": 4.2838,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968921547495584e-05,
+      "loss": 4.2811,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.2874,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2843,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2902,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2668,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.2798,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2814,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2691,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2865,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889348077750837e-05,
+      "loss": 4.268,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888509482999785e-05,
+      "loss": 4.2792,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887670888248733e-05,
+      "loss": 4.2694,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886832293497681e-05,
+      "loss": 4.2477,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885995336627002e-05,
+      "loss": 4.2505,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515674187595e-05,
+      "loss": 4.2539,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8843197850052706e-05,
+      "loss": 4.2607,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834811902542186e-05,
+      "loss": 4.2614,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8826425955031666e-05,
+      "loss": 4.2615,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8818040007521146e-05,
+      "loss": 4.2509,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8809654060010626e-05,
+      "loss": 4.2453,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8801268112500106e-05,
+      "loss": 4.2399,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8792882164989586e-05,
+      "loss": 4.2349,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2444,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877612664877228e-05,
+      "loss": 4.2537,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876775708006549e-05,
+      "loss": 4.2458,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875937113255497e-05,
+      "loss": 4.2263,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875098518504445e-05,
+      "loss": 4.2293,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.1957807540893555,
+      "eval_runtime": 622.6197,
+      "eval_samples_per_second": 612.88,
+      "eval_steps_per_second": 19.153,
+      "step": 76319
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.874259923753393e-05,
+      "loss": 4.232,
+      "step": 76800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.873421329002341e-05,
+      "loss": 4.2268,
+      "step": 77312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.872582734251289e-05,
+      "loss": 4.2213,
+      "step": 77824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.871744139500237e-05,
+      "loss": 4.2219,
+      "step": 78336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870905544749185e-05,
+      "loss": 4.2131,
+      "step": 78848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870066949998133e-05,
+      "loss": 4.1968,
+      "step": 79360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.869228355247081e-05,
+      "loss": 4.2042,
+      "step": 79872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.868389760496029e-05,
+      "loss": 4.2168,
+      "step": 80384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.867551165744977e-05,
+      "loss": 4.2047,
+      "step": 80896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8667142088742986e-05,
+      "loss": 4.2081,
+      "step": 81408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865875614123246e-05,
+      "loss": 4.2105,
+      "step": 81920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865037019372194e-05,
+      "loss": 4.2093,
+      "step": 82432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.864198424621142e-05,
+      "loss": 4.1997,
+      "step": 82944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.86335982987009e-05,
+      "loss": 4.2022,
+      "step": 83456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.862521235119038e-05,
+      "loss": 4.1959,
+      "step": 83968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.1815,
+      "step": 84480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860844045616934e-05,
+      "loss": 4.1899,
+      "step": 84992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860005450865882e-05,
+      "loss": 4.194,
+      "step": 85504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.85916685611483e-05,
+      "loss": 4.1926,
+      "step": 86016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.1942,
+      "step": 86528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.1794,
+      "step": 87040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.191,
+      "step": 87552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.1757,
+      "step": 88064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8549771581203164e-05,
+      "loss": 4.1871,
+      "step": 88576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.1779,
+      "step": 89088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8532999686182124e-05,
+      "loss": 4.1756,
+      "step": 89600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8524613738671604e-05,
+      "loss": 4.1801,
+      "step": 90112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8516227791161084e-05,
+      "loss": 4.1611,
+      "step": 90624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.1765,
+      "step": 91136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.1609,
+      "step": 91648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.1653,
+      "step": 92160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.1668,
+      "step": 92672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.847433081121595e-05,
+      "loss": 4.1709,
+      "step": 93184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.846594486370543e-05,
+      "loss": 4.1598,
+      "step": 93696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.845755891619491e-05,
+      "loss": 4.1543,
+      "step": 94208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.844917296868439e-05,
+      "loss": 4.1683,
+      "step": 94720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.84408033999776e-05,
+      "loss": 4.1573,
+      "step": 95232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.1646,
+      "step": 95744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.842403150495656e-05,
+      "loss": 4.149,
+      "step": 96256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.841564555744604e-05,
+      "loss": 4.1394,
+      "step": 96768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.840725960993552e-05,
+      "loss": 4.1427,
+      "step": 97280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839889004122873e-05,
+      "loss": 4.1387,
+      "step": 97792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839050409371821e-05,
+      "loss": 4.1534,
+      "step": 98304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8382118146207687e-05,
+      "loss": 4.1413,
+      "step": 98816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8373732198697167e-05,
+      "loss": 4.1469,
+      "step": 99328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8365346251186646e-05,
+      "loss": 4.1412,
+      "step": 99840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.835697668247986e-05,
+      "loss": 4.1466,
+      "step": 100352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834859073496934e-05,
+      "loss": 4.1437,
+      "step": 100864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834020478745882e-05,
+      "loss": 4.1199,
+      "step": 101376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8331818839948295e-05,
+      "loss": 4.1272,
+      "step": 101888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.832344927124151e-05,
+      "loss": 4.1237,
+      "step": 102400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.831506332373099e-05,
+      "loss": 4.1395,
+      "step": 102912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.830667737622047e-05,
+      "loss": 4.1172,
+      "step": 103424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8298291428709944e-05,
+      "loss": 4.1395,
+      "step": 103936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828992186000316e-05,
+      "loss": 4.1154,
+      "step": 104448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828153591249264e-05,
+      "loss": 4.1135,
+      "step": 104960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.827314996498212e-05,
+      "loss": 4.1231,
+      "step": 105472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.82647640174716e-05,
+      "loss": 4.1178,
+      "step": 105984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.825637806996108e-05,
+      "loss": 4.1172,
+      "step": 106496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8248008501254296e-05,
+      "loss": 4.1104,
+      "step": 107008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.823962255374377e-05,
+      "loss": 4.1057,
+      "step": 107520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.823123660623325e-05,
+      "loss": 4.1201,
+      "step": 108032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.822285065872273e-05,
+      "loss": 4.1074,
+      "step": 108544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8214481090015945e-05,
+      "loss": 4.1061,
+      "step": 109056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.820609514250542e-05,
+      "loss": 4.105,
+      "step": 109568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.81977091949949e-05,
