diff --git "a/checkpoint-992150/trainer_state.json" "b/checkpoint-992150/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-992150/trainer_state.json"
@@ -0,0 +1,11751 @@
+{
+  "best_metric": 3.837082862854004,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full/transformer/4/checkpoints/checkpoint-992150",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 992150,
+  "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
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24976398143823e-05,
+      "loss": 3.7636,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248925386687178e-05,
+      "loss": 3.7501,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248086791936126e-05,
+      "loss": 3.7465,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247248197185074e-05,
+      "loss": 3.7468,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246409602434022e-05,
+      "loss": 3.7413,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245572645563343e-05,
+      "loss": 3.7391,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244734050812291e-05,
+      "loss": 3.7373,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243895456061239e-05,
+      "loss": 3.7442,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243056861310187e-05,
+      "loss": 3.7474,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2422199044395076e-05,
+      "loss": 3.7479,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2413813096884556e-05,
+      "loss": 3.747,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2405427149374036e-05,
+      "loss": 3.7391,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2397041201863516e-05,
+      "loss": 3.7501,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2388655254352996e-05,
+      "loss": 3.7459,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238028568564621e-05,
+      "loss": 3.7461,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237189973813569e-05,
+      "loss": 3.7311,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236351379062517e-05,
+      "loss": 3.7308,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235512784311465e-05,
+      "loss": 3.7481,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234675827440786e-05,
+      "loss": 3.7402,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233837232689734e-05,
+      "loss": 3.7553,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232998637938682e-05,
+      "loss": 3.7378,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23216004318763e-05,
+      "loss": 3.75,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2313214484365774e-05,
+      "loss": 3.7443,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230484491565899e-05,
+      "loss": 3.7425,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229645896814847e-05,
+      "loss": 3.7384,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228807302063795e-05,
+      "loss": 3.7384,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227968707312743e-05,
+      "loss": 3.7486,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2271317504420645e-05,
+      "loss": 3.732,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2262931556910125e-05,
+      "loss": 3.7442,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22545456093996e-05,
+      "loss": 3.738,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224615966188908e-05,
+      "loss": 3.7258,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237790093182294e-05,
+      "loss": 3.7418,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2229404145671774e-05,
+      "loss": 3.7445,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222101819816125e-05,
+      "loss": 3.7475,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221263225065073e-05,
+      "loss": 3.735,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220424630314021e-05,
+      "loss": 3.7387,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219587673443342e-05,
+      "loss": 3.738,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21874907869229e-05,
+      "loss": 3.7513,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217910483941238e-05,
+      "loss": 3.7383,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217071889190186e-05,
+      "loss": 3.7252,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216234932319507e-05,
+      "loss": 3.7295,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215396337568455e-05,
+      "loss": 3.7262,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214557742817403e-05,
+      "loss": 3.7453,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213719148066351e-05,
+      "loss": 3.7348,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212882191195672e-05,
+      "loss": 3.7353,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21204359644462e-05,
+      "loss": 3.7423,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211205001693568e-05,
+      "loss": 3.7326,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210366406942516e-05,
+      "loss": 3.7454,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209529450071838e-05,
+      "loss": 3.7273,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208690855320786e-05,
+      "loss": 3.7204,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207852260569734e-05,
+      "loss": 3.7338,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207013665818682e-05,
+      "loss": 3.727,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20617507106763e-05,
+      "loss": 3.7323,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2053381141969506e-05,
+      "loss": 3.7344,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2044995194458986e-05,
+      "loss": 3.7249,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036609246948466e-05,
+      "loss": 3.7226,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2028223299437946e-05,
+      "loss": 3.7251,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2019853730731155e-05,
+      "loss": 3.7248,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011467783220635e-05,
+      "loss": 3.7389,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2003081835710115e-05,
+      "loss": 3.7283,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1994695888199595e-05,
+      "loss": 3.7188,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198632631949281e-05,
+      "loss": 3.7254,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197794037198229e-05,
+      "loss": 3.7121,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196955442447177e-05,
+      "loss": 3.7398,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196116847696125e-05,
+      "loss": 3.7212,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195278252945073e-05,
+      "loss": 3.7289,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194441296074394e-05,
+      "loss": 3.718,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193602701323342e-05,
+      "loss": 3.7182,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19276410657229e-05,
+      "loss": 3.723,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191925511821238e-05,
+      "loss": 3.7205,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191088554950559e-05,
+      "loss": 3.729,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190249960199507e-05,
+      "loss": 3.7252,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189411365448455e-05,
+      "loss": 3.7305,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188572770697403e-05,
+      "loss": 3.7394,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1877341759463515e-05,
+      "loss": 3.7374,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1868972190756724e-05,
+      "loss": 3.7345,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860586243246204e-05,
+      "loss": 3.7229,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1852200295735684e-05,
+      "loss": 3.7167,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843814348225164e-05,
+      "loss": 3.7219,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183544477951837e-05,
+      "loss": 3.7387,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182705883200785e-05,
+      "loss": 3.7314,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181867288449733e-05,
+      "loss": 3.7197,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181028693698681e-05,
+      "loss": 3.7083,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180191736828002e-05,
+      "loss": 3.7381,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17935314207695e-05,
+      "loss": 3.7035,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178514547325898e-05,
+      "loss": 3.722,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177675952574847e-05,
+      "loss": 3.7243,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176838995704168e-05,
+      "loss": 3.7217,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176000400953116e-05,
+      "loss": 3.73,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175161806202064e-05,
+      "loss": 3.7358,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174323211451012e-05,
+      "loss": 3.7104,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173486254580333e-05,
+      "loss": 3.7242,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172647659829281e-05,
+      "loss": 3.7264,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171809065078229e-05,
+      "loss": 3.7176,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170970470327177e-05,
+      "loss": 3.7138,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1701335134564976e-05,
+      "loss": 3.7316,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1692949187054456e-05,
+      "loss": 3.7304,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1684563239543936e-05,
+      "loss": 3.7385,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167617729203342e-05,
+      "loss": 3.7184,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166780772332663e-05,
+      "loss": 3.7251,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165942177581611e-05,
+      "loss": 3.7122,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165103582830559e-05,
+      "loss": 3.7259,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164264988079507e-05,
+      "loss": 3.7317,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163426393328455e-05,
+      "loss": 3.7252,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162589436457776e-05,
+      "loss": 3.7321,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161750841706724e-05,
+      "loss": 3.7085,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160912246955672e-05,
+      "loss": 3.7099,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16007365220462e-05,
+      "loss": 3.7077,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1592350574535674e-05,
+      "loss": 3.7183,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158398100582889e-05,
+      "loss": 3.7276,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1575595058318376e-05,
+      "loss": 3.7257,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1567209110807856e-05,
+      "loss": 3.6992,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155882316329733e-05,
+      "loss": 3.7186,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1550453594590545e-05,
+      "loss": 3.7282,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1542067647080025e-05,
+      "loss": 3.7271,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15336816995695e-05,
+      "loss": 3.7247,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152529575205898e-05,
+      "loss": 3.718,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1516926183352194e-05,
+      "loss": 3.7156,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1508540235841674e-05,
+      "loss": 3.7269,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150015428833115e-05,
+      "loss": 3.7201,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149176834082063e-05,
+      "loss": 3.7139,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1483382393310114e-05,
+      "loss": 3.7079,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147501282460333e-05,
+      "loss": 3.7137,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14666268770928e-05,
+      "loss": 3.7211,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145824092958228e-05,
+      "loss": 3.7201,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144985498207176e-05,
+      "loss": 3.7232,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144146903456124e-05,
+      "loss": 3.7137,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143309946585445e-05,
+      "loss": 3.7115,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142471351834393e-05,
