diff --git "a/checkpoint-457914/trainer_state.json" "b/checkpoint-457914/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-457914/trainer_state.json"
@@ -0,0 +1,5437 @@
+{
+  "best_metric": 3.919722318649292,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/pp-mod-subj/transformer/0/checkpoints/checkpoint-457914",
+  "epoch": 0.025000278439663435,
+  "eval_steps": 10,
+  "global_step": 457914,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.9248,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8446,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1909,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9688,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8177,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.6907,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.5974,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5201,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4566,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.3848,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3377,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.2926,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.2597,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.1964,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1551,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.127,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.0856,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0578,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0367,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0035,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983231380739706e-05,
+      "loss": 4.9911,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9647,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9332,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9263,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.8997,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.8761,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8612,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8522,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8181,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8087,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.7864,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7768,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7772,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 4.7543,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7432,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7315,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7236,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 4.7098,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 4.6959,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 4.6883,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 4.6707,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 4.6663,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 4.6414,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6486,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6443,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 4.6319,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 4.6124,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6117,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.6073,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 4.5876,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 4.578,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.5738,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.583,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 4.5493,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 4.5732,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 4.5352,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5441,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 4.5357,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 4.5319,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 4.5215,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5167,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.4999,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 4.5103,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 4.4935,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4955,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.4901,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 4.4795,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 4.4624,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4754,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9421582546222625e-05,
+      "loss": 4.4665,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4523,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4574,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4485,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4488,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4501,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4431,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.4345,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.419,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4263,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.4221,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.4254,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.4202,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4047,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 4.39,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.3942,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 4.3955,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 4.3955,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 4.3922,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.3856,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925391273242342e-05,
+      "loss": 4.3756,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92455267849129e-05,
+      "loss": 4.3721,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923714083740238e-05,
+      "loss": 4.3716,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 4.378,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 4.368,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.3452,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.3519,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.3568,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.3588,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.3504,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.3464,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.3429,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.3379,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3427,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913657498249106e-05,
+      "loss": 4.3405,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912820541378427e-05,
+      "loss": 4.3194,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911981946627375e-05,
+      "loss": 4.3303,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911143351876323e-05,
+      "loss": 4.3093,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.3194,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3191,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.312,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.3133,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.301,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.2962,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3028,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.3145,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.2876,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027606801265495e-05,
+      "loss": 4.2832,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.2858,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.2988,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.2871,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899407939002715e-05,
+      "loss": 4.2886,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985693442516624e-05,
+      "loss": 4.284,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977307495006104e-05,
+      "loss": 4.2864,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968921547495584e-05,
+      "loss": 4.2774,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.277,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2714,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2806,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893539413625723e-05,
+      "loss": 4.2595,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.2724,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2663,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2552,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890186672501889e-05,
+      "loss": 4.2668,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889348077750837e-05,
+      "loss": 4.2602,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888509482999785e-05,
+      "loss": 4.2646,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887670888248733e-05,
+      "loss": 4.2618,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886833931378054e-05,
+      "loss": 4.243,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885995336627002e-05,
+      "loss": 4.2421,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515674187595e-05,
+      "loss": 4.2444,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2538,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834811902542186e-05,
+      "loss": 4.2543,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8826425955031666e-05,
+      "loss": 4.2484,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8818040007521146e-05,
+      "loss": 4.2356,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8809654060010626e-05,
+      "loss": 4.2374,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8801268112500106e-05,
+      "loss": 4.2385,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879289854379332e-05,
+      "loss": 4.2352,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878452897508653e-05,
+      "loss": 4.2312,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877614302757601e-05,
+      "loss": 4.2485,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876775708006549e-05,
+      "loss": 4.2321,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875937113255497e-05,
+      "loss": 4.2237,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875098518504445e-05,
+      "loss": 4.2225,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.238885879516602,
+      "eval_runtime": 600.8169,
+      "eval_samples_per_second": 635.12,
+      "eval_steps_per_second": 19.848,
+      "step": 76319
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.874259923753393e-05,
+      "loss": 4.2297,
+      "step": 76800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.873421329002341e-05,