+      "loss": 4.1106,
+      "step": 110080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818932324748438e-05,
+      "loss": 4.0907,
+      "step": 110592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8180953678777594e-05,
+      "loss": 4.1112,
+      "step": 111104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8172567731267074e-05,
+      "loss": 4.0985,
+      "step": 111616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8164181783756554e-05,
+      "loss": 4.0997,
+      "step": 112128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8155795836246034e-05,
+      "loss": 4.1061,
+      "step": 112640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.814742626753924e-05,
+      "loss": 4.0959,
+      "step": 113152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813904032002872e-05,
+      "loss": 4.1145,
+      "step": 113664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.81306543725182e-05,
+      "loss": 4.1171,
+      "step": 114176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.812226842500768e-05,
+      "loss": 4.1107,
+      "step": 114688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.811389885630089e-05,
+      "loss": 4.0947,
+      "step": 115200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.810551290879037e-05,
+      "loss": 4.0886,
+      "step": 115712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.809714334008359e-05,
+      "loss": 4.0839,
+      "step": 116224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808875739257306e-05,
+      "loss": 4.0951,
+      "step": 116736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808037144506255e-05,
+      "loss": 4.109,
+      "step": 117248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.807198549755203e-05,
+      "loss": 4.0976,
+      "step": 117760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.806359955004151e-05,
+      "loss": 4.0828,
+      "step": 118272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.805521360253099e-05,
+      "loss": 4.0772,
+      "step": 118784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.804682765502047e-05,
+      "loss": 4.0903,
+      "step": 119296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.803844170750995e-05,
+      "loss": 4.0786,
+      "step": 119808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8030072138803156e-05,
+      "loss": 4.0844,
+      "step": 120320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8021686191292636e-05,
+      "loss": 4.0859,
+      "step": 120832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8013300243782116e-05,
+      "loss": 4.0904,
+      "step": 121344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8004914296271596e-05,
+      "loss": 4.081,
+      "step": 121856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7996561106368535e-05,
+      "loss": 4.0798,
+      "step": 122368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7988175158858015e-05,
+      "loss": 4.0729,
+      "step": 122880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7979789211347494e-05,
+      "loss": 4.0835,
+      "step": 123392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797140326383698e-05,
+      "loss": 4.0806,
+      "step": 123904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.796301731632646e-05,
+      "loss": 4.0614,
+      "step": 124416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.795463136881594e-05,
+      "loss": 4.0723,
+      "step": 124928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.794624542130542e-05,
+      "loss": 4.0813,
+      "step": 125440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.79378594737949e-05,
+      "loss": 4.0875,
+      "step": 125952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792948990508811e-05,
+      "loss": 4.076,
+      "step": 126464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792110395757759e-05,
+      "loss": 4.0674,
+      "step": 126976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.791271801006707e-05,
+      "loss": 4.0799,
+      "step": 127488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.790433206255655e-05,
+      "loss": 4.0567,
+      "step": 128000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.789596249384976e-05,
+      "loss": 4.0724,
+      "step": 128512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.788757654633924e-05,
+      "loss": 4.0801,
+      "step": 129024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.787920697763245e-05,
+      "loss": 4.0702,
+      "step": 129536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7870821030121935e-05,
+      "loss": 4.0703,
+      "step": 130048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7862435082611415e-05,
+      "loss": 4.0482,
+      "step": 130560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7854049135100895e-05,
+      "loss": 4.0531,
+      "step": 131072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7845663187590375e-05,
+      "loss": 4.0572,
+      "step": 131584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7837277240079855e-05,
+      "loss": 4.0579,
+      "step": 132096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7828891292569335e-05,
+      "loss": 4.0657,
+      "step": 132608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7820505345058815e-05,
+      "loss": 4.0537,
+      "step": 133120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7812135776352024e-05,
+      "loss": 4.0492,
+      "step": 133632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7803749828841504e-05,
+      "loss": 4.0607,
+      "step": 134144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7795363881330984e-05,
+      "loss": 4.0674,
+      "step": 134656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7786977933820464e-05,
+      "loss": 4.0529,
+      "step": 135168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777860836511367e-05,
+      "loss": 4.0605,
+      "step": 135680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777022241760315e-05,
+      "loss": 4.0505,
+      "step": 136192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.776183647009263e-05,
+      "loss": 4.0533,
+      "step": 136704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.775345052258212e-05,
+      "loss": 4.0583,
+      "step": 137216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.774509733267906e-05,
+      "loss": 4.046,
+      "step": 137728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.773671138516854e-05,
+      "loss": 4.0456,
+      "step": 138240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.772832543765802e-05,
+      "loss": 4.0442,
+      "step": 138752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77199394901475e-05,
+      "loss": 4.0423,
+      "step": 139264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771155354263698e-05,
+      "loss": 4.0476,
+      "step": 139776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.770316759512646e-05,
+      "loss": 4.05,
+      "step": 140288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.769478164761594e-05,
+      "loss": 4.0542,
+      "step": 140800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.768639570010542e-05,
+      "loss": 4.0354,
+      "step": 141312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7678026131398626e-05,
+      "loss": 4.0421,
+      "step": 141824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7669640183888106e-05,
+      "loss": 4.0555,
+      "step": 142336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7661254236377586e-05,
+      "loss": 4.0397,
+      "step": 142848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.765286828886707e-05,
+      "loss": 4.0539,
+      "step": 143360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.764449872016028e-05,
+      "loss": 4.0402,
+      "step": 143872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.763611277264976e-05,
+      "loss": 4.0558,
+      "step": 144384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.762772682513924e-05,
+      "loss": 4.0446,
+      "step": 144896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761934087762872e-05,
+      "loss": 4.0286,
+      "step": 145408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761097130892193e-05,
+      "loss": 4.0282,
+      "step": 145920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.760258536141141e-05,
+      "loss": 4.032,
+      "step": 146432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.759419941390089e-05,
+      "loss": 4.0449,
+      "step": 146944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.758581346639037e-05,
+      "loss": 4.0412,
+      "step": 147456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.757744389768358e-05,
+      "loss": 4.0451,
+      "step": 147968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7569074328976796e-05,
+      "loss": 4.0366,
+      "step": 148480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7560688381466276e-05,
+      "loss": 4.0289,
+      "step": 148992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7552302433955756e-05,
+      "loss": 4.0287,
+      "step": 149504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7543916486445236e-05,
+      "loss": 4.0237,
+      "step": 150016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7535530538934716e-05,
+      "loss": 4.0329,
+      "step": 150528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7527144591424196e-05,
+      "loss": 4.0465,
+      "step": 151040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751875864391367e-05,
+      "loss": 4.0318,
+      "step": 151552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751037269640315e-05,