+      "loss": 3.7291,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141632757083341e-05,
+      "loss": 3.7096,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140794162332289e-05,
+      "loss": 3.7258,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13995720546161e-05,
+      "loss": 3.7294,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139118610710558e-05,
+      "loss": 3.7258,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138280015959507e-05,
+      "loss": 3.7238,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137441421208455e-05,
+      "loss": 3.7173,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136602826457403e-05,
+      "loss": 3.705,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357658695867237e-05,
+      "loss": 3.7124,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1349272748356717e-05,
+      "loss": 3.7179,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340886800846197e-05,
+      "loss": 3.7302,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332500853335676e-05,
+      "loss": 3.7166,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1324131284628886e-05,
+      "loss": 3.7257,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315745337118366e-05,
+      "loss": 3.7132,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1307359389607845e-05,
+      "loss": 3.7162,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1298973442097325e-05,
+      "loss": 3.7031,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1290587494586805e-05,
+      "loss": 3.7143,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1282201547076285e-05,
+      "loss": 3.7259,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12738319783695e-05,
+      "loss": 3.7264,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126544603085898e-05,
+      "loss": 3.7071,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125706008334846e-05,
+      "loss": 3.7143,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.857569456100464,
+      "eval_runtime": 575.5623,
+      "eval_samples_per_second": 662.988,
+      "eval_steps_per_second": 20.719,
+      "step": 534233
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124867413583794e-05,
+      "loss": 3.732,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124028818832742e-05,
+      "loss": 3.7154,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.12319022408169e-05,
+      "loss": 3.7108,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.122351629330638e-05,
+      "loss": 3.713,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.121514672459959e-05,
+      "loss": 3.7025,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.120676077708907e-05,
+      "loss": 3.7078,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119837482957855e-05,
+      "loss": 3.6969,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118998888206803e-05,
+      "loss": 3.7098,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118161931336124e-05,
+      "loss": 3.7093,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117323336585072e-05,
+      "loss": 3.7155,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11648474183402e-05,
+      "loss": 3.7101,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1156461470829686e-05,
+      "loss": 3.707,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114807552331916e-05,
+      "loss": 3.7159,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1139705954612375e-05,
+      "loss": 3.7062,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1131320007101855e-05,
+      "loss": 3.7163,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1122934059591335e-05,
+      "loss": 3.6943,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111454811208081e-05,
+      "loss": 3.6996,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1106178543374024e-05,
+      "loss": 3.7115,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1097792595863504e-05,
+      "loss": 3.7029,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1089406648352984e-05,
+      "loss": 3.7191,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108102070084246e-05,
+      "loss": 3.7078,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.107265113213567e-05,
+      "loss": 3.7118,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106426518462515e-05,
+      "loss": 3.7103,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.105587923711463e-05,
+      "loss": 3.7036,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.104749328960411e-05,
+      "loss": 3.7075,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103910734209359e-05,
+      "loss": 3.7006,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103073777338681e-05,
+      "loss": 3.7144,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.102235182587628e-05,
+      "loss": 3.705,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101396587836576e-05,
+      "loss": 3.7089,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100557993085524e-05,
+      "loss": 3.7028,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099721036214846e-05,
+      "loss": 3.6953,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098882441463793e-05,
+      "loss": 3.7072,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098043846712741e-05,
+      "loss": 3.7092,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097205251961689e-05,
+      "loss": 3.7085,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0963682950910106e-05,
+      "loss": 3.7061,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0955297003399586e-05,
+      "loss": 3.7001,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0946911055889066e-05,
+      "loss": 3.7052,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0938525108378546e-05,
+      "loss": 3.7149,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0930155539671755e-05,
+      "loss": 3.7055,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0921769592161235e-05,
+      "loss": 3.6941,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0913383644650715e-05,
+      "loss": 3.6924,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0904997697140195e-05,
+      "loss": 3.6931,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0896628128433404e-05,
+      "loss": 3.7118,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0888242180922884e-05,
+      "loss": 3.7007,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0879856233412364e-05,
+      "loss": 3.6987,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0871470285901844e-05,
+      "loss": 3.7137,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.086310071719506e-05,
+      "loss": 3.6987,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085471476968454e-05,
+      "loss": 3.71,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084632882217402e-05,
+      "loss": 3.694,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08379428746635e-05,
+      "loss": 3.686,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082957330595671e-05,
+      "loss": 3.6982,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082118735844619e-05,
+      "loss": 3.6967,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081280141093567e-05,
+      "loss": 3.693,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080441546342515e-05,
+      "loss": 3.7012,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079604589471836e-05,
+      "loss": 3.6936,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078765994720784e-05,
+      "loss": 3.687,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077927399969732e-05,
+      "loss": 3.6989,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.07708880521868e-05,
+      "loss": 3.688,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0762518483480014e-05,
+      "loss": 3.7048,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0754132535969494e-05,
+      "loss": 3.6947,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0745746588458974e-05,
+      "loss": 3.6884,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0737360640948453e-05,
+      "loss": 3.6908,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072899107224166e-05,
+      "loss": 3.6824,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072060512473114e-05,
+      "loss": 3.7039,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071221917722062e-05,
+      "loss": 3.6878,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.07038332297101e-05,
+      "loss": 3.6956,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069544728219958e-05,
+      "loss": 3.6896,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068707771349279e-05,
+      "loss": 3.6861,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067869176598227e-05,
+      "loss": 3.6866,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067030581847175e-05,
+      "loss": 3.6902,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066191987096124e-05,
+      "loss": 3.6945,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065353392345072e-05,
+      "loss": 3.6928,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064516435474393e-05,
+      "loss": 3.6978,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063677840723341e-05,
+      "loss": 3.7063,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062839245972289e-05,
+      "loss": 3.7029,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062000651221237e-05,
+      "loss": 3.6996,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0611636943505576e-05,
+      "loss": 3.6976,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0603250995995056e-05,
+      "loss": 3.6781,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0594865048484536e-05,
+      "loss": 3.6914,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0586479100974016e-05,
+      "loss": 3.7076,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0578109532267225e-05,
+      "loss": 3.6981,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0569723584756705e-05,
+      "loss": 3.6921,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056133763724619e-05,
+      "loss": 3.6754,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05529680685394e-05,
+      "loss": 3.7,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054458212102888e-05,
+      "loss": 3.6767,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053619617351836e-05,
+      "loss": 3.6881,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.052781022600784e-05,
+      "loss": 3.6867,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051942427849732e-05,
+      "loss": 3.6931,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051105470979053e-05,
+      "loss": 3.6985,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.050266876228001e-05,
+      "loss": 3.6989,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049428281476949e-05,
+      "loss": 3.6817,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.048589686725897e-05,
+      "loss": 3.6954,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.047752729855218e-05,
+      "loss": 3.6897,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046914135104166e-05,
+      "loss": 3.6859,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046075540353114e-05,
+      "loss": 3.6822,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0452369456020625e-05,
+      "loss": 3.6969,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0443999887313835e-05,
+      "loss": 3.6977,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0435613939803314e-05,
+      "loss": 3.7064,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0427227992292794e-05,
+      "loss": 3.6903,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0418842044782274e-05,
+      "loss": 3.6928,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0410472476075483e-05,
+      "loss": 3.6806,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0402086528564963e-05,
+      "loss": 3.6954,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039370058105444e-05,
+      "loss": 3.6995,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038531463354392e-05,
+      "loss": 3.6963,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037694506483713e-05,
+      "loss": 3.697,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036855911732661e-05,
+      "loss": 3.6806,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036017316981609e-05,
+      "loss": 3.6768,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035178722230558e-05,
+      "loss": 3.6788,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034341765359879e-05,
+      "loss": 3.6847,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033503170608827e-05,
+      "loss": 3.6976,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032664575857775e-05,