+      "loss": 4.2254,
+      "step": 77312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.872582734251289e-05,
+      "loss": 4.2143,
+      "step": 77824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.871744139500237e-05,
+      "loss": 4.2142,
+      "step": 78336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870905544749185e-05,
+      "loss": 4.207,
+      "step": 78848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870066949998133e-05,
+      "loss": 4.2042,
+      "step": 79360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.869228355247081e-05,
+      "loss": 4.1988,
+      "step": 79872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.868389760496029e-05,
+      "loss": 4.1998,
+      "step": 80384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.867551165744977e-05,
+      "loss": 4.2057,
+      "step": 80896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.866712570993925e-05,
+      "loss": 4.2033,
+      "step": 81408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865873976242873e-05,
+      "loss": 4.1984,
+      "step": 81920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865035381491821e-05,
+      "loss": 4.2014,
+      "step": 82432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.864198424621142e-05,
+      "loss": 4.1945,
+      "step": 82944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.86335982987009e-05,
+      "loss": 4.1894,
+      "step": 83456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.862521235119038e-05,
+      "loss": 4.1864,
+      "step": 83968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.1594,
+      "step": 84480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.1812,
+      "step": 84992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860007088746255e-05,
+      "loss": 4.1806,
+      "step": 85504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.859168493995203e-05,
+      "loss": 4.1804,
+      "step": 86016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8583315371245244e-05,
+      "loss": 4.1917,
+      "step": 86528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8574929423734724e-05,
+      "loss": 4.1742,
+      "step": 87040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8566543476224204e-05,
+      "loss": 4.1773,
+      "step": 87552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8558157528713684e-05,
+      "loss": 4.1781,
+      "step": 88064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.854978796000689e-05,
+      "loss": 4.1733,
+      "step": 88576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.854140201249637e-05,
+      "loss": 4.1666,
+      "step": 89088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.853301606498585e-05,
+      "loss": 4.1633,
+      "step": 89600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.852463011747533e-05,
+      "loss": 4.171,
+      "step": 90112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.851626054876854e-05,
+      "loss": 4.156,
+      "step": 90624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.850787460125802e-05,
+      "loss": 4.162,
+      "step": 91136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.84994886537475e-05,
+      "loss": 4.1472,
+      "step": 91648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.849110270623698e-05,
+      "loss": 4.1507,
+      "step": 92160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.84827331375302e-05,
+      "loss": 4.1631,
+      "step": 92672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.847434719001968e-05,
+      "loss": 4.1554,
+      "step": 93184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.846596124250916e-05,
+      "loss": 4.1553,
+      "step": 93696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.845757529499864e-05,
+      "loss": 4.1517,
+      "step": 94208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.844920572629185e-05,
+      "loss": 4.1513,
+      "step": 94720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.844081977878133e-05,
+      "loss": 4.1534,
+      "step": 95232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.843243383127081e-05,
+      "loss": 4.149,
+      "step": 95744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.842404788376029e-05,
+      "loss": 4.1474,
+      "step": 96256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8415678315053496e-05,
+      "loss": 4.1361,
+      "step": 96768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8407292367542976e-05,
+      "loss": 4.1442,
+      "step": 97280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8398906420032456e-05,
+      "loss": 4.1276,
+      "step": 97792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8390520472521936e-05,
+      "loss": 4.1409,
+      "step": 98304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.838215090381515e-05,
+      "loss": 4.1438,
+      "step": 98816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.837376495630463e-05,
+      "loss": 4.139,
+      "step": 99328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.836537900879411e-05,
+      "loss": 4.1311,
+      "step": 99840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.835699306128359e-05,
+      "loss": 4.1341,
+      "step": 100352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834860711377307e-05,
+      "loss": 4.1383,
+      "step": 100864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834023754506628e-05,
+      "loss": 4.1171,
+      "step": 101376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.833185159755576e-05,
+      "loss": 4.1214,
+      "step": 101888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.832346565004524e-05,
+      "loss": 4.1193,
+      "step": 102400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.831509608133845e-05,
+      "loss": 4.135,
+      "step": 102912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.830671013382793e-05,
+      "loss": 4.1121,
+      "step": 103424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.829832418631741e-05,
+      "loss": 4.1336,
+      "step": 103936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828993823880689e-05,
+      "loss": 4.1051,
+      "step": 104448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828155229129637e-05,
+      "loss": 4.1189,
+      "step": 104960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8273166343785856e-05,
+      "loss": 4.1179,
+      "step": 105472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8264780396275336e-05,
+      "loss": 4.1195,
+      "step": 105984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8256410827568545e-05,
+      "loss": 4.111,
+      "step": 106496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8248024880058025e-05,
+      "loss": 4.1157,
+      "step": 107008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8239638932547505e-05,
+      "loss": 4.1031,
+      "step": 107520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8231252985036985e-05,
+      "loss": 4.1129,
+      "step": 108032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8222883416330194e-05,
+      "loss": 4.1034,
+      "step": 108544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8214497468819674e-05,
+      "loss": 4.1089,
+      "step": 109056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8206111521309154e-05,
+      "loss": 4.1051,
+      "step": 109568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8197725573798634e-05,
+      "loss": 4.1022,
+      "step": 110080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818935600509184e-05,
+      "loss": 4.0856,
+      "step": 110592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818097005758132e-05,
+      "loss": 4.1094,
+      "step": 111104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.817258411007081e-05,
+      "loss": 4.0966,
+      "step": 111616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.816419816256029e-05,
+      "loss": 4.0914,
+      "step": 112128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.81558285938535e-05,
+      "loss": 4.0993,
+      "step": 112640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.814745902514671e-05,
+      "loss": 4.095,
+      "step": 113152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813907307763619e-05,
+      "loss": 4.0974,
+      "step": 113664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813068713012567e-05,
+      "loss": 4.1036,
+      "step": 114176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.812230118261515e-05,
+      "loss": 4.0999,
+      "step": 114688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.811393161390836e-05,
+      "loss": 4.0922,
+      "step": 115200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.810554566639784e-05,
+      "loss": 4.0834,
+      "step": 115712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.809715971888732e-05,
+      "loss": 4.0888,
+      "step": 116224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.80887737713768e-05,
+      "loss": 4.0909,
+      "step": 116736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808038782386628e-05,
+      "loss": 4.1005,
+      "step": 117248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.807200187635576e-05,
+      "loss": 4.0909,
+      "step": 117760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.806363230764897e-05,
+      "loss": 4.0851,
+      "step": 118272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.805524636013845e-05,
+      "loss": 4.0712,
+      "step": 118784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.804686041262793e-05,
+      "loss": 4.0757,
+      "step": 119296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.803847446511741e-05,
+      "loss": 4.079,
+      "step": 119808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8030088517606886e-05,
+      "loss": 4.0815,
+      "step": 120320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8021702570096366e-05,
+      "loss": 4.0868,
+      "step": 120832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8013316622585846e-05,
+      "loss": 4.0785,
+      "step": 121344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8004930675075325e-05,
+      "loss": 4.0721,
+      "step": 121856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7996561106368535e-05,
+      "loss": 4.0769,
+      "step": 122368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7988175158858015e-05,
+      "loss": 4.0702,
+      "step": 122880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7979789211347494e-05,
+      "loss": 4.0781,
+      "step": 123392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797140326383698e-05,
+      "loss": 4.0757,
+      "step": 123904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.796301731632646e-05,