+      "loss": 4.0258,
+      "step": 152064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7502003127696365e-05,
+      "loss": 4.0241,
+      "step": 152576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.025460720062256,
+      "eval_runtime": 588.8804,
+      "eval_samples_per_second": 647.994,
+      "eval_steps_per_second": 20.25,
+      "step": 152638
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493633558989574e-05,
+      "loss": 4.0321,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485247611479054e-05,
+      "loss": 4.0269,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476861663968534e-05,
+      "loss": 4.0179,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468475716458014e-05,
+      "loss": 4.0223,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460089768947494e-05,
+      "loss": 4.0194,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451703821436974e-05,
+      "loss": 3.9993,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443317873926453e-05,
+      "loss": 4.0086,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434931926415933e-05,
+      "loss": 4.0242,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426545978905413e-05,
+      "loss": 4.0067,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741817641019862e-05,
+      "loss": 4.0143,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74097904626881e-05,
+      "loss": 4.0201,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740140451517758e-05,
+      "loss": 4.0178,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739303494647079e-05,
+      "loss": 4.0068,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738464899896027e-05,
+      "loss": 4.0138,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737626305144975e-05,
+      "loss": 4.0061,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736787710393923e-05,
+      "loss": 3.9991,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735949115642871e-05,
+      "loss": 4.0034,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73511052089182e-05,
+      "loss": 4.0082,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734271926140768e-05,
+      "loss": 4.0109,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733433331389716e-05,
+      "loss": 4.0134,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732594736638664e-05,
+      "loss": 3.9979,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731756141887612e-05,
+      "loss": 4.0119,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73091754713656e-05,
+      "loss": 3.9996,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730080590265881e-05,
+      "loss": 4.0055,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729241995514829e-05,
+      "loss": 4.0014,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728403400763777e-05,
+      "loss": 3.9995,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727564806012725e-05,
+      "loss": 4.0071,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726726211261673e-05,
+      "loss": 3.9887,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725887616510621e-05,
+      "loss": 4.0015,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725049021759568e-05,
+      "loss": 3.9891,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724210427008517e-05,
+      "loss": 3.9933,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723373470137838e-05,
+      "loss": 3.996,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722536513267159e-05,
+      "loss": 3.9996,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721697918516107e-05,
+      "loss": 3.9942,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720859323765055e-05,
+      "loss": 3.9859,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720020729014003e-05,
+      "loss": 4.0009,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191821342629505e-05,
+      "loss": 3.9919,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183435395118985e-05,
+      "loss": 4.0014,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175049447608465e-05,
+      "loss": 3.9866,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7166663500097945e-05,
+      "loss": 3.9749,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158293931391154e-05,
+      "loss": 3.9776,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714992436268437e-05,
+      "loss": 3.9817,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714153841517385e-05,
+      "loss": 3.9964,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713315246766333e-05,
+      "loss": 3.9758,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712476652015281e-05,
+      "loss": 3.9894,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711638057264229e-05,
+      "loss": 3.9841,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7108011003935505e-05,
+      "loss": 3.9848,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709962505642498e-05,
+      "loss": 3.9902,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709123910891446e-05,
+      "loss": 3.9649,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708285316140394e-05,
+      "loss": 3.9707,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074483592697154e-05,
+      "loss": 3.966,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706609764518663e-05,
+      "loss": 3.9875,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705771169767611e-05,
+      "loss": 3.9669,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704932575016559e-05,
+      "loss": 3.9867,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70409561814588e-05,
+      "loss": 3.9631,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703257023394828e-05,
+      "loss": 3.9636,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702418428643776e-05,
+      "loss": 3.9698,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701579833892724e-05,
+      "loss": 3.964,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700742877022045e-05,
+      "loss": 3.9708,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699904282270993e-05,
+      "loss": 3.9663,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699065687519941e-05,
+      "loss": 3.9542,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698227092768889e-05,
+      "loss": 3.971,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697388498017837e-05,
+      "loss": 3.9607,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696549903266785e-05,
+      "loss": 3.9651,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695711308515733e-05,
+      "loss": 3.9583,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694872713764681e-05,
+      "loss": 3.9677,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694035756894003e-05,
+      "loss": 3.9448,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693198800023324e-05,
+      "loss": 3.9681,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692360205272272e-05,
+      "loss": 3.9535,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69152161052122e-05,
+      "loss": 3.962,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906830157701677e-05,
+      "loss": 3.9618,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898444210191157e-05,
+      "loss": 3.9554,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6890074641484366e-05,
+      "loss": 3.9716,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881688693973846e-05,
+      "loss": 3.9778,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873302746463326e-05,
+      "loss": 3.9727,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864916798952806e-05,
+      "loss": 3.9615,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856547230246015e-05,
+      "loss": 3.9505,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848161282735495e-05,
+      "loss": 3.9429,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683977533522498e-05,
+      "loss": 3.9613,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683138938771446e-05,
+      "loss": 3.9694,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682301981900767e-05,
+      "loss": 3.9632,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681463387149715e-05,
+      "loss": 3.9478,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680624792398663e-05,
+      "loss": 3.9426,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679786197647611e-05,
+      "loss": 3.9527,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678947602896559e-05,
+      "loss": 3.9415,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67811064602588e-05,
+      "loss": 3.9484,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677272051274828e-05,
+      "loss": 3.9545,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676433456523776e-05,
+      "loss": 3.9556,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675594861772724e-05,
+      "loss": 3.951,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674757904902045e-05,
+      "loss": 3.9533,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739193101509935e-05,
+      "loss": 3.9419,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730807153999415e-05,
+      "loss": 3.9539,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722421206488895e-05,
+      "loss": 3.9478,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6714035258978375e-05,
+      "loss": 3.9376,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705665690271584e-05,
+      "loss": 3.9417,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697279742761064e-05,
+      "loss": 3.952,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688893795250544e-05,
+      "loss": 3.9605,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680507847740024e-05,