+      "loss": 3.6934,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031825981106723e-05,
+      "loss": 3.6647,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030989024236044e-05,
+      "loss": 3.6933,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030150429484992e-05,
+      "loss": 3.6972,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.02931183473394e-05,
+      "loss": 3.6914,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028473239982888e-05,
+      "loss": 3.6938,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0276362831122086e-05,
+      "loss": 3.6889,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0267976883611566e-05,
+      "loss": 3.6811,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0259590936101046e-05,
+      "loss": 3.6963,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025120498859053e-05,
+      "loss": 3.6889,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.024281904108001e-05,
+      "loss": 3.6829,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.023444947237322e-05,
+      "loss": 3.6752,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.02260635248627e-05,
+      "loss": 3.6842,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.021767757735218e-05,
+      "loss": 3.6909,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0209291629841655e-05,
+      "loss": 3.6868,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020092206113487e-05,
+      "loss": 3.6967,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.019253611362435e-05,
+      "loss": 3.6827,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.018415016611383e-05,
+      "loss": 3.6818,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0175764218603304e-05,
+      "loss": 3.7011,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.016739464989652e-05,
+      "loss": 3.679,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0159008702386e-05,
+      "loss": 3.6952,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0150622754875486e-05,
+      "loss": 3.7009,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014223680736496e-05,
+      "loss": 3.6921,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0133867238658175e-05,
+      "loss": 3.6931,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0125481291147655e-05,
+      "loss": 3.6862,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011709534363713e-05,
+      "loss": 3.6744,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010870939612661e-05,
+      "loss": 3.6808,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0100339827419824e-05,
+      "loss": 3.689,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0091953879909304e-05,
+      "loss": 3.6944,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008356793239878e-05,
+      "loss": 3.6891,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007518198488826e-05,
+      "loss": 3.6935,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006679603737774e-05,
+      "loss": 3.6807,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005842646867095e-05,
+      "loss": 3.6862,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005004052116043e-05,
+      "loss": 3.6727,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004165457364991e-05,
+      "loss": 3.684,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.003326862613939e-05,
+      "loss": 3.6937,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00248990574326e-05,
+      "loss": 3.6958,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.001651310992208e-05,
+      "loss": 3.675,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000812716241156e-05,
+      "loss": 3.6846,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.851036787033081,
+      "eval_runtime": 579.0615,
+      "eval_samples_per_second": 658.982,
+      "eval_steps_per_second": 20.594,
+      "step": 610552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999974121490104e-05,
+      "loss": 3.7001,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999135526739052e-05,
+      "loss": 3.6871,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998296931988e-05,
+      "loss": 3.6814,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997458337236948e-05,
+      "loss": 3.6842,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996621380366269e-05,
+      "loss": 3.6722,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995782785615218e-05,
+      "loss": 3.6753,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994944190864166e-05,
+      "loss": 3.6743,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994105596113114e-05,
+      "loss": 3.6779,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993268639242435e-05,
+      "loss": 3.6789,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992430044491383e-05,
+      "loss": 3.6856,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991591449740331e-05,
+      "loss": 3.6791,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990752854989279e-05,
+      "loss": 3.6804,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9899158981185996e-05,
+      "loss": 3.687,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890773033675476e-05,
+      "loss": 3.6729,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9882387086164956e-05,
+      "loss": 3.6894,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9874001138654436e-05,
+      "loss": 3.6647,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865631569947645e-05,
+      "loss": 3.6683,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985724562243713e-05,
+      "loss": 3.683,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984885967492661e-05,
+      "loss": 3.6747,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984047372741609e-05,
+      "loss": 3.6925,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98321041587093e-05,
+      "loss": 3.6789,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982371821119878e-05,
+      "loss": 3.6801,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981533226368826e-05,
+      "loss": 3.6809,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980694631617774e-05,
+      "loss": 3.6747,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979857674747095e-05,
+      "loss": 3.6802,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979019079996043e-05,
+      "loss": 3.6696,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978180485244991e-05,
+      "loss": 3.6843,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977341890493939e-05,
+      "loss": 3.6746,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97650493362326e-05,
+      "loss": 3.6817,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9756663388722085e-05,
+      "loss": 3.6705,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9748277441211565e-05,
+      "loss": 3.6645,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9739891493701045e-05,
+      "loss": 3.6791,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9731521924994254e-05,
+      "loss": 3.6806,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9723135977483734e-05,
+      "loss": 3.6789,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9714750029973214e-05,
+      "loss": 3.6758,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9706364082462694e-05,
+      "loss": 3.6739,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9697978134952174e-05,
+      "loss": 3.6737,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968960856624538e-05,
+      "loss": 3.6844,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968122261873486e-05,
+      "loss": 3.679,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967283667122434e-05,
+      "loss": 3.6648,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966445072371382e-05,
+      "loss": 3.6635,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965608115500703e-05,
+      "loss": 3.6633,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964769520749652e-05,
+      "loss": 3.682,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9639309259986e-05,
+      "loss": 3.6776,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963092331247548e-05,
+      "loss": 3.6661,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962255374376869e-05,
+      "loss": 3.6856,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.961416779625817e-05,
+      "loss": 3.6699,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960578184874765e-05,
+      "loss": 3.6802,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959739590123713e-05,
+      "loss": 3.6711,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958902633253034e-05,
+      "loss": 3.6612,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958064038501982e-05,
+      "loss": 3.6659,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95722544375093e-05,
+      "loss": 3.6656,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956386848999878e-05,
+      "loss": 3.6726,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9555498921291986e-05,
+      "loss": 3.6671,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954711297378147e-05,
+      "loss": 3.6697,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953872702627095e-05,
+      "loss": 3.6576,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953034107876043e-05,
+      "loss": 3.6702,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952197151005364e-05,
+      "loss": 3.6602,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951358556254312e-05,
+      "loss": 3.6751,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95051996150326e-05,
+      "loss": 3.6613,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949681366752208e-05,
+      "loss": 3.6644,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948844409881529e-05,
+      "loss": 3.6644,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948005815130477e-05,
+      "loss": 3.6545,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947167220379425e-05,
+      "loss": 3.6742,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946328625628373e-05,
+      "loss": 3.6586,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945491668757694e-05,
+      "loss": 3.6686,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9446530740066426e-05,
+      "loss": 3.6633,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9438144792555906e-05,
+      "loss": 3.6566,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429758845045386e-05,
+      "loss": 3.6579,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9421389276338595e-05,
+      "loss": 3.6626,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9413003328828075e-05,
+      "loss": 3.6656,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9404617381317555e-05,
+      "loss": 3.6673,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939623143380703e-05,
+      "loss": 3.6684,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9387861865100244e-05,
+      "loss": 3.6862,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9379475917589724e-05,
+      "loss": 3.6744,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9371089970079204e-05,
+      "loss": 3.6675,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936270402256868e-05,
+      "loss": 3.6678,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935433445386189e-05,
+      "loss": 3.6536,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934594850635138e-05,
+      "loss": 3.6612,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933756255884086e-05,
+      "loss": 3.677,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932917661133033e-05,
+      "loss": 3.6744,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932080704262355e-05,
+      "loss": 3.6636,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931242109511303e-05,
+      "loss": 3.6492,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93040351476025e-05,
+      "loss": 3.6705,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929564920009198e-05,
+      "loss": 3.6462,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92872796313852e-05,
+      "loss": 3.662,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927889368387468e-05,
+      "loss": 3.6619,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927050773636415e-05,
+      "loss": 3.6602,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926212178885363e-05,
+      "loss": 3.6739,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925375222014685e-05,
+      "loss": 3.669,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9245366272636333e-05,
+      "loss": 3.6592,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923698032512581e-05,
+      "loss": 3.6613,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922859437761529e-05,