+      "loss": 4.057,
+      "step": 124416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.795463136881594e-05,
+      "loss": 4.0601,
+      "step": 124928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.794626180010915e-05,
+      "loss": 4.0652,
+      "step": 125440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.793787585259863e-05,
+      "loss": 4.0779,
+      "step": 125952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792948990508811e-05,
+      "loss": 4.0624,
+      "step": 126464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792110395757759e-05,
+      "loss": 4.0714,
+      "step": 126976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.791271801006707e-05,
+      "loss": 4.0631,
+      "step": 127488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.790433206255655e-05,
+      "loss": 4.0593,
+      "step": 128000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.789596249384976e-05,
+      "loss": 4.0676,
+      "step": 128512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.788757654633924e-05,
+      "loss": 4.0651,
+      "step": 129024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.787920697763245e-05,
+      "loss": 4.0522,
+      "step": 129536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7870821030121935e-05,
+      "loss": 4.0657,
+      "step": 130048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7862435082611415e-05,
+      "loss": 4.0461,
+      "step": 130560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7854049135100895e-05,
+      "loss": 4.0513,
+      "step": 131072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7845663187590375e-05,
+      "loss": 4.0597,
+      "step": 131584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7837277240079855e-05,
+      "loss": 4.0474,
+      "step": 132096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7828891292569335e-05,
+      "loss": 4.054,
+      "step": 132608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7820505345058815e-05,
+      "loss": 4.0458,
+      "step": 133120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7812119397548295e-05,
+      "loss": 4.0417,
+      "step": 133632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7803733450037775e-05,
+      "loss": 4.0478,
+      "step": 134144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.779534750252725e-05,
+      "loss": 4.0611,
+      "step": 134656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.778696155501673e-05,
+      "loss": 4.0406,
+      "step": 135168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777860836511367e-05,
+      "loss": 4.0414,
+      "step": 135680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777022241760315e-05,
+      "loss": 4.0391,
+      "step": 136192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.776183647009263e-05,
+      "loss": 4.0492,
+      "step": 136704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.775345052258212e-05,
+      "loss": 4.0463,
+      "step": 137216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.774508095387533e-05,
+      "loss": 4.0441,
+      "step": 137728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.773669500636481e-05,
+      "loss": 4.0454,
+      "step": 138240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.772830905885429e-05,
+      "loss": 4.047,
+      "step": 138752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771992311134377e-05,
+      "loss": 4.0418,
+      "step": 139264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771153716383325e-05,
+      "loss": 4.035,
+      "step": 139776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.770316759512646e-05,
+      "loss": 4.0389,
+      "step": 140288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.769478164761594e-05,
+      "loss": 4.0474,
+      "step": 140800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.768639570010542e-05,
+      "loss": 4.0311,
+      "step": 141312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.76780097525949e-05,
+      "loss": 4.0338,
+      "step": 141824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.766962380508437e-05,
+      "loss": 4.046,
+      "step": 142336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.766123785757386e-05,
+      "loss": 4.0239,
+      "step": 142848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.765285191006334e-05,
+      "loss": 4.0382,
+      "step": 143360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.764448234135655e-05,
+      "loss": 4.0327,
+      "step": 143872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7636096393846026e-05,
+      "loss": 4.0449,
+      "step": 144384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7627710446335506e-05,
+      "loss": 4.0373,
+      "step": 144896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7619324498824986e-05,
+      "loss": 4.0243,
+      "step": 145408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7610954930118195e-05,
+      "loss": 4.0216,
+      "step": 145920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7602568982607675e-05,
+      "loss": 4.0261,
+      "step": 146432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7594183035097155e-05,
+      "loss": 4.0381,
+      "step": 146944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7585797087586635e-05,
+      "loss": 4.0393,
+      "step": 147456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7577411140076115e-05,
+      "loss": 4.0362,
+      "step": 147968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7569041571369324e-05,
+      "loss": 4.0193,
+      "step": 148480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756065562385881e-05,
+      "loss": 4.0246,
+      "step": 148992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.755226967634829e-05,
+      "loss": 4.0305,
+      "step": 149504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.754388372883777e-05,
+      "loss": 4.0273,
+      "step": 150016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.753551416013098e-05,
+      "loss": 4.0213,
+      "step": 150528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.752712821262046e-05,
+      "loss": 4.0403,
+      "step": 151040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751874226510994e-05,
+      "loss": 4.0202,
+      "step": 151552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751035631759942e-05,
+      "loss": 4.0231,
+      "step": 152064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.750198674889263e-05,
+      "loss": 4.0185,
+      "step": 152576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.071280002593994,
+      "eval_runtime": 570.0359,
+      "eval_samples_per_second": 669.416,
+      "eval_steps_per_second": 20.92,
+      "step": 152638
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749360080138211e-05,
+      "loss": 4.0287,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748521485387159e-05,
+      "loss": 4.0254,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747682890636107e-05,
+      "loss": 4.0146,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746844295885055e-05,
+      "loss": 4.0153,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746005701134003e-05,
+      "loss": 4.0079,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745167106382951e-05,
+      "loss": 4.009,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744328511631899e-05,
+      "loss": 4.0078,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434899168808475e-05,
+      "loss": 4.0068,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426529600101684e-05,
+      "loss": 4.0118,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418143652591164e-05,
+      "loss": 4.011,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409757705080644e-05,
+      "loss": 4.0101,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401371757570124e-05,
+      "loss": 4.0098,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739300218886333e-05,
+      "loss": 4.0051,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738461624135281e-05,
+      "loss": 4.0004,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737623029384229e-05,
+      "loss": 4.0033,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736784434633177e-05,
+      "loss": 3.9751,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735945839882125e-05,
+      "loss": 4.0009,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735107245131073e-05,
+      "loss": 3.9977,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734268650380021e-05,
+      "loss": 3.9964,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733430055628969e-05,
+      "loss": 4.0149,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732593098758291e-05,
+      "loss": 3.9962,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731754504007238e-05,
+      "loss": 4.0006,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730915909256186e-05,
+      "loss": 3.9992,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730077314505134e-05,
+      "loss": 3.9953,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729238719754082e-05,
+      "loss": 3.9918,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728401762883403e-05,
+      "loss": 3.9867,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727563168132351e-05,
+      "loss": 3.998,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726724573381299e-05,
+      "loss": 3.9847,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725885978630247e-05,
+      "loss": 3.9902,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725047383879195e-05,
+      "loss": 3.9749,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724210427008517e-05,
+      "loss": 3.9835,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233718322574647e-05,
+      "loss": 3.9952,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225332375064127e-05,
+      "loss": 3.984,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216946427553606e-05,
+      "loss": 3.9903,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208576858846816e-05,
+      "loss": 3.984,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7200190911336296e-05,
+      "loss": 3.987,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191804963825775e-05,
+      "loss": 3.9912,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183419016315255e-05,
+      "loss": 3.9862,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175033068804735e-05,
+      "loss": 3.9827,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7166647121294215e-05,
+      "loss": 3.9737,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158277552587424e-05,
+      "loss": 3.9822,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149891605076904e-05,
+      "loss": 3.9697,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141505657566384e-05,