+      "loss": 3.9463,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6672121900229504e-05,
+      "loss": 3.9441,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666375233152271e-05,
+      "loss": 3.9504,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665536638401219e-05,
+      "loss": 3.9319,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664698043650167e-05,
+      "loss": 3.9498,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663859448899115e-05,
+      "loss": 3.9527,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663022492028437e-05,
+      "loss": 3.9429,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662183897277385e-05,
+      "loss": 3.9509,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661345302526333e-05,
+      "loss": 3.9271,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660506707775281e-05,
+      "loss": 3.9255,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659669750904602e-05,
+      "loss": 3.9342,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65883115615355e-05,
+      "loss": 3.9357,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657992561402498e-05,
+      "loss": 3.9467,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657153966651446e-05,
+      "loss": 3.9312,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563170097807667e-05,
+      "loss": 3.9274,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554784150297146e-05,
+      "loss": 3.9351,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546398202786626e-05,
+      "loss": 3.947,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6538012255276106e-05,
+      "loss": 3.9371,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529626307765586e-05,
+      "loss": 3.9456,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65212567390588e-05,
+      "loss": 3.9301,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651287079154828e-05,
+      "loss": 3.9306,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650448484403776e-05,
+      "loss": 3.9443,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6496098896527235e-05,
+      "loss": 3.9284,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648772932782045e-05,
+      "loss": 3.928,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647934338030993e-05,
+      "loss": 3.928,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470957432799404e-05,
+      "loss": 3.9244,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462571485288884e-05,
+      "loss": 3.9327,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64542019165821e-05,
+      "loss": 3.9315,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644581596907158e-05,
+      "loss": 3.9404,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643743002156106e-05,
+      "loss": 3.9211,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642904407405054e-05,
+      "loss": 3.9244,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642065812654002e-05,
+      "loss": 3.9456,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412288557833236e-05,
+      "loss": 3.9236,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640390261032271e-05,
+      "loss": 3.9404,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639551666281219e-05,
+      "loss": 3.928,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638713071530167e-05,
+      "loss": 3.9409,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637876114659488e-05,
+      "loss": 3.9318,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637037519908436e-05,
+      "loss": 3.9184,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636198925157384e-05,
+      "loss": 3.9172,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635360330406332e-05,
+      "loss": 3.9191,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634523373535653e-05,
+      "loss": 3.9308,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336847787846014e-05,
+      "loss": 3.9364,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328461840335494e-05,
+      "loss": 3.932,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6320075892824974e-05,
+      "loss": 3.9266,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631170632411818e-05,
+      "loss": 3.9182,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630332037660766e-05,
+      "loss": 3.9199,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629493442909714e-05,
+      "loss": 3.9134,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628654848158662e-05,
+      "loss": 3.9251,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62781625340761e-05,
+      "loss": 3.933,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626979296536931e-05,
+      "loss": 3.9224,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626140701785879e-05,
+      "loss": 3.9198,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625302107034827e-05,
+      "loss": 3.9167,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.950981855392456,
+      "eval_runtime": 570.4475,
+      "eval_samples_per_second": 668.933,
+      "eval_steps_per_second": 20.905,
+      "step": 228957
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.624463512283775e-05,
+      "loss": 3.922,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.623624917532723e-05,
+      "loss": 3.9261,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.622786322781671e-05,
+      "loss": 3.9129,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.62194772803062e-05,
+      "loss": 3.9125,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621109133279568e-05,
+      "loss": 3.9154,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.620270538528516e-05,
+      "loss": 3.8958,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.619431943777464e-05,
+      "loss": 3.9039,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.618593349026412e-05,
+      "loss": 3.917,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.617756392155733e-05,
+      "loss": 3.9061,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6169194352850536e-05,
+      "loss": 3.913,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6160808405340016e-05,
+      "loss": 3.9137,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6152422457829496e-05,
+      "loss": 3.9117,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6144036510318976e-05,
+      "loss": 3.9054,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6135650562808456e-05,
+      "loss": 3.9151,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6127264615297936e-05,
+      "loss": 3.9028,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6118878667787416e-05,
+      "loss": 3.9001,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6110492720276896e-05,
+      "loss": 3.8998,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.610212315157011e-05,
+      "loss": 3.9074,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.609373720405959e-05,
+      "loss": 3.9084,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6085351256549065e-05,
+      "loss": 3.9138,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6076965309038545e-05,
+      "loss": 3.8993,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606859574033176e-05,
+      "loss": 3.9119,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606020979282124e-05,
+      "loss": 3.9023,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6051823845310714e-05,
+      "loss": 3.9022,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.604345427660393e-05,
+      "loss": 3.9045,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.603506832909341e-05,
+      "loss": 3.9008,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.602668238158289e-05,
+      "loss": 3.9079,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.601829643407237e-05,
+      "loss": 3.8921,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.600991048656185e-05,
+      "loss": 3.9049,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600152453905133e-05,
+      "loss": 3.8927,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599313859154081e-05,
+      "loss": 3.8942,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598475264403029e-05,
+      "loss": 3.8996,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.59763830753235e-05,
+      "loss": 3.9062,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.596799712781298e-05,
+      "loss": 3.8943,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595961118030246e-05,
+      "loss": 3.8957,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595122523279194e-05,
+      "loss": 3.8999,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.594285566408515e-05,
+      "loss": 3.8962,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.593446971657463e-05,
+      "loss": 3.9104,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.592608376906411e-05,
+      "loss": 3.8925,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.591769782155359e-05,
+      "loss": 3.88,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.59093282528468e-05,
+      "loss": 3.8868,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590094230533628e-05,
+      "loss": 3.8825,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.589255635782576e-05,
+      "loss": 3.9028,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.588417041031524e-05,
+      "loss": 3.886,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.587580084160845e-05,
+      "loss": 3.8951,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.586741489409793e-05,
+      "loss": 3.8893,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585902894658741e-05,
+      "loss": 3.8932,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585064299907689e-05,
+      "loss": 3.8969,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58422734303701e-05,