+      "loss": 3.6652,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.92202248089085e-05,
+      "loss": 3.6579,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211838861397976e-05,
+      "loss": 3.6539,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9203452913887456e-05,
+      "loss": 3.6672,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9195066966376936e-05,
+      "loss": 3.6722,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918669739767015e-05,
+      "loss": 3.6814,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9178311450159625e-05,
+      "loss": 3.6593,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9169925502649105e-05,
+      "loss": 3.6658,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9161539555138585e-05,
+      "loss": 3.6558,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91531699864318e-05,
+      "loss": 3.6631,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914478403892128e-05,
+      "loss": 3.6738,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913639809141076e-05,
+      "loss": 3.6688,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912801214390024e-05,
+      "loss": 3.6731,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911964257519345e-05,
+      "loss": 3.6518,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911125662768293e-05,
+      "loss": 3.6525,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910287068017241e-05,
+      "loss": 3.6517,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909448473266189e-05,
+      "loss": 3.6579,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90861151639551e-05,
+      "loss": 3.665,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907772921644458e-05,
+      "loss": 3.668,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906934326893406e-05,
+      "loss": 3.6368,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906095732142354e-05,
+      "loss": 3.6656,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9052587752716754e-05,
+      "loss": 3.6694,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9044201805206234e-05,
+      "loss": 3.6634,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9035815857695714e-05,
+      "loss": 3.6669,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9027429910185194e-05,
+      "loss": 3.666,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90190603414784e-05,
+      "loss": 3.6509,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901067439396788e-05,
+      "loss": 3.6685,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900228844645736e-05,
+      "loss": 3.6679,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899390249894684e-05,
+      "loss": 3.6528,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898553293024005e-05,
+      "loss": 3.6484,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897714698272953e-05,
+      "loss": 3.6575,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896876103521901e-05,
+      "loss": 3.6604,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896037508770849e-05,
+      "loss": 3.6588,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895200551900171e-05,
+      "loss": 3.6679,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894361957149119e-05,
+      "loss": 3.6599,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893523362398067e-05,
+      "loss": 3.6518,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892684767647015e-05,
+      "loss": 3.6753,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891847810776336e-05,
+      "loss": 3.6518,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891009216025284e-05,
+      "loss": 3.6678,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890170621274232e-05,
+      "loss": 3.6783,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8893320265231797e-05,
+      "loss": 3.6646,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8884950696525006e-05,
+      "loss": 3.6675,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8876564749014486e-05,
+      "loss": 3.6603,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8868178801503966e-05,
+      "loss": 3.6489,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8859792853993446e-05,
+      "loss": 3.6568,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885142328528666e-05,
+      "loss": 3.6585,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884303733777614e-05,
+      "loss": 3.6699,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883465139026562e-05,
+      "loss": 3.6624,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88262654427551e-05,
+      "loss": 3.663,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881789587404831e-05,
+      "loss": 3.6606,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880950992653779e-05,
+      "loss": 3.6574,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880112397902727e-05,
+      "loss": 3.6463,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879273803151675e-05,
+      "loss": 3.658,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878436846280996e-05,
+      "loss": 3.6666,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877598251529944e-05,
+      "loss": 3.6701,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876759656778892e-05,
+      "loss": 3.6501,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.87592106202784e-05,
+      "loss": 3.6544,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8464159965515137,
+      "eval_runtime": 319.7597,
+      "eval_samples_per_second": 1193.368,
+      "eval_steps_per_second": 37.294,
+      "step": 686871
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8750841051571615e-05,
+      "loss": 3.6669,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8742455104061095e-05,
+      "loss": 3.6604,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8734069156550575e-05,
+      "loss": 3.6582,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8725683209040055e-05,
+      "loss": 3.6562,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8717313640333264e-05,
+      "loss": 3.6488,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8708927692822744e-05,
+      "loss": 3.6483,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8700541745312224e-05,
+      "loss": 3.6475,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8692155797801704e-05,
+      "loss": 3.65,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.868378622909491e-05,
+      "loss": 3.6505,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.867540028158439e-05,
+      "loss": 3.6608,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.866701433407387e-05,
+      "loss": 3.6498,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865862838656335e-05,
+      "loss": 3.6538,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.865025881785657e-05,
+      "loss": 3.6619,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864187287034605e-05,
+      "loss": 3.6478,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.863348692283553e-05,
+      "loss": 3.6586,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862510097532501e-05,
+      "loss": 3.6388,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861673140661822e-05,
+      "loss": 3.6417,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86083454591077e-05,
+      "loss": 3.6585,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859995951159718e-05,
+      "loss": 3.6508,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859157356408666e-05,
+      "loss": 3.665,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858320399537987e-05,
+      "loss": 3.6499,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857481804786935e-05,
+      "loss": 3.6637,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8566432100358827e-05,
+      "loss": 3.6476,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8558046152848307e-05,
+      "loss": 3.6486,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854967658414152e-05,
+      "loss": 3.6533,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8541290636631e-05,
+      "loss": 3.65,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853290468912048e-05,
+      "loss": 3.6573,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852451874160996e-05,
+      "loss": 3.6482,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851614917290317e-05,
+      "loss": 3.6549,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.850776322539265e-05,
+      "loss": 3.6488,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849937727788213e-05,
+      "loss": 3.6341,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849099133037161e-05,
+      "loss": 3.6567,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848262176166482e-05,
+      "loss": 3.654,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84742358141543e-05,
+      "loss": 3.651,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846584986664378e-05,
+      "loss": 3.652,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.845746391913326e-05,
+      "loss": 3.6461,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8449094350426476e-05,
+      "loss": 3.6491,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8440708402915956e-05,
+      "loss": 3.6558,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8432322455405436e-05,
+      "loss": 3.6557,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8423936507894916e-05,
+      "loss": 3.6374,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8415550560384396e-05,
+      "loss": 3.6424,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8407180991677605e-05,
+      "loss": 3.642,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8398795044167085e-05,
+      "loss": 3.6496,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8390409096656565e-05,
+      "loss": 3.6537,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8382023149146045e-05,
+      "loss": 3.6411,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8373653580439254e-05,
+      "loss": 3.6586,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8365267632928734e-05,
+      "loss": 3.6428,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8356881685418214e-05,
+      "loss": 3.6573,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8348495737907694e-05,
+      "loss": 3.6445,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834012616920091e-05,
+      "loss": 3.6367,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833174022169039e-05,
+      "loss": 3.6441,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832335427417987e-05,
+      "loss": 3.6357,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831496832666935e-05,
+      "loss": 3.6511,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830659875796256e-05,
+      "loss": 3.64,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.829821281045204e-05,
+      "loss": 3.647,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828982686294152e-05,
+      "loss": 3.6299,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8281440915431e-05,
+      "loss": 3.6438,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.827307134672421e-05,
+      "loss": 3.6331,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826468539921369e-05,
+      "loss": 3.6517,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.825629945170317e-05,
+      "loss": 3.6368,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824791350419265e-05,
+      "loss": 3.6393,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823954393548586e-05,
+      "loss": 3.638,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823115798797534e-05,
+      "loss": 3.6291,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822277204046482e-05,
+      "loss": 3.6488,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.82143860929543e-05,
+      "loss": 3.6378,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820601652424751e-05,
+      "loss": 3.6456,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819763057673699e-05,
+      "loss": 3.6342,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818924462922647e-05,
+      "loss": 3.6338,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818087506051968e-05,
+      "loss": 3.632,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817248911300916e-05,
+      "loss": 3.6406,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816410316549864e-05,
+      "loss": 3.637,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815571721798812e-05,
+      "loss": 3.6412,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814734764928134e-05,
+      "loss": 3.6426,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813896170177082e-05,