+      "loss": 3.9861,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133119710055864e-05,
+      "loss": 3.983,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712473376254535e-05,
+      "loss": 3.9828,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711634781503483e-05,
+      "loss": 3.9737,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710797824632804e-05,
+      "loss": 3.9735,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709959229881752e-05,
+      "loss": 3.9828,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091206351307e-05,
+      "loss": 3.9619,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708282040379648e-05,
+      "loss": 3.9673,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707443445628596e-05,
+      "loss": 3.9629,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706606488757917e-05,
+      "loss": 3.9826,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705767894006865e-05,
+      "loss": 3.9619,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704929299255813e-05,
+      "loss": 3.9813,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704090704504761e-05,
+      "loss": 3.9541,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703253747634082e-05,
+      "loss": 3.9656,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024151528830305e-05,
+      "loss": 3.9695,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015765581319785e-05,
+      "loss": 3.9701,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007379633809265e-05,
+      "loss": 3.9648,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6999010065102474e-05,
+      "loss": 3.9769,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990624117591954e-05,
+      "loss": 3.9487,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982238170081434e-05,
+      "loss": 3.9631,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973852222570914e-05,
+      "loss": 3.9601,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696548265386412e-05,
+      "loss": 3.968,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69570967063536e-05,
+      "loss": 3.9577,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694871075884308e-05,
+      "loss": 3.9603,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694032481133256e-05,
+      "loss": 3.9439,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693195524262577e-05,
+      "loss": 3.9645,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692356929511526e-05,
+      "loss": 3.9539,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691518334760474e-05,
+      "loss": 3.9522,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690679740009422e-05,
+      "loss": 3.9589,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689842783138743e-05,
+      "loss": 3.9563,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689004188387691e-05,
+      "loss": 3.9547,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688165593636639e-05,
+      "loss": 3.9686,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687326998885587e-05,
+      "loss": 3.9617,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864900420149076e-05,
+      "loss": 3.9554,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856514472638556e-05,
+      "loss": 3.947,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848128525128036e-05,
+      "loss": 3.9501,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683974257761751e-05,
+      "loss": 3.955,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831373008910725e-05,
+      "loss": 3.9619,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822987061400205e-05,
+      "loss": 3.9527,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681460111388969e-05,
+      "loss": 3.9545,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806215166379165e-05,
+      "loss": 3.9367,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679784559767238e-05,
+      "loss": 3.9461,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678945965016186e-05,
+      "loss": 3.9399,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678107370265134e-05,
+      "loss": 3.9494,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772687755140814e-05,
+      "loss": 3.9551,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764301807630294e-05,
+      "loss": 3.9475,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675593223892351e-05,
+      "loss": 3.9391,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674756267021672e-05,
+      "loss": 3.9464,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.67391767227062e-05,
+      "loss": 3.9395,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673079077519568e-05,
+      "loss": 3.9476,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672240482768516e-05,
+      "loss": 3.9461,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671401888017464e-05,
+      "loss": 3.9308,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670563293266412e-05,
+      "loss": 3.9309,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66972469851536e-05,
+      "loss": 3.9375,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668886103764308e-05,
+      "loss": 3.9492,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668049146893629e-05,
+      "loss": 3.9376,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667210552142577e-05,
+      "loss": 3.9475,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666371957391525e-05,
+      "loss": 3.9343,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665533362640473e-05,
+      "loss": 3.9359,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664696405769794e-05,
+      "loss": 3.9429,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663857811018742e-05,
+      "loss": 3.94,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663020854148063e-05,
+      "loss": 3.9296,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662182259397011e-05,
+      "loss": 3.9461,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661343664645959e-05,
+      "loss": 3.9199,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660505069894907e-05,
+      "loss": 3.9249,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659666475143855e-05,
+      "loss": 3.9394,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658829518273176e-05,
+      "loss": 3.923,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657990923522124e-05,
+      "loss": 3.937,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657152328771072e-05,
+      "loss": 3.9256,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65631373402002e-05,
+      "loss": 3.9197,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655475139268968e-05,
+      "loss": 3.9268,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654636544517916e-05,
+      "loss": 3.9445,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653797949766864e-05,
+      "loss": 3.9216,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652959355015812e-05,
+      "loss": 3.9258,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652122398145134e-05,
+      "loss": 3.9174,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651283803394082e-05,
+      "loss": 3.9269,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65044520864303e-05,
+      "loss": 3.932,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649606613891978e-05,
+      "loss": 3.9268,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487696570212986e-05,
+      "loss": 3.9254,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479310622702466e-05,
+      "loss": 3.928,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470924675191946e-05,
+      "loss": 3.9261,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462538727681426e-05,
+      "loss": 3.9233,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454169158974635e-05,
+      "loss": 3.9243,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445783211464115e-05,
+      "loss": 3.9311,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437397263953595e-05,
+      "loss": 3.9176,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429011316443075e-05,
+      "loss": 3.9198,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642064174773629e-05,
+      "loss": 3.9322,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641225580022577e-05,
+      "loss": 3.9107,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640386985271525e-05,
+      "loss": 3.9235,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639548390520473e-05,
+      "loss": 3.9224,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638711433649794e-05,
+      "loss": 3.9322,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637872838898742e-05,
+      "loss": 3.9225,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63703424414769e-05,
+      "loss": 3.9169,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636195649396638e-05,
+      "loss": 3.9096,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635357054645586e-05,
+      "loss": 3.9146,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634520097774907e-05,
+      "loss": 3.9258,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633681503023855e-05,
+      "loss": 3.9292,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632842908272803e-05,
+      "loss": 3.9275,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632004313521751e-05,
+      "loss": 3.912,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311673566510724e-05,
+      "loss": 3.911,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303287619000204e-05,
+      "loss": 3.9268,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294901671489684e-05,
+      "loss": 3.9161,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286515723979164e-05,
+      "loss": 3.9162,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6278146155272373e-05,
+      "loss": 3.9305,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626976020776185e-05,
+      "loss": 3.911,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626137426025133e-05,
+      "loss": 3.9166,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625298831274081e-05,
+      "loss": 3.9131,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9977993965148926,
+      "eval_runtime": 567.9532,
+      "eval_samples_per_second": 671.87,
+      "eval_steps_per_second": 20.996,
+      "step": 228957
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.624460236523029e-05,
+      "loss": 3.9213,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6236216417719766e-05,
+      "loss": 3.921,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6227830470209246e-05,
+      "loss": 3.9085,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6219444522698726e-05,