+      "loss": 3.8766,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.583388748285958e-05,
+      "loss": 3.8738,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.582550153534906e-05,
+      "loss": 3.8813,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.581711558783854e-05,
+      "loss": 3.8953,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580874601913176e-05,
+      "loss": 3.876,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580036007162124e-05,
+      "loss": 3.8933,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.579197412411072e-05,
+      "loss": 3.8734,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5783604555403926e-05,
+      "loss": 3.8709,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5775218607893406e-05,
+      "loss": 3.8834,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5766832660382886e-05,
+      "loss": 3.8716,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5758446712872366e-05,
+      "loss": 3.8882,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5750060765361846e-05,
+      "loss": 3.8801,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5741674817851326e-05,
+      "loss": 3.8652,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5733288870340806e-05,
+      "loss": 3.88,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5724919301634015e-05,
+      "loss": 3.8717,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5716533354123495e-05,
+      "loss": 3.8797,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5708147406612975e-05,
+      "loss": 3.8679,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569976145910246e-05,
+      "loss": 3.8817,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569139189039567e-05,
+      "loss": 3.8612,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.568300594288515e-05,
+      "loss": 3.8789,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.567461999537463e-05,
+      "loss": 3.8655,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.566623404786411e-05,
+      "loss": 3.8725,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.565784810035359e-05,
+      "loss": 3.8767,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.56494785316468e-05,
+      "loss": 3.8737,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564109258413628e-05,
+      "loss": 3.8791,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.563270663662576e-05,
+      "loss": 3.8915,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.562433706791897e-05,
+      "loss": 3.8858,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561595112040845e-05,
+      "loss": 3.875,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560756517289793e-05,
+      "loss": 3.868,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5599179225387415e-05,
+      "loss": 3.8605,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5590809656680624e-05,
+      "loss": 3.8755,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5582423709170104e-05,
+      "loss": 3.8855,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5574037761659584e-05,
+      "loss": 3.8809,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5565651814149064e-05,
+      "loss": 3.8619,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.555728224544227e-05,
+      "loss": 3.8596,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554889629793175e-05,
+      "loss": 3.8742,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554051035042123e-05,
+      "loss": 3.8539,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.553212440291071e-05,
+      "loss": 3.8681,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.552375483420392e-05,
+      "loss": 3.8691,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.55153688866934e-05,
+      "loss": 3.8716,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.550698293918288e-05,
+      "loss": 3.8701,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549859699167237e-05,
+      "loss": 3.8729,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549022742296558e-05,
+      "loss": 3.8573,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548184147545506e-05,
+      "loss": 3.8682,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.547345552794454e-05,
+      "loss": 3.8656,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546506958043402e-05,
+      "loss": 3.8587,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.545670001172723e-05,
+      "loss": 3.8582,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544831406421671e-05,
+      "loss": 3.8747,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543992811670619e-05,
+      "loss": 3.8734,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5431542169195667e-05,
+      "loss": 3.8702,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5423172600488876e-05,
+      "loss": 3.8617,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5414786652978356e-05,
+      "loss": 3.8708,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5406400705467836e-05,
+      "loss": 3.855,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.539801475795732e-05,
+      "loss": 3.8669,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538964518925053e-05,
+      "loss": 3.8742,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538125924174001e-05,
+      "loss": 3.8618,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.537287329422949e-05,
+      "loss": 3.8762,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.536448734671897e-05,
+      "loss": 3.8425,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.535611777801218e-05,
+      "loss": 3.8488,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.534773183050166e-05,
+      "loss": 3.8573,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533934588299114e-05,
+      "loss": 3.8581,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5330959935480613e-05,
+      "loss": 3.8695,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.532259036677383e-05,
+      "loss": 3.851,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.531420441926331e-05,
+      "loss": 3.8488,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5305834850556525e-05,
+      "loss": 3.8574,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5297448903046005e-05,
+      "loss": 3.8687,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5289062955535485e-05,
+      "loss": 3.8616,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5280677008024965e-05,
+      "loss": 3.867,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5272291060514445e-05,
+      "loss": 3.854,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.526390511300392e-05,
+      "loss": 3.8524,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.52555191654934e-05,
+      "loss": 3.8693,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.524713321798288e-05,
+      "loss": 3.8515,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523876364927609e-05,
+      "loss": 3.8488,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523037770176557e-05,
+      "loss": 3.8538,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.522199175425505e-05,
+      "loss": 3.846,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.521360580674453e-05,
+      "loss": 3.8587,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.520523623803774e-05,
+      "loss": 3.8549,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.519685029052722e-05,
+      "loss": 3.8669,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51884643430167e-05,
+      "loss": 3.8458,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.518009477430992e-05,
+      "loss": 3.8466,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517170882679939e-05,
+      "loss": 3.8728,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.516332287928887e-05,
+      "loss": 3.8497,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515493693177835e-05,
+      "loss": 3.8625,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.514656736307156e-05,
+      "loss": 3.8575,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.513818141556104e-05,
+      "loss": 3.8642,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512979546805052e-05,
+      "loss": 3.8561,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512140952054e-05,
+      "loss": 3.8488,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.511303995183322e-05,
+      "loss": 3.8418,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5104654004322697e-05,
+      "loss": 3.8454,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5096268056812177e-05,
+      "loss": 3.8546,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5087882109301657e-05,
+      "loss": 3.8637,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5079512540594866e-05,
+      "loss": 3.8556,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5071126593084346e-05,
+      "loss": 3.8583,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5062740645573826e-05,
+      "loss": 3.8444,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5054354698063305e-05,
+      "loss": 3.8486,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5045985129356515e-05,
+      "loss": 3.838,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5037599181845995e-05,
+      "loss": 3.8521,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5029213234335474e-05,
+      "loss": 3.8555,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5020827286824954e-05,
+      "loss": 3.8576,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.501245771811817e-05,
+      "loss": 3.8456,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.500407177060765e-05,
+      "loss": 3.8423,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9093804359436035,