+      "loss": 3.6619,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81305757542603e-05,
+      "loss": 3.6469,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812218980674977e-05,
+      "loss": 3.6457,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8113820238042986e-05,
+      "loss": 3.6412,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8105434290532466e-05,
+      "loss": 3.6335,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8097048343021946e-05,
+      "loss": 3.6335,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808866239551142e-05,
+      "loss": 3.6508,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8080292826804635e-05,
+      "loss": 3.651,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8071906879294115e-05,
+      "loss": 3.6416,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8063520931783595e-05,
+      "loss": 3.6247,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8055134984273075e-05,
+      "loss": 3.6416,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804676541556629e-05,
+      "loss": 3.623,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803837946805577e-05,
+      "loss": 3.6371,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8029993520545244e-05,
+      "loss": 3.637,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8021607573034724e-05,
+      "loss": 3.6369,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.801323800432794e-05,
+      "loss": 3.6478,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.800485205681742e-05,
+      "loss": 3.6421,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.799646610930689e-05,
+      "loss": 3.635,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798808016179637e-05,
+      "loss": 3.6362,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797971059308959e-05,
+      "loss": 3.6419,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797132464557907e-05,
+      "loss": 3.633,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796293869806855e-05,
+      "loss": 3.6322,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.795455275055803e-05,
+      "loss": 3.6413,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7946183181851244e-05,
+      "loss": 3.6473,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793779723434072e-05,
+      "loss": 3.6574,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79294112868302e-05,
+      "loss": 3.6354,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792102533931968e-05,
+      "loss": 3.6447,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7912655770612893e-05,
+      "loss": 3.6336,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7904269823102367e-05,
+      "loss": 3.6397,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7895883875591847e-05,
+      "loss": 3.6409,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7887497928081326e-05,
+      "loss": 3.6467,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787912835937454e-05,
+      "loss": 3.6485,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7870742411864016e-05,
+      "loss": 3.6284,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78623564643535e-05,
+      "loss": 3.6297,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785397051684298e-05,
+      "loss": 3.6269,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.784560094813619e-05,
+      "loss": 3.63,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.783721500062567e-05,
+      "loss": 3.6425,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782882905311515e-05,
+      "loss": 3.6428,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782044310560463e-05,
+      "loss": 3.6179,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781207353689784e-05,
+      "loss": 3.641,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780368758938732e-05,
+      "loss": 3.6451,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77953016418768e-05,
+      "loss": 3.6398,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778691569436628e-05,
+      "loss": 3.642,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777854612565949e-05,
+      "loss": 3.6391,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777016017814897e-05,
+      "loss": 3.6304,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776177423063845e-05,
+      "loss": 3.6409,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7753388283127936e-05,
+      "loss": 3.6426,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7745018714421145e-05,
+      "loss": 3.6358,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7736632766910625e-05,
+      "loss": 3.6201,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7728246819400105e-05,
+      "loss": 3.6343,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7719860871889585e-05,
+      "loss": 3.6356,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7711491303182794e-05,
+      "loss": 3.6382,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7703105355672274e-05,
+      "loss": 3.6371,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7694719408161754e-05,
+      "loss": 3.6427,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7686333460651234e-05,
+      "loss": 3.6286,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767796389194444e-05,
+      "loss": 3.6493,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766957794443392e-05,
+      "loss": 3.6312,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76611919969234e-05,
+      "loss": 3.6436,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765280604941289e-05,
+      "loss": 3.6509,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76444364807061e-05,
+      "loss": 3.6397,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763605053319558e-05,
+      "loss": 3.6431,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.762766458568506e-05,
+      "loss": 3.6383,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761927863817454e-05,
+      "loss": 3.626,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761090906946775e-05,
+      "loss": 3.6341,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760252312195723e-05,
+      "loss": 3.6319,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759413717444671e-05,
+      "loss": 3.6438,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758575122693619e-05,
+      "loss": 3.6396,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7577381658229397e-05,
+      "loss": 3.6409,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7568995710718877e-05,
+      "loss": 3.6408,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7560609763208356e-05,
+      "loss": 3.6327,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755222381569784e-05,
+      "loss": 3.6222,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754385424699105e-05,
+      "loss": 3.6333,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753546829948053e-05,
+      "loss": 3.6446,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752708235197001e-05,
+      "loss": 3.644,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751869640445949e-05,
+      "loss": 3.6272,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75103268357527e-05,
+      "loss": 3.6297,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8441436290740967,
+      "eval_runtime": 319.558,
+      "eval_samples_per_second": 1194.121,
+      "eval_steps_per_second": 37.317,
+      "step": 763190
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750194088824218e-05,
+      "loss": 3.6267,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749355494073166e-05,
+      "loss": 3.6341,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748516899322114e-05,
+      "loss": 3.6378,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747678304571062e-05,
+      "loss": 3.6422,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746841347700383e-05,
+      "loss": 3.6479,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746002752949331e-05,
+      "loss": 3.6358,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.74516415819828e-05,
+      "loss": 3.6389,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744325563447228e-05,
+      "loss": 3.6365,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7434886065765486e-05,
+      "loss": 3.6301,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7426500118254966e-05,
+      "loss": 3.633,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7418114170744446e-05,
+      "loss": 3.643,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7409728223233926e-05,
+      "loss": 3.6414,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7401375033330864e-05,
+      "loss": 3.6393,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7392989085820344e-05,
+      "loss": 3.6144,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7384603138309824e-05,
+      "loss": 3.6405,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7376217190799304e-05,
+      "loss": 3.6223,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7367831243288784e-05,
+      "loss": 3.6333,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7359445295778264e-05,
+      "loss": 3.6312,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735105934826775e-05,
+      "loss": 3.632,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734267340075723e-05,
+      "loss": 3.6419,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733430383205044e-05,
+      "loss": 3.6461,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732591788453992e-05,
+      "loss": 3.6402,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73175319370294e-05,
+      "loss": 3.6313,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730914598951888e-05,
+      "loss": 3.626,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730077642081209e-05,
+      "loss": 3.6377,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729239047330157e-05,
+      "loss": 3.6347,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728400452579105e-05,
+      "loss": 3.6305,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727561857828053e-05,
+      "loss": 3.6293,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726724900957374e-05,
+      "loss": 3.6281,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725886306206322e-05,
+      "loss": 3.6314,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7250477114552704e-05,
+      "loss": 3.62,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7242091167042184e-05,
+      "loss": 3.6436,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723372159833539e-05,
+      "loss": 3.6388,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722533565082487e-05,
+      "loss": 3.6365,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.721694970331435e-05,
+      "loss": 3.6312,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720856375580383e-05,
+      "loss": 3.6285,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720019418709704e-05,
+      "loss": 3.6353,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719180823958652e-05,
+      "loss": 3.6294,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7183422292076e-05,
+      "loss": 3.6208,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717503634456548e-05,
+      "loss": 3.6242,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.716666677585869e-05,
+      "loss": 3.6251,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.715828082834817e-05,
+      "loss": 3.627,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714989488083766e-05,
+      "loss": 3.6308,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714150893332714e-05,
+      "loss": 3.6294,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.713313936462035e-05,
+      "loss": 3.6399,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.712475341710983e-05,
+      "loss": 3.6287,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711636746959931e-05,
+      "loss": 3.6338,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.710798152208879e-05,
+      "loss": 3.6333,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7099611953381996e-05,
+      "loss": 3.6063,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7091226005871476e-05,
+      "loss": 3.6405,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7082840058360956e-05,
+      "loss": 3.6235,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7074454110850436e-05,
+      "loss": 3.6291,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7066084542143645e-05,
+      "loss": 3.6341,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7057698594633125e-05,
+      "loss": 3.623,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704931264712261e-05,
+      "loss": 3.6105,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704092669961209e-05,