+      "loss": 3.9086,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.621105857518821e-05,
+      "loss": 3.9072,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.620267262767769e-05,
+      "loss": 3.9029,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.619428668016717e-05,
+      "loss": 3.9032,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.618590073265665e-05,
+      "loss": 3.9024,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.617753116394986e-05,
+      "loss": 3.9083,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616914521643934e-05,
+      "loss": 3.9115,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616075926892882e-05,
+      "loss": 3.9052,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.61523733214183e-05,
+      "loss": 3.9048,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.614400375271151e-05,
+      "loss": 3.9058,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.613561780520099e-05,
+      "loss": 3.9015,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.612723185769047e-05,
+      "loss": 3.8991,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611884591017995e-05,
+      "loss": 3.8769,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611045996266943e-05,
+      "loss": 3.8965,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.610207401515891e-05,
+      "loss": 3.8992,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.609368806764839e-05,
+      "loss": 3.8969,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.608530212013788e-05,
+      "loss": 3.9121,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6076932551431087e-05,
+      "loss": 3.8963,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6068546603920566e-05,
+      "loss": 3.9057,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6060160656410046e-05,
+      "loss": 3.8959,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6051791087703256e-05,
+      "loss": 3.8948,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6043405140192735e-05,
+      "loss": 3.8927,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6035019192682215e-05,
+      "loss": 3.8927,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6026633245171695e-05,
+      "loss": 3.8958,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6018247297661175e-05,
+      "loss": 3.889,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6009877728954384e-05,
+      "loss": 3.8905,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6001491781443864e-05,
+      "loss": 3.878,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5993105833933344e-05,
+      "loss": 3.8878,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598471988642283e-05,
+      "loss": 3.8957,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.597635031771604e-05,
+      "loss": 3.891,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.596796437020552e-05,
+      "loss": 3.8932,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5959578422695e-05,
+      "loss": 3.8894,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595119247518448e-05,
+      "loss": 3.892,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.594282290647769e-05,
+      "loss": 3.8927,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.593443695896717e-05,
+      "loss": 3.8971,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.592605101145665e-05,
+      "loss": 3.8864,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.591766506394613e-05,
+      "loss": 3.8799,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590929549523934e-05,
+      "loss": 3.8898,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590090954772882e-05,
+      "loss": 3.8726,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58925236002183e-05,
+      "loss": 3.8964,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5884137652707785e-05,
+      "loss": 3.8892,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5875768084000994e-05,
+      "loss": 3.8886,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5867382136490474e-05,
+      "loss": 3.8812,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5858996188979954e-05,
+      "loss": 3.8816,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585061024146943e-05,
+      "loss": 3.893,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584224067276264e-05,
+      "loss": 3.8691,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.583385472525212e-05,
+      "loss": 3.8771,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58254687777416e-05,
+      "loss": 3.8772,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5817082830231076e-05,
+      "loss": 3.8877,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580871326152429e-05,
+      "loss": 3.8688,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580032731401377e-05,
+      "loss": 3.8928,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.579194136650325e-05,
+      "loss": 3.8675,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.578355541899273e-05,
+      "loss": 3.8706,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.577518585028595e-05,
+      "loss": 3.8839,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.576679990277543e-05,
+      "loss": 3.8744,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.57584139552649e-05,
+      "loss": 3.8849,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.575002800775438e-05,
+      "loss": 3.8881,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5741658439047596e-05,
+      "loss": 3.8637,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5733272491537076e-05,
+      "loss": 3.8696,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.572488654402655e-05,
+      "loss": 3.8747,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.571650059651603e-05,
+      "loss": 3.8781,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5708131027809245e-05,
+      "loss": 3.8688,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5699745080298725e-05,
+      "loss": 3.8717,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5691359132788205e-05,
+      "loss": 3.857,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5682973185277685e-05,
+      "loss": 3.8796,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.56746036165709e-05,
+      "loss": 3.8664,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5666217669060374e-05,
+      "loss": 3.8653,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5657831721549854e-05,
+      "loss": 3.8722,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564946215284307e-05,
+      "loss": 3.8711,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564107620533255e-05,
+      "loss": 3.8678,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.563269025782202e-05,
+      "loss": 3.878,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.56243043103115e-05,
+      "loss": 3.8776,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561591836280098e-05,
+      "loss": 3.8714,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560753241529046e-05,
+      "loss": 3.8646,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559914646777994e-05,
+      "loss": 3.8667,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559076052026943e-05,
+      "loss": 3.8674,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.558239095156264e-05,
+      "loss": 3.8802,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.557400500405212e-05,
+      "loss": 3.8708,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.55656190565416e-05,
+      "loss": 3.8644,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.555723310903108e-05,
+      "loss": 3.856,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554886354032429e-05,
+      "loss": 3.8614,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554047759281377e-05,
+      "loss": 3.8582,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.553209164530325e-05,
+      "loss": 3.8656,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.552370569779273e-05,
+      "loss": 3.8655,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.551533612908594e-05,
+      "loss": 3.8672,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.550695018157542e-05,
+      "loss": 3.859,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549858061286863e-05,
+      "loss": 3.8679,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549019466535811e-05,
+      "loss": 3.8585,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548180871784759e-05,
+      "loss": 3.8616,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.547342277033707e-05,
+      "loss": 3.8631,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546503682282655e-05,
+      "loss": 3.8521,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.545665087531603e-05,
+      "loss": 3.85,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544826492780551e-05,
+      "loss": 3.8627,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543987898029499e-05,
+      "loss": 3.8602,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54315094115882e-05,
+      "loss": 3.8659,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.542312346407768e-05,
+      "loss": 3.8619,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.541473751656716e-05,
+      "loss": 3.8558,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.540635156905664e-05,
+      "loss": 3.8545,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.539798200034985e-05,
+      "loss": 3.8652,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538959605283933e-05,
+      "loss": 3.861,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5381226484132546e-05,
+      "loss": 3.8508,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5372840536622026e-05,
+      "loss": 3.8661,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5364454589111506e-05,
+      "loss": 3.8401,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5356068641600986e-05,
+      "loss": 3.8494,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5347682694090466e-05,
+      "loss": 3.8615,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5339313125383675e-05,
+      "loss": 3.8454,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5330927177873155e-05,
+      "loss": 3.8581,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5322541230362635e-05,
+      "loss": 3.8462,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5314155282852115e-05,
+      "loss": 3.8396,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5305769335341595e-05,
+      "loss": 3.8496,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5297383387831075e-05,
+      "loss": 3.8668,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5288997440320555e-05,
+      "loss": 3.8486,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5280611492810035e-05,
+      "loss": 3.8453,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.527224192410325e-05,