+      "eval_runtime": 570.3829,
+      "eval_samples_per_second": 669.008,
+      "eval_steps_per_second": 20.907,
+      "step": 305276
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499570220190086e-05,
+      "loss": 3.8518,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498731625439034e-05,
+      "loss": 3.85,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497893030687982e-05,
+      "loss": 3.843,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49705443593693e-05,
+      "loss": 3.842,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496215841185878e-05,
+      "loss": 3.8437,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495378884315199e-05,
+      "loss": 3.8304,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494540289564147e-05,
+      "loss": 3.8344,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493701694813095e-05,
+      "loss": 3.8411,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492863100062043e-05,
+      "loss": 3.8398,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492026143191364e-05,
+      "loss": 3.8443,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911875484403124e-05,
+      "loss": 3.8425,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903489536892604e-05,
+      "loss": 3.8374,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4895103589382084e-05,
+      "loss": 3.8373,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886717641871564e-05,
+      "loss": 3.8445,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878331694361044e-05,
+      "loss": 3.8353,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869945746850524e-05,
+      "loss": 3.8293,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486157617814373e-05,
+      "loss": 3.8295,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485319023063321e-05,
+      "loss": 3.8417,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484480428312269e-05,
+      "loss": 3.8355,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483641833561217e-05,
+      "loss": 3.8469,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482803238810165e-05,
+      "loss": 3.8297,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481964644059113e-05,
+      "loss": 3.8435,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481126049308061e-05,
+      "loss": 3.8354,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480287454557009e-05,
+      "loss": 3.8349,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479450497686331e-05,
+      "loss": 3.8333,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478611902935279e-05,
+      "loss": 3.8313,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477773308184227e-05,
+      "loss": 3.8414,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476934713433175e-05,
+      "loss": 3.825,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476096118682122e-05,
+      "loss": 3.8355,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47525752393107e-05,
+      "loss": 3.8287,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474420567060392e-05,
+      "loss": 3.8228,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47358197230934e-05,
+      "loss": 3.8344,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472743377558287e-05,
+      "loss": 3.8379,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471904782807235e-05,
+      "loss": 3.8304,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471066188056183e-05,
+      "loss": 3.8273,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470227593305131e-05,
+      "loss": 3.8303,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469388998554079e-05,
+      "loss": 3.832,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468550403803028e-05,
+      "loss": 3.8423,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4677134469323486e-05,
+      "loss": 3.827,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668748521812966e-05,
+      "loss": 3.8135,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660362574302446e-05,
+      "loss": 3.8216,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651976626791926e-05,
+      "loss": 3.8191,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643607058085135e-05,
+      "loss": 3.8337,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4635221110574615e-05,
+      "loss": 3.8236,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626835163064095e-05,
+      "loss": 3.8331,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618449215553575e-05,
+      "loss": 3.8254,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4610079646846784e-05,
+      "loss": 3.8255,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460171007814e-05,
+      "loss": 3.8344,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459332413062948e-05,
+      "loss": 3.8155,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458493818311896e-05,
+      "loss": 3.8078,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457655223560844e-05,
+      "loss": 3.8212,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456818266690165e-05,
+      "loss": 3.826,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455979671939113e-05,
+      "loss": 3.8121,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455141077188061e-05,
+      "loss": 3.829,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454302482437009e-05,
+      "loss": 3.8106,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453463887685957e-05,
+      "loss": 3.8056,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452625292934905e-05,
+      "loss": 3.8197,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451788336064226e-05,
+      "loss": 3.8087,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450949741313174e-05,
+      "loss": 3.8296,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450111146562122e-05,
+      "loss": 3.819,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44927255181107e-05,
+      "loss": 3.7982,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448433957060018e-05,
+      "loss": 3.8152,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475953623089664e-05,
+      "loss": 3.8092,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467567675579144e-05,
+      "loss": 3.8223,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4459181728068624e-05,
+      "loss": 3.8047,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445081215936183e-05,
+      "loss": 3.8163,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444242621185131e-05,
+      "loss": 3.8003,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443404026434079e-05,
+      "loss": 3.8144,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442565431683027e-05,
+      "loss": 3.8059,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441726836931975e-05,
+      "loss": 3.8087,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440888242180923e-05,
+      "loss": 3.8165,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400496474298706e-05,
+      "loss": 3.8091,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439212690559192e-05,
+      "loss": 3.8184,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43837409580814e-05,
+      "loss": 3.8277,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437535501057088e-05,
+      "loss": 3.8258,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436696906306036e-05,
+      "loss": 3.8147,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435859949435358e-05,
+      "loss": 3.8074,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435022992564679e-05,
+      "loss": 3.8002,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434184397813627e-05,
+      "loss": 3.8123,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433345803062575e-05,
+      "loss": 3.8281,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432507208311523e-05,
+      "loss": 3.8165,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316702514408436e-05,
+      "loss": 3.8048,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308316566897916e-05,
+      "loss": 3.7946,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299930619387396e-05,
+      "loss": 3.8176,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291544671876876e-05,
+      "loss": 3.7932,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4283158724366356e-05,
+      "loss": 3.8049,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274772776855836e-05,
+      "loss": 3.8068,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266386829345316e-05,
+      "loss": 3.8097,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4258000881834796e-05,
+      "loss": 3.812,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424964769193174e-05,
+      "loss": 3.8158,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424126174442122e-05,
+      "loss": 3.7994,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42328757969107e-05,
+      "loss": 3.8085,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422448984940018e-05,
+      "loss": 3.8075,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216103901889654e-05,
+      "loss": 3.8005,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207717954379134e-05,
+      "loss": 3.7988,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4199332006868614e-05,
+      "loss": 3.8149,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419096243816182e-05,
+      "loss": 3.8138,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418257649065131e-05,
+      "loss": 3.8145,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417419054314079e-05,
+      "loss": 3.8038,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416580459563027e-05,