+      "loss": 3.623,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70325571309053e-05,
+      "loss": 3.6245,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702417118339478e-05,
+      "loss": 3.6199,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.701578523588426e-05,
+      "loss": 3.6358,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.700741566717747e-05,
+      "loss": 3.6234,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699902971966695e-05,
+      "loss": 3.6161,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699064377215643e-05,
+      "loss": 3.6441,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698225782464591e-05,
+      "loss": 3.6105,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697388825593912e-05,
+      "loss": 3.614,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69655023084286e-05,
+      "loss": 3.6249,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695711636091808e-05,
+      "loss": 3.6217,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694873041340756e-05,
+      "loss": 3.6174,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6940360844700774e-05,
+      "loss": 3.623,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6931974897190254e-05,
+      "loss": 3.6073,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6923588949679734e-05,
+      "loss": 3.615,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6915203002169214e-05,
+      "loss": 3.6338,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.690683343346242e-05,
+      "loss": 3.6212,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68984474859519e-05,
+      "loss": 3.6336,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689006153844138e-05,
+      "loss": 3.624,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688167559093086e-05,
+      "loss": 3.6411,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.687330602222407e-05,
+      "loss": 3.617,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686492007471355e-05,
+      "loss": 3.6312,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685653412720303e-05,
+      "loss": 3.6291,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684814817969251e-05,
+      "loss": 3.6074,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683977861098573e-05,
+      "loss": 3.6339,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683139266347521e-05,
+      "loss": 3.6268,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682300671596469e-05,
+      "loss": 3.6267,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681462076845416e-05,
+      "loss": 3.6293,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680625119974738e-05,
+      "loss": 3.607,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679786525223686e-05,
+      "loss": 3.6083,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678947930472634e-05,
+      "loss": 3.6176,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678109335721581e-05,
+      "loss": 3.6263,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6772723788509026e-05,
+      "loss": 3.6158,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6764337840998506e-05,
+      "loss": 3.6164,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6755951893487986e-05,
+      "loss": 3.6204,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6747565945977466e-05,
+      "loss": 3.6252,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673919637727068e-05,
+      "loss": 3.6173,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673081042976016e-05,
+      "loss": 3.6152,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6722424482249635e-05,
+      "loss": 3.6345,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6714038534739115e-05,
+      "loss": 3.6347,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.670566896603233e-05,
+      "loss": 3.6278,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669728301852181e-05,
+      "loss": 3.6113,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6688897071011284e-05,
+      "loss": 3.621,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6680511123500764e-05,
+      "loss": 3.6193,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667214155479398e-05,
+      "loss": 3.6307,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.666375560728346e-05,
+      "loss": 3.6192,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.665536965977294e-05,
+      "loss": 3.6199,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664698371226242e-05,
+      "loss": 3.6182,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6638614143555635e-05,
+      "loss": 3.6316,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663022819604511e-05,
+      "loss": 3.6196,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662184224853459e-05,
+      "loss": 3.6232,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.661345630102407e-05,
+      "loss": 3.611,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6605086732317284e-05,
+      "loss": 3.6179,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.659670078480676e-05,
+      "loss": 3.6144,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.658831483729624e-05,
+      "loss": 3.611,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657992888978572e-05,
+      "loss": 3.6162,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6571559321078926e-05,
+      "loss": 3.6168,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656317337356841e-05,
+      "loss": 3.6232,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655478742605789e-05,
+      "loss": 3.6141,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654640147854737e-05,
+      "loss": 3.6259,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.653803190984058e-05,
+      "loss": 3.6309,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652964596233006e-05,
+      "loss": 3.6211,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652126001481954e-05,
+      "loss": 3.6321,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651287406730902e-05,
+      "loss": 3.6146,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650450449860223e-05,
+      "loss": 3.6158,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649611855109171e-05,
+      "loss": 3.6155,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648773260358119e-05,
+      "loss": 3.6093,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647934665607067e-05,
+      "loss": 3.6203,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647097708736388e-05,
+      "loss": 3.62,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646259113985337e-05,
+      "loss": 3.6211,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.645420519234285e-05,
+      "loss": 3.6165,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644581924483233e-05,
+      "loss": 3.6255,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6437449676125536e-05,
+      "loss": 3.6179,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6429063728615016e-05,
+      "loss": 3.6136,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6420677781104496e-05,
+      "loss": 3.6221,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6412291833593976e-05,
+      "loss": 3.6222,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6403922264887185e-05,
+      "loss": 3.629,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6395536317376665e-05,
+      "loss": 3.6126,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6387150369866145e-05,
+      "loss": 3.6316,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6378764422355625e-05,
+      "loss": 3.618,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6370394853648834e-05,
+      "loss": 3.6246,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636200890613832e-05,
+      "loss": 3.6253,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63536229586278e-05,
+      "loss": 3.6131,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634523701111728e-05,
+      "loss": 3.6129,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633686744241049e-05,
+      "loss": 3.6298,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632848149489997e-05,
+      "loss": 3.6067,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632009554738945e-05,
+      "loss": 3.6223,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631170959987893e-05,
+      "loss": 3.6211,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630334003117214e-05,
+      "loss": 3.6175,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629495408366162e-05,
+      "loss": 3.6103,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.62865681361511e-05,
+      "loss": 3.6211,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627818218864058e-05,
+      "loss": 3.6212,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626981261993379e-05,
+      "loss": 3.6211,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6261426672423274e-05,
+      "loss": 3.6212,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8373005390167236,
+      "eval_runtime": 303.2718,
+      "eval_samples_per_second": 1258.247,
+      "eval_steps_per_second": 39.321,
+      "step": 839510
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6253040724912754e-05,
+      "loss": 3.6049,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6244654777402234e-05,
+      "loss": 3.6121,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.623628520869544e-05,
+      "loss": 3.6132,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622789926118492e-05,
+      "loss": 3.617,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.62195133136744e-05,
+      "loss": 3.6263,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621112736616388e-05,
+      "loss": 3.6109,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620275779745709e-05,
+      "loss": 3.6167,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.619437184994657e-05,
+      "loss": 3.6132,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.618598590243605e-05,
+      "loss": 3.6051,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617759995492553e-05,
+      "loss": 3.6063,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616923038621874e-05,
+      "loss": 3.6177,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616084443870823e-05,
+      "loss": 3.6225,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615247487000144e-05,
+      "loss": 3.6144,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.614408892249092e-05,
+      "loss": 3.5953,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61357029749804e-05,
+      "loss": 3.6157,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612731702746988e-05,
+      "loss": 3.5982,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611893107995936e-05,
+      "loss": 3.609,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611054513244884e-05,
+      "loss": 3.611,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610215918493832e-05,
+      "loss": 3.6086,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60937732374278e-05,
+      "loss": 3.6154,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6085403668721006e-05,
+      "loss": 3.6232,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6077017721210486e-05,
+      "loss": 3.6141,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6068631773699966e-05,
+      "loss": 3.6098,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6060245826189446e-05,
+      "loss": 3.6007,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605187625748266e-05,
+      "loss": 3.6158,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.604349030997214e-05,
+      "loss": 3.6121,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603510436246162e-05,
+      "loss": 3.6077,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60267184149511e-05,
+      "loss": 3.6084,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601834884624431e-05,
+      "loss": 3.6066,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600996289873379e-05,
+      "loss": 3.6079,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600157695122327e-05,
+      "loss": 3.602,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599319100371275e-05,
+      "loss": 3.6121,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598482143500596e-05,
+      "loss": 3.613,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597643548749544e-05,
+      "loss": 3.6149,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596804953998492e-05,
+      "loss": 3.609,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59596635924744e-05,
+      "loss": 3.6065,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5951294023767615e-05,
+      "loss": 3.6136,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5942908076257095e-05,