+      "loss": 3.8421,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.526385597659273e-05,
+      "loss": 3.849,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.525547002908221e-05,
+      "loss": 3.8541,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.524710046037542e-05,
+      "loss": 3.8511,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.52387145128649e-05,
+      "loss": 3.8465,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.523032856535438e-05,
+      "loss": 3.8535,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.522194261784386e-05,
+      "loss": 3.85,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.521355667033333e-05,
+      "loss": 3.8485,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.520517072282281e-05,
+      "loss": 3.8462,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.519678477531229e-05,
+      "loss": 3.8544,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.518839882780177e-05,
+      "loss": 3.8417,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.518002925909499e-05,
+      "loss": 3.8438,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517164331158447e-05,
+      "loss": 3.8565,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.516325736407395e-05,
+      "loss": 3.8371,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515487141656343e-05,
+      "loss": 3.8465,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.514650184785664e-05,
+      "loss": 3.8514,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.513811590034612e-05,
+      "loss": 3.8563,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51297299528356e-05,
+      "loss": 3.847,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512134400532508e-05,
+      "loss": 3.8455,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5112974436618287e-05,
+      "loss": 3.8365,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5104588489107766e-05,
+      "loss": 3.8403,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5096202541597246e-05,
+      "loss": 3.8499,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5087816594086726e-05,
+      "loss": 3.8535,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507944702537994e-05,
+      "loss": 3.8532,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507106107786942e-05,
+      "loss": 3.8405,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.50626751303589e-05,
+      "loss": 3.8354,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.505428918284838e-05,
+      "loss": 3.8521,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.504591961414159e-05,
+      "loss": 3.8421,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503753366663107e-05,
+      "loss": 3.8457,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502914771912055e-05,
+      "loss": 3.8584,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502076177161003e-05,
+      "loss": 3.8419,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.501239220290324e-05,
+      "loss": 3.843,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.500400625539272e-05,
+      "loss": 3.8409,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9589924812316895,
+      "eval_runtime": 577.3589,
+      "eval_samples_per_second": 660.925,
+      "eval_steps_per_second": 20.654,
+      "step": 305276
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49956203078822e-05,
+      "loss": 3.8517,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498723436037168e-05,
+      "loss": 3.8472,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497884841286116e-05,
+      "loss": 3.8369,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497046246535065e-05,
+      "loss": 3.8351,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496207651784013e-05,
+      "loss": 3.8381,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495369057032961e-05,
+      "loss": 3.8321,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945304622819087e-05,
+      "loss": 3.8343,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936918675308567e-05,
+      "loss": 3.8289,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928549106601776e-05,
+      "loss": 3.839,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4920163159091256e-05,
+      "loss": 3.8453,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911777211580736e-05,
+      "loss": 3.8328,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903391264070216e-05,
+      "loss": 3.8305,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4895021695363425e-05,
+      "loss": 3.8382,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886635747852905e-05,
+      "loss": 3.8342,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878249800342385e-05,
+      "loss": 3.8307,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869863852831864e-05,
+      "loss": 3.8066,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486149428412508e-05,
+      "loss": 3.8265,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485310833661456e-05,
+      "loss": 3.8295,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484472238910404e-05,
+      "loss": 3.8288,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483633644159352e-05,
+      "loss": 3.8469,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482796687288673e-05,
+      "loss": 3.8247,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481958092537621e-05,
+      "loss": 3.8378,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481119497786569e-05,
+      "loss": 3.8254,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480280903035516e-05,
+      "loss": 3.8305,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479443946164838e-05,
+      "loss": 3.8207,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478605351413786e-05,
+      "loss": 3.8251,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477766756662734e-05,
+      "loss": 3.8271,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476928161911682e-05,
+      "loss": 3.8227,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4760912050410034e-05,
+      "loss": 3.823,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752526102899514e-05,
+      "loss": 3.8132,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4744140155388994e-05,
+      "loss": 3.8159,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473575420787847e-05,
+      "loss": 3.8304,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472738463917168e-05,
+      "loss": 3.8246,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471899869166116e-05,
+      "loss": 3.8264,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710612744150636e-05,
+      "loss": 3.8229,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4702226796640116e-05,
+      "loss": 3.8284,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469385722793333e-05,
+      "loss": 3.8248,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468547128042281e-05,
+      "loss": 3.8308,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4677085332912285e-05,
+      "loss": 3.8222,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466869938540177e-05,
+      "loss": 3.8127,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466032981669499e-05,
+      "loss": 3.828,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465194386918447e-05,
+      "loss": 3.8065,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464355792167394e-05,
+      "loss": 3.8296,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463517197416342e-05,
+      "loss": 3.8225,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462680240545664e-05,
+      "loss": 3.8283,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461841645794611e-05,
+      "loss": 3.8151,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461003051043559e-05,
+      "loss": 3.814,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460164456292507e-05,
+      "loss": 3.827,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4593274994218286e-05,
+      "loss": 3.8096,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458488904670776e-05,
+      "loss": 3.8111,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457650309919724e-05,
+      "loss": 3.815,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4568133530490455e-05,
+      "loss": 3.8194,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455974758297994e-05,
+      "loss": 3.8102,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551361635469415e-05,
+      "loss": 3.8227,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542975687958895e-05,
+      "loss": 3.8041,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453460611925211e-05,
+      "loss": 3.8079,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4526220171741584e-05,
+      "loss": 3.8187,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517834224231064e-05,
+      "loss": 3.8076,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509448276720543e-05,
+      "loss": 3.8261,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450107870801376e-05,
+      "loss": 3.8276,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449269276050323e-05,
+      "loss": 3.7988,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448430681299271e-05,
+      "loss": 3.806,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447592086548219e-05,
+      "loss": 3.8135,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446755129677541e-05,
+      "loss": 3.8202,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445916534926489e-05,
+      "loss": 3.8046,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445077940175437e-05,
+      "loss": 3.8068,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444239345424385e-05,
+      "loss": 3.7998,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443402388553706e-05,
+      "loss": 3.8136,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442563793802654e-05,
+      "loss": 3.8059,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441725199051602e-05,
+      "loss": 3.8007,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44088660430055e-05,
+      "loss": 3.8115,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400496474298706e-05,
+      "loss": 3.8074,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4392110526788186e-05,
+      "loss": 3.8056,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4383724579277666e-05,
+      "loss": 3.8173,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4375338631767146e-05,
+      "loss": 3.8162,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436695268425663e-05,
+      "loss": 3.8104,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435856673674611e-05,
+      "loss": 3.8053,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435018078923559e-05,