+      "loss": 3.8095,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415741864811975e-05,
+      "loss": 3.7967,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414903270060923e-05,
+      "loss": 3.8106,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414064675309871e-05,
+      "loss": 3.8132,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413227718439192e-05,
+      "loss": 3.8044,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41238912368814e-05,
+      "loss": 3.8188,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411550528937088e-05,
+      "loss": 3.7879,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410711934186036e-05,
+      "loss": 3.792,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409874977315357e-05,
+      "loss": 3.7962,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409036382564305e-05,
+      "loss": 3.7997,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408197787813253e-05,
+      "loss": 3.8122,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407359193062201e-05,
+      "loss": 3.7991,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4065205983111494e-05,
+      "loss": 3.7871,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40568364144047e-05,
+      "loss": 3.7987,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404845046689418e-05,
+      "loss": 3.8115,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404006451938366e-05,
+      "loss": 3.8032,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403167857187314e-05,
+      "loss": 3.8119,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402329262436262e-05,
+      "loss": 3.7967,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401492305565583e-05,
+      "loss": 3.7948,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400653710814531e-05,
+      "loss": 3.8117,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399815116063479e-05,
+      "loss": 3.7982,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398976521312427e-05,
+      "loss": 3.7941,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398137926561375e-05,
+      "loss": 3.7939,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397300969690696e-05,
+      "loss": 3.7921,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396462374939645e-05,
+      "loss": 3.8019,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395623780188593e-05,
+      "loss": 3.8006,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394785185437541e-05,
+      "loss": 3.8089,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393946590686489e-05,
+      "loss": 3.7896,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393107995935437e-05,
+      "loss": 3.7932,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392269401184384e-05,
+      "loss": 3.8121,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391430806433332e-05,
+      "loss": 3.7965,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905938495626536e-05,
+      "loss": 3.8046,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897552548116016e-05,
+      "loss": 3.8068,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388916660060549e-05,
+      "loss": 3.8058,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880797031898705e-05,
+      "loss": 3.8031,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872411084388185e-05,
+      "loss": 3.7895,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3864025136877665e-05,
+      "loss": 3.7883,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855639189367145e-05,
+      "loss": 3.794,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3847253241856625e-05,
+      "loss": 3.7996,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838867294346105e-05,
+      "loss": 3.8042,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830481346835585e-05,
+      "loss": 3.8018,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3822111778128794e-05,
+      "loss": 3.8015,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813725830618274e-05,
+      "loss": 3.7922,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805339883107754e-05,
+      "loss": 3.7934,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796953935597234e-05,
+      "loss": 3.7824,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788567988086714e-05,
+      "loss": 3.8016,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3780182040576194e-05,
+      "loss": 3.8023,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771796093065674e-05,
+      "loss": 3.8014,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763410145555154e-05,
+      "loss": 3.787,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375504057684836e-05,
+      "loss": 3.79,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.884523630142212,
+      "eval_runtime": 566.2085,
+      "eval_samples_per_second": 673.941,
+      "eval_steps_per_second": 21.061,
+      "step": 381595
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374667100814158e-05,
+      "loss": 3.8052,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373828506063106e-05,
+      "loss": 3.7939,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372989911312054e-05,
+      "loss": 3.7922,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372151316561002e-05,
+      "loss": 3.786,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.37131272180995e-05,
+      "loss": 3.7883,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370475764939271e-05,
+      "loss": 3.7789,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369637170188219e-05,
+      "loss": 3.7811,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368798575437167e-05,
+      "loss": 3.7823,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367959980686115e-05,
+      "loss": 3.7876,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367123023815436e-05,
+      "loss": 3.7949,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366284429064384e-05,
+      "loss": 3.7851,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365445834313332e-05,
+      "loss": 3.7849,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36460723956228e-05,
+      "loss": 3.7881,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363768644811228e-05,
+      "loss": 3.7911,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362931687940549e-05,
+      "loss": 3.7834,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362093093189497e-05,
+      "loss": 3.7769,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361254498438445e-05,
+      "loss": 3.776,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360415903687393e-05,
+      "loss": 3.7866,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359577308936341e-05,
+      "loss": 3.783,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358738714185289e-05,
+      "loss": 3.7957,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357900119434237e-05,
+      "loss": 3.7801,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3570615246831845e-05,
+      "loss": 3.7897,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356224567812506e-05,
+      "loss": 3.787,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355387610941827e-05,
+      "loss": 3.7838,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354549016190776e-05,
+      "loss": 3.7818,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353710421439724e-05,
+      "loss": 3.7807,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352871826688672e-05,
+      "loss": 3.7893,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35203323193762e-05,
+      "loss": 3.7725,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351194637186568e-05,
+      "loss": 3.7867,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350356042435515e-05,
+      "loss": 3.776,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3495190855648366e-05,
+      "loss": 3.7686,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3486804908137846e-05,
+      "loss": 3.7851,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347841896062732e-05,
+      "loss": 3.7843,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34700330131168e-05,
+      "loss": 3.7836,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3461663444410015e-05,
+      "loss": 3.7787,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3453277496899495e-05,
+      "loss": 3.7769,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3444891549388975e-05,
+      "loss": 3.7801,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3436505601878455e-05,
+      "loss": 3.7947,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3428119654367935e-05,
+      "loss": 3.7772,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3419733706857415e-05,
+      "loss": 3.7641,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3411347759346895e-05,
+      "loss": 3.7722,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402978190640104e-05,
+      "loss": 3.7682,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3394592243129584e-05,
+      "loss": 3.7827,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3386206295619064e-05,
+      "loss": 3.7749,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377820348108544e-05,
+      "loss": 3.7787,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3369434400598024e-05,
+      "loss": 3.7788,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3361048453087503e-05,
+      "loss": 3.7748,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335267888438071e-05,
+      "loss": 3.7852,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334429293687019e-05,
+      "loss": 3.7653,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333590698935968e-05,
+      "loss": 3.7561,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332752104184916e-05,