+      "loss": 3.6034,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5934522128746575e-05,
+      "loss": 3.5993,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5926136181236055e-05,
+      "loss": 3.6071,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5917766612529264e-05,
+      "loss": 3.5968,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5909380665018744e-05,
+      "loss": 3.6036,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5900994717508224e-05,
+      "loss": 3.6061,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5892608769997704e-05,
+      "loss": 3.6083,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588423920129091e-05,
+      "loss": 3.6124,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587585325378039e-05,
+      "loss": 3.6114,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586746730626987e-05,
+      "loss": 3.6106,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585908135875935e-05,
+      "loss": 3.606,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585071179005257e-05,
+      "loss": 3.5847,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584232584254205e-05,
+      "loss": 3.6165,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583393989503153e-05,
+      "loss": 3.6025,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582557032632474e-05,
+      "loss": 3.6065,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581718437881422e-05,
+      "loss": 3.6105,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58087984313037e-05,
+      "loss": 3.602,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580041248379318e-05,
+      "loss": 3.5861,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579204291508639e-05,
+      "loss": 3.5963,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.578365696757587e-05,
+      "loss": 3.6063,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577527102006535e-05,
+      "loss": 3.5975,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.576688507255483e-05,
+      "loss": 3.6166,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5758515503848036e-05,
+      "loss": 3.6007,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.575012955633752e-05,
+      "loss": 3.5903,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5741743608827e-05,
+      "loss": 3.6196,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573335766131648e-05,
+      "loss": 3.5938,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572498809260969e-05,
+      "loss": 3.5922,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571660214509917e-05,
+      "loss": 3.6041,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570821619758865e-05,
+      "loss": 3.5972,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569983025007813e-05,
+      "loss": 3.5968,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569146068137134e-05,
+      "loss": 3.6006,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568307473386082e-05,
+      "loss": 3.5831,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56746887863503e-05,
+      "loss": 3.5935,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566630283883978e-05,
+      "loss": 3.6072,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565793327013299e-05,
+      "loss": 3.6025,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5649547322622476e-05,
+      "loss": 3.6114,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5641161375111956e-05,
+      "loss": 3.6022,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563277542760143e-05,
+      "loss": 3.6201,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5624405858894645e-05,
+      "loss": 3.5937,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5616019911384125e-05,
+      "loss": 3.6104,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5607633963873605e-05,
+      "loss": 3.6027,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559924801636308e-05,
+      "loss": 3.5872,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5590878447656294e-05,
+      "loss": 3.61,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5582492500145774e-05,
+      "loss": 3.6091,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5574106552635254e-05,
+      "loss": 3.6051,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556572060512473e-05,
+      "loss": 3.6064,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555735103641794e-05,
+      "loss": 3.5878,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554896508890743e-05,
+      "loss": 3.5841,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55405791413969e-05,
+      "loss": 3.5946,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553219319388638e-05,
+      "loss": 3.6045,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55238236251796e-05,
+      "loss": 3.5966,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551543767766908e-05,
+      "loss": 3.5935,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550705173015855e-05,
+      "loss": 3.5966,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549866578264803e-05,
+      "loss": 3.6006,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549029621394125e-05,
+      "loss": 3.5958,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548191026643072e-05,
+      "loss": 3.5977,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54735243189202e-05,
+      "loss": 3.6076,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.546513837140968e-05,
+      "loss": 3.6173,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54567688027029e-05,
+      "loss": 3.6089,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.544838285519238e-05,
+      "loss": 3.591,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543999690768186e-05,
+      "loss": 3.5949,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543161096017134e-05,
+      "loss": 3.5991,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.542324139146455e-05,
+      "loss": 3.6084,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5414855443954026e-05,
+      "loss": 3.5997,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5406469496443506e-05,
+      "loss": 3.598,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5398083548932986e-05,
+      "loss": 3.5971,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5389713980226195e-05,
+      "loss": 3.6052,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5381328032715675e-05,
+      "loss": 3.6034,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5372942085205155e-05,
+      "loss": 3.5956,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5364556137694635e-05,
+      "loss": 3.5904,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535618656898785e-05,
+      "loss": 3.5936,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534780062147733e-05,
+      "loss": 3.5975,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533941467396681e-05,
+      "loss": 3.5886,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533102872645629e-05,
+      "loss": 3.5959,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53226591577495e-05,
+      "loss": 3.5932,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531427321023898e-05,
+      "loss": 3.5997,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.530588726272846e-05,
+      "loss": 3.5929,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.529750131521794e-05,
+      "loss": 3.6065,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528913174651115e-05,
+      "loss": 3.6106,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528074579900063e-05,
+      "loss": 3.598,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527235985149011e-05,
+      "loss": 3.607,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5263990282783324e-05,
+      "loss": 3.5999,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5255604335272804e-05,
+      "loss": 3.5954,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5247218387762284e-05,
+      "loss": 3.5948,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5238832440251764e-05,
+      "loss": 3.5863,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523046287154497e-05,
+      "loss": 3.597,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522207692403445e-05,
+      "loss": 3.5983,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521369097652393e-05,
+      "loss": 3.6016,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520530502901341e-05,
+      "loss": 3.5928,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519693546030662e-05,
+      "loss": 3.6049,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51885495127961e-05,
+      "loss": 3.5971,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518016356528558e-05,
+      "loss": 3.5931,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517177761777506e-05,
+      "loss": 3.601,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516340804906828e-05,
+      "loss": 3.603,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515502210155776e-05,
+      "loss": 3.6079,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.514663615404724e-05,
+      "loss": 3.5898,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.513825020653672e-05,
+      "loss": 3.6135,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512988063782993e-05,
+      "loss": 3.5943,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512149469031941e-05,
+      "loss": 3.605,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.511310874280889e-05,
+      "loss": 3.6019,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.510472279529837e-05,
+      "loss": 3.5959,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5096353226591576e-05,
+      "loss": 3.5895,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5087967279081056e-05,
+      "loss": 3.6117,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5079581331570536e-05,
+      "loss": 3.585,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5071195384060016e-05,
+      "loss": 3.602,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506282581535323e-05,
+      "loss": 3.5967,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505443986784271e-05,
+      "loss": 3.597,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504605392033219e-05,
+      "loss": 3.5905,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.503766797282167e-05,
+      "loss": 3.598,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502929840411488e-05,
+      "loss": 3.5988,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502091245660436e-05,
+      "loss": 3.6062,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501252650909384e-05,
+      "loss": 3.5951,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.83720064163208,
+      "eval_runtime": 302.3365,
+      "eval_samples_per_second": 1262.14,
+      "eval_steps_per_second": 39.443,
+      "step": 915830
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.500414056158332e-05,
+      "loss": 3.5839,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.49957546140728e-05,
+      "loss": 3.5875,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498736866656228e-05,
+      "loss": 3.5937,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497898271905176e-05,
+      "loss": 3.5996,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497061315034497e-05,
+      "loss": 3.6001,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496222720283445e-05,
+      "loss": 3.5926,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495384125532393e-05,
+      "loss": 3.5992,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4945455307813416e-05,
+      "loss": 3.5946,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4937085739106625e-05,
+      "loss": 3.586,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4928699791596105e-05,
+      "loss": 3.5806,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4920313844085585e-05,
+      "loss": 3.6026,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4911927896575065e-05,
+      "loss": 3.6026,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4903574706672e-05,
+      "loss": 3.5944,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489518875916148e-05,
+      "loss": 3.5739,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.488680281165096e-05,
+      "loss": 3.5951,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487841686414044e-05,
+      "loss": 3.5759,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487003091662992e-05,
+      "loss": 3.5922,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48616449691194e-05,
+      "loss": 3.5898,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.485325902160888e-05,
+      "loss": 3.5869,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484487307409837e-05,