+      "loss": 3.8094,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434179484172507e-05,
+      "loss": 3.8014,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433342527301828e-05,
+      "loss": 3.8236,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432503932550776e-05,
+      "loss": 3.8085,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431665337799724e-05,
+      "loss": 3.8058,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430826743048672e-05,
+      "loss": 3.7926,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429989786177993e-05,
+      "loss": 3.8057,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429151191426941e-05,
+      "loss": 3.7958,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428312596675889e-05,
+      "loss": 3.8087,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427474001924837e-05,
+      "loss": 3.8017,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266370450541586e-05,
+      "loss": 3.806,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257984503031066e-05,
+      "loss": 3.7994,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249614934324276e-05,
+      "loss": 3.812,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4241228986813756e-05,
+      "loss": 3.7985,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232843039303235e-05,
+      "loss": 3.8035,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224457091792715e-05,
+      "loss": 3.8082,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216071144282195e-05,
+      "loss": 3.7897,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207685196771675e-05,
+      "loss": 3.7904,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4199299249261155e-05,
+      "loss": 3.8046,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190913301750635e-05,
+      "loss": 3.7981,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182543733043844e-05,
+      "loss": 3.8135,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4174157785533324e-05,
+      "loss": 3.8019,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165771838022804e-05,
+      "loss": 3.7915,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157385890512284e-05,
+      "loss": 3.798,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41490163218055e-05,
+      "loss": 3.805,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414063037429498e-05,
+      "loss": 3.8026,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413226080558819e-05,
+      "loss": 3.7951,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412387485807767e-05,
+      "loss": 3.8061,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411548891056715e-05,
+      "loss": 3.7829,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410710296305663e-05,
+      "loss": 3.7914,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409871701554611e-05,
+      "loss": 3.8049,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409034744683932e-05,
+      "loss": 3.7834,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40819614993288e-05,
+      "loss": 3.8005,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407357555181828e-05,
+      "loss": 3.7958,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406518960430776e-05,
+      "loss": 3.7762,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056820035600974e-05,
+      "loss": 3.7926,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4048434088090454e-05,
+      "loss": 3.808,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4040048140579934e-05,
+      "loss": 3.7901,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031662193069414e-05,
+      "loss": 3.7923,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4023276245558894e-05,
+      "loss": 3.7861,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401489029804837e-05,
+      "loss": 3.7907,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400650435053785e-05,
+      "loss": 3.7992,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399813478183106e-05,
+      "loss": 3.7946,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398974883432054e-05,
+      "loss": 3.7887,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981362886810016e-05,
+      "loss": 3.7921,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3972976939299496e-05,
+      "loss": 3.7961,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3964590991788976e-05,
+      "loss": 3.7908,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3956205044278456e-05,
+      "loss": 3.7908,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394783547557167e-05,
+      "loss": 3.7969,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393944952806115e-05,
+      "loss": 3.7828,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393106358055063e-05,
+      "loss": 3.7931,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392267763304011e-05,
+      "loss": 3.7972,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391429168552959e-05,
+      "loss": 3.7842,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390590573801907e-05,
+      "loss": 3.79,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389751979050855e-05,
+      "loss": 3.7974,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388913384299803e-05,
+      "loss": 3.7972,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388076427429124e-05,
+      "loss": 3.7952,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387237832678072e-05,
+      "loss": 3.7897,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38639923792702e-05,
+      "loss": 3.7813,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385562281056341e-05,
+      "loss": 3.7862,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3847236863052896e-05,
+      "loss": 3.7927,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838850915542376e-05,
+      "loss": 3.798,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830464968031856e-05,
+      "loss": 3.8001,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3822095399325065e-05,
+      "loss": 3.7893,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813709451814545e-05,
+      "loss": 3.7793,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805323504304025e-05,
+      "loss": 3.7967,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796937556793505e-05,
+      "loss": 3.7846,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788567988086714e-05,
+      "loss": 3.7901,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3780182040576194e-05,
+      "loss": 3.8044,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771796093065674e-05,
+      "loss": 3.7913,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763410145555154e-05,
+      "loss": 3.7848,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375504057684836e-05,
+      "loss": 3.788,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.934941053390503,
+      "eval_runtime": 572.972,
+      "eval_samples_per_second": 665.985,
+      "eval_steps_per_second": 20.813,
+      "step": 381595
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374667100814158e-05,
+      "loss": 3.7973,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373828506063106e-05,
+      "loss": 3.7962,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372989911312054e-05,
+      "loss": 3.7814,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372151316561002e-05,
+      "loss": 3.7798,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.37131272180995e-05,
+      "loss": 3.788,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370474127058898e-05,
+      "loss": 3.7789,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369635532307846e-05,
+      "loss": 3.7798,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368796937556794e-05,
+      "loss": 3.7717,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367959980686115e-05,
+      "loss": 3.787,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367123023815436e-05,
+      "loss": 3.7967,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366284429064384e-05,
+      "loss": 3.7761,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365445834313332e-05,
+      "loss": 3.7778,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36460723956228e-05,
+      "loss": 3.7862,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363768644811228e-05,
+      "loss": 3.7791,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362930050060176e-05,
+      "loss": 3.7807,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362093093189497e-05,
+      "loss": 3.7549,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361254498438445e-05,
+      "loss": 3.7713,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360415903687393e-05,
+      "loss": 3.7756,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359577308936341e-05,
+      "loss": 3.7774,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358738714185289e-05,
+      "loss": 3.7964,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357900119434237e-05,
+      "loss": 3.7728,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3570615246831845e-05,
+      "loss": 3.7868,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3562229299321325e-05,
+      "loss": 3.7755,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355385973061454e-05,
+      "loss": 3.7782,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354547378310402e-05,
+      "loss": 3.7678,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35370878355935e-05,
+      "loss": 3.7731,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352871826688672e-05,
+      "loss": 3.7782,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35203323193762e-05,
+      "loss": 3.7732,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351194637186568e-05,
+      "loss": 3.7715,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350356042435515e-05,
+      "loss": 3.765,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3495190855648366e-05,
+      "loss": 3.7591,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3486804908137846e-05,
+      "loss": 3.7795,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347841896062732e-05,
+      "loss": 3.7772,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34700330131168e-05,
+      "loss": 3.7744,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3461663444410015e-05,
+      "loss": 3.7723,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3453277496899495e-05,
+      "loss": 3.7764,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3444891549388975e-05,
+      "loss": 3.7774,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3436505601878455e-05,
+      "loss": 3.774,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342813603317167e-05,
+      "loss": 3.773,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341975008566115e-05,