+      "loss": 3.7746,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331915147314237e-05,
+      "loss": 3.7711,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331076552563185e-05,
+      "loss": 3.7676,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330237957812133e-05,
+      "loss": 3.7794,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329399363061081e-05,
+      "loss": 3.7598,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328560768310029e-05,
+      "loss": 3.7617,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327722173558977e-05,
+      "loss": 3.767,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326885216688298e-05,
+      "loss": 3.7602,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326046621937246e-05,
+      "loss": 3.7804,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325208027186194e-05,
+      "loss": 3.7672,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324369432435142e-05,
+      "loss": 3.7591,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32353083768409e-05,
+      "loss": 3.7628,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322692242933038e-05,
+      "loss": 3.7558,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321853648181986e-05,
+      "loss": 3.7769,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321015053430934e-05,
+      "loss": 3.7582,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320178096560255e-05,
+      "loss": 3.7732,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319339501809203e-05,
+      "loss": 3.7516,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3185009070581506e-05,
+      "loss": 3.7646,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3176623123070986e-05,
+      "loss": 3.7583,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3168237175560466e-05,
+      "loss": 3.7602,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3159851228049946e-05,
+      "loss": 3.7664,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3151465280539426e-05,
+      "loss": 3.7645,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3143095711832635e-05,
+      "loss": 3.767,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3134709764322115e-05,
+      "loss": 3.7791,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3126323816811595e-05,
+      "loss": 3.7777,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311795424810481e-05,
+      "loss": 3.7707,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310956830059429e-05,
+      "loss": 3.7603,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310118235308377e-05,
+      "loss": 3.7556,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309279640557325e-05,
+      "loss": 3.7613,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308442683686646e-05,
+      "loss": 3.7784,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307604088935594e-05,
+      "loss": 3.769,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.306765494184542e-05,
+      "loss": 3.7588,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30592689943349e-05,
+      "loss": 3.7466,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305089942562811e-05,
+      "loss": 3.7737,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304251347811759e-05,
+      "loss": 3.7416,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.303412753060707e-05,
+      "loss": 3.76,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.302574158309655e-05,
+      "loss": 3.7595,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3017355635586035e-05,
+      "loss": 3.7633,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3008986066879244e-05,
+      "loss": 3.7633,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3000600119368724e-05,
+      "loss": 3.7712,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2992214171858204e-05,
+      "loss": 3.7507,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2983828224347684e-05,
+      "loss": 3.7604,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297545865564089e-05,
+      "loss": 3.7643,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296707270813037e-05,
+      "loss": 3.7553,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295868676061985e-05,
+      "loss": 3.7521,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295030081310933e-05,
+      "loss": 3.7691,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294193124440254e-05,
+      "loss": 3.7663,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293354529689202e-05,
+      "loss": 3.7754,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29251593493815e-05,
+      "loss": 3.7561,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291677340187099e-05,
+      "loss": 3.7614,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290838745436047e-05,
+      "loss": 3.7494,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290001788565368e-05,
+      "loss": 3.7641,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289163193814316e-05,
+      "loss": 3.7704,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288324599063264e-05,
+      "loss": 3.7565,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287486004312212e-05,
+      "loss": 3.772,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28664740956116e-05,
+      "loss": 3.7419,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285808814810108e-05,
+      "loss": 3.7506,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284971857939429e-05,
+      "loss": 3.748,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284133263188377e-05,
+      "loss": 3.7547,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2832946684373247e-05,
+      "loss": 3.7655,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2824560736862727e-05,
+      "loss": 3.7592,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2816174789352207e-05,
+      "loss": 3.7366,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280780522064542e-05,
+      "loss": 3.7582,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27994192731349e-05,
+      "loss": 3.7646,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279103332562438e-05,
+      "loss": 3.7611,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278264737811386e-05,
+      "loss": 3.7623,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277426143060334e-05,
+      "loss": 3.7523,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276589186189655e-05,
+      "loss": 3.7491,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275750591438603e-05,
+      "loss": 3.768,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274911996687551e-05,
+      "loss": 3.7533,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274073401936499e-05,
+      "loss": 3.7504,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2732348071854464e-05,
+      "loss": 3.7445,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272397850314768e-05,
+      "loss": 3.7499,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271559255563716e-05,
+      "loss": 3.7577,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270720660812664e-05,
+      "loss": 3.7575,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269882066061612e-05,
+      "loss": 3.7642,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26904347131056e-05,
+      "loss": 3.7488,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2682065144398816e-05,
+      "loss": 3.7478,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267367919688829e-05,
+      "loss": 3.768,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266529324937777e-05,
+      "loss": 3.7506,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265690730186725e-05,
+      "loss": 3.7587,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264852135435673e-05,
+      "loss": 3.764,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264013540684621e-05,
+      "loss": 3.7647,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263174945933569e-05,
+      "loss": 3.7607,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26233798906289e-05,
+      "loss": 3.7484,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261499394311838e-05,
+      "loss": 3.7406,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260660799560786e-05,
+      "loss": 3.7502,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2598222048097345e-05,
+      "loss": 3.7563,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2589852479390554e-05,
+      "loss": 3.7637,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2581466531880034e-05,
+      "loss": 3.7572,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2573080584369514e-05,
+      "loss": 3.7565,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2564694636858994e-05,
+      "loss": 3.7525,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25563250681522e-05,
+      "loss": 3.7522,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254793912064168e-05,
+      "loss": 3.7415,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253955317313116e-05,
+      "loss": 3.749,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253116722562064e-05,
+      "loss": 3.7613,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252278127811012e-05,
+      "loss": 3.7597,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25143953305996e-05,
+      "loss": 3.7446,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250602576189281e-05,
+      "loss": 3.7451,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.868739366531372,
+      "eval_runtime": 570.4041,
+      "eval_samples_per_second": 668.984,
+      "eval_steps_per_second": 20.906,
+      "step": 457914
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 3.183624312697221e+17,
+  "trial_name": null,
+  "trial_params": null
+}