+      "loss": 3.5964,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483650350539158e-05,
+      "loss": 3.6021,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482811755788106e-05,
+      "loss": 3.5907,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481973161037054e-05,
+      "loss": 3.5896,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481134566286002e-05,
+      "loss": 3.5855,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480297609415323e-05,
+      "loss": 3.5942,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479459014664271e-05,
+      "loss": 3.5912,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478620419913219e-05,
+      "loss": 3.5827,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477781825162167e-05,
+      "loss": 3.5891,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4769448682914877e-05,
+      "loss": 3.5883,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4761062735404357e-05,
+      "loss": 3.586,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4752676787893837e-05,
+      "loss": 3.5858,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.474429084038332e-05,
+      "loss": 3.593,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.473592127167653e-05,
+      "loss": 3.5886,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.472753532416601e-05,
+      "loss": 3.5984,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.471914937665549e-05,
+      "loss": 3.5874,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.471076342914497e-05,
+      "loss": 3.5861,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470239386043818e-05,
+      "loss": 3.5964,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469400791292766e-05,
+      "loss": 3.5803,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468562196541714e-05,
+      "loss": 3.5804,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467723601790662e-05,
+      "loss": 3.5856,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466886644919983e-05,
+      "loss": 3.5808,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466048050168931e-05,
+      "loss": 3.5825,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465209455417879e-05,
+      "loss": 3.5868,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464370860666828e-05,
+      "loss": 3.5891,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4635339037961486e-05,
+      "loss": 3.5905,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4626953090450966e-05,
+      "loss": 3.5927,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4618567142940446e-05,
+      "loss": 3.5891,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4610181195429926e-05,
+      "loss": 3.5905,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4601811626723135e-05,
+      "loss": 3.5696,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4593425679212615e-05,
+      "loss": 3.5951,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4585039731702095e-05,
+      "loss": 3.5824,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4576653784191575e-05,
+      "loss": 3.5847,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4568284215484784e-05,
+      "loss": 3.5893,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4559898267974264e-05,
+      "loss": 3.5826,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4551512320463744e-05,
+      "loss": 3.5684,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.454314275175696e-05,
+      "loss": 3.5796,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.453475680424644e-05,
+      "loss": 3.5858,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.452637085673592e-05,
+      "loss": 3.5756,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45179849092254e-05,
+      "loss": 3.5956,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450961534051861e-05,
+      "loss": 3.5856,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.450122939300809e-05,
+      "loss": 3.5676,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449284344549757e-05,
+      "loss": 3.5994,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448445749798705e-05,
+      "loss": 3.5746,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447608792928026e-05,
+      "loss": 3.5727,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.446770198176974e-05,
+      "loss": 3.5853,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445931603425922e-05,
+      "loss": 3.5769,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44509300867487e-05,
+      "loss": 3.5743,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4442560518041913e-05,
+      "loss": 3.582,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443417457053139e-05,
+      "loss": 3.5677,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442578862302087e-05,
+      "loss": 3.5691,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4417402675510346e-05,
+      "loss": 3.5908,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440903310680356e-05,
+      "loss": 3.5846,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440064715929304e-05,
+      "loss": 3.5873,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439226121178252e-05,
+      "loss": 3.5843,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4383875264271995e-05,
+      "loss": 3.5989,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.437550569556521e-05,
+      "loss": 3.5746,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436711974805469e-05,
+      "loss": 3.5935,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435873380054417e-05,
+      "loss": 3.5801,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435034785303365e-05,
+      "loss": 3.572,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434197828432687e-05,
+      "loss": 3.5858,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433359233681635e-05,
+      "loss": 3.594,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432520638930582e-05,
+      "loss": 3.5822,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.43168204417953e-05,
+      "loss": 3.5919,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4308450873088516e-05,
+      "loss": 3.5683,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4300064925577996e-05,
+      "loss": 3.5656,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429167897806747e-05,
+      "loss": 3.5752,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.428329303055695e-05,
+      "loss": 3.5867,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4274923461850165e-05,
+      "loss": 3.578,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4266537514339645e-05,
+      "loss": 3.5759,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4258151566829125e-05,
+      "loss": 3.5792,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4249765619318605e-05,
+      "loss": 3.5823,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424139605061182e-05,
+      "loss": 3.5737,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4233010103101294e-05,
+      "loss": 3.5777,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4224624155590774e-05,
+      "loss": 3.5859,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4216238208080254e-05,
+      "loss": 3.5968,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420786863937347e-05,
+      "loss": 3.5898,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419948269186294e-05,
+      "loss": 3.5706,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419109674435242e-05,
+      "loss": 3.5776,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41827107968419e-05,
+      "loss": 3.5785,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417434122813511e-05,
+      "loss": 3.5882,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41659552806246e-05,
+      "loss": 3.5829,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.415756933311408e-05,
+      "loss": 3.5778,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414918338560356e-05,
+      "loss": 3.579,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414081381689677e-05,
+      "loss": 3.5841,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413242786938625e-05,
+      "loss": 3.5855,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.412404192187573e-05,
+      "loss": 3.5786,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.411565597436521e-05,
+      "loss": 3.5729,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4107286405658417e-05,
+      "loss": 3.5727,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4098900458147897e-05,
+      "loss": 3.576,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4090514510637376e-05,
+      "loss": 3.5727,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4082128563126856e-05,
+      "loss": 3.5741,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4073758994420066e-05,
+      "loss": 3.5741,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406537304690955e-05,
+      "loss": 3.5814,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405698709939903e-05,
+      "loss": 3.5715,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404860115188851e-05,
+      "loss": 3.5855,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404023158318172e-05,
+      "loss": 3.595,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40318456356712e-05,
+      "loss": 3.5787,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402345968816068e-05,
+      "loss": 3.5844,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401507374065016e-05,
+      "loss": 3.5828,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400670417194337e-05,
+      "loss": 3.5752,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399831822443285e-05,
+      "loss": 3.5783,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398993227692233e-05,
+      "loss": 3.563,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398154632941181e-05,
+      "loss": 3.5824,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397317676070502e-05,
+      "loss": 3.5775,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39647908131945e-05,
+      "loss": 3.5834,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3956404865683986e-05,
+      "loss": 3.5728,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3948018918173466e-05,
+      "loss": 3.5872,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3939649349466675e-05,
+      "loss": 3.5781,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3931263401956155e-05,
+      "loss": 3.5746,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3922877454445635e-05,
+      "loss": 3.5859,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3914491506935115e-05,
+      "loss": 3.5823,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3906121938228324e-05,
+      "loss": 3.5907,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3897735990717804e-05,
+      "loss": 3.574,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3889350043207284e-05,
+      "loss": 3.5919,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.388098047450049e-05,
+      "loss": 3.5777,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387259452698997e-05,
+      "loss": 3.5841,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.386420857947945e-05,
+      "loss": 3.5858,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.385582263196894e-05,
+      "loss": 3.5769,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.384743668445842e-05,
+      "loss": 3.5726,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383906711575163e-05,
+      "loss": 3.5918,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383068116824111e-05,
+      "loss": 3.5695,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382229522073059e-05,
+      "loss": 3.5807,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38139256520238e-05,
+      "loss": 3.5799,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380553970451328e-05,
+      "loss": 3.5786,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379715375700276e-05,
+      "loss": 3.5754,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378876780949224e-05,
+      "loss": 3.5767,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3780398240785447e-05,
+      "loss": 3.5815,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3772012293274927e-05,
+      "loss": 3.5832,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3763626345764407e-05,
+      "loss": 3.5801,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.837082862854004,
+      "eval_runtime": 360.0266,
+      "eval_samples_per_second": 1059.897,
+      "eval_steps_per_second": 33.123,
+      "step": 992150
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 6.899544304446044e+17,
+  "trial_name": null,
+  "trial_params": null
+}