+      "loss": 3.7623,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3411364138150624e-05,
+      "loss": 3.7757,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402978190640104e-05,
+      "loss": 3.7586,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339460862193332e-05,
+      "loss": 3.778,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.338622267442279e-05,
+      "loss": 3.77,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337783672691227e-05,
+      "loss": 3.7786,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336945077940175e-05,
+      "loss": 3.766,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336108121069497e-05,
+      "loss": 3.7637,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335269526318445e-05,
+      "loss": 3.7807,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334430931567393e-05,
+      "loss": 3.7568,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333592336816341e-05,
+      "loss": 3.7614,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3327553799456624e-05,
+      "loss": 3.7653,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33191678519461e-05,
+      "loss": 3.7668,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331078190443558e-05,
+      "loss": 3.7626,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330239595692506e-05,
+      "loss": 3.7715,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3294026388218266e-05,
+      "loss": 3.7597,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3285640440707746e-05,
+      "loss": 3.7583,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3277254493197226e-05,
+      "loss": 3.7724,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3268868545686706e-05,
+      "loss": 3.7609,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3260498976979915e-05,
+      "loss": 3.7739,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32521130294694e-05,
+      "loss": 3.7773,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324372708195888e-05,
+      "loss": 3.7535,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323534113444836e-05,
+      "loss": 3.7559,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322697156574157e-05,
+      "loss": 3.761,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321858561823105e-05,
+      "loss": 3.7751,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321019967072053e-05,
+      "loss": 3.758,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320181372321001e-05,
+      "loss": 3.7609,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319344415450322e-05,
+      "loss": 3.7494,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31850582069927e-05,
+      "loss": 3.7663,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317667225948218e-05,
+      "loss": 3.7545,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316828631197166e-05,
+      "loss": 3.7548,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315991674326487e-05,
+      "loss": 3.7604,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3151530795754356e-05,
+      "loss": 3.7611,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3143144848243836e-05,
+      "loss": 3.7571,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3134758900733316e-05,
+      "loss": 3.7674,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3126389332026525e-05,
+      "loss": 3.7675,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3118003384516005e-05,
+      "loss": 3.762,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109617437005485e-05,
+      "loss": 3.7551,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3101231489494965e-05,
+      "loss": 3.7656,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3092861920788174e-05,
+      "loss": 3.7536,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3084475973277654e-05,
+      "loss": 3.7722,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3076090025767134e-05,
+      "loss": 3.7573,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067704078256614e-05,
+      "loss": 3.762,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305933450954982e-05,
+      "loss": 3.7462,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30509485620393e-05,
+      "loss": 3.7575,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304256261452879e-05,
+      "loss": 3.7477,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3034193045822e-05,
+      "loss": 3.7637,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.302580709831148e-05,
+      "loss": 3.7557,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.301742115080096e-05,
+      "loss": 3.7576,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300903520329044e-05,
+      "loss": 3.7562,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300066563458365e-05,
+      "loss": 3.7684,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299227968707313e-05,
+      "loss": 3.7443,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298389373956261e-05,
+      "loss": 3.7561,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297550779205209e-05,
+      "loss": 3.7606,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296712184454157e-05,
+      "loss": 3.7465,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295873589703105e-05,
+      "loss": 3.7457,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295034994952053e-05,
+      "loss": 3.76,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294196400201001e-05,
+      "loss": 3.7489,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293359443330322e-05,
+      "loss": 3.7698,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292522486459643e-05,
+      "loss": 3.754,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291683891708591e-05,
+      "loss": 3.7468,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290845296957539e-05,
+      "loss": 3.75,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290006702206487e-05,
+      "loss": 3.762,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289169745335808e-05,
+      "loss": 3.7584,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288331150584756e-05,
+      "loss": 3.7481,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287492555833704e-05,
+      "loss": 3.7593,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286653961082652e-05,
+      "loss": 3.7388,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2858153663316e-05,
+      "loss": 3.745,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284976771580548e-05,
+      "loss": 3.7569,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28413981470987e-05,
+      "loss": 3.7403,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283301219958818e-05,
+      "loss": 3.7537,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282462625207766e-05,
+      "loss": 3.7557,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281624030456714e-05,
+      "loss": 3.7296,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2807854357056617e-05,
+      "loss": 3.7481,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2799484788349826e-05,
+      "loss": 3.762,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2791098840839306e-05,
+      "loss": 3.7524,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2782712893328786e-05,
+      "loss": 3.7445,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2774326945818266e-05,
+      "loss": 3.7421,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2765940998307746e-05,
+      "loss": 3.7449,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2757555050797225e-05,
+      "loss": 3.754,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2749185482090435e-05,
+      "loss": 3.7513,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2740799534579915e-05,
+      "loss": 3.7434,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2732413587069394e-05,
+      "loss": 3.7493,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2724027639558874e-05,
+      "loss": 3.7494,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715641692048354e-05,
+      "loss": 3.7492,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2707255744537834e-05,
+      "loss": 3.7429,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2698869797027314e-05,
+      "loss": 3.7508,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2690483849516794e-05,
+      "loss": 3.7422,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268211428081e-05,
+      "loss": 3.7492,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267372833329948e-05,
+      "loss": 3.752,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266534238578896e-05,
+      "loss": 3.7379,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265695643827844e-05,
+      "loss": 3.7462,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264858686957165e-05,
+      "loss": 3.7566,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264020092206113e-05,
+      "loss": 3.7525,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263181497455062e-05,
+      "loss": 3.7507,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26234290270401e-05,
+      "loss": 3.7477,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261505945833331e-05,
+      "loss": 3.7344,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260667351082279e-05,
+      "loss": 3.7403,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.259828756331227e-05,
+      "loss": 3.7488,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258990161580175e-05,
+      "loss": 3.7608,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258153204709496e-05,
+      "loss": 3.7521,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257314609958444e-05,
+      "loss": 3.7424,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256476015207392e-05,
+      "loss": 3.7417,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25563742045634e-05,
+      "loss": 3.7481,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2548004635856606e-05,
+      "loss": 3.7444,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2539618688346086e-05,
+      "loss": 3.7438,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253123274083557e-05,
+      "loss": 3.7628,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252284679332505e-05,
+      "loss": 3.7464,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251447722461826e-05,
+      "loss": 3.7412,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250609127710774e-05,
+      "loss": 3.7444,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.919722318649292,
+      "eval_runtime": 569.191,
+      "eval_samples_per_second": 670.409,
+      "eval_steps_per_second": 20.951,
+      "step": 457914
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 2.968475802494792e+17,
+  "trial_name": null,
+  "trial_params": null
+}