diff --git "a/checkpoint-457920/trainer_state.json" "b/checkpoint-457920/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-457920/trainer_state.json"
@@ -0,0 +1,5437 @@
+{
+  "best_metric": 4.1629862785339355,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/binding-case/lstm/2/checkpoints/checkpoint-457920",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 457920,
+  "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": 10.8202,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5553,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0623,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9941,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9472,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9029,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7368,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6252,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5249,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4532,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.3903,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3296,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2678,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.2035,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1464,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.0874,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.0408,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985743889232116e-05,
+      "loss": 5.9952,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984905294481064e-05,
+      "loss": 5.9558,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984066699730012e-05,
+      "loss": 5.9191,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8856,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8505,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8194,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7915,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7743,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7353,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7141,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6887,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6688,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6429,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6182,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.6126,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5839,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.5749,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714926921053515e-05,
+      "loss": 5.5602,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706540973542995e-05,
+      "loss": 5.5357,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5246,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4853,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4867,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4507,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4514,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 5.4323,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 5.4321,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.4055,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.3987,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 5.3903,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.3724,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.3771,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3481,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3317,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 5.3419,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.336,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.3088,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.2988,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.2973,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.2805,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.295,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2557,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2554,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 5.248,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 5.236,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 5.2383,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2159,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.2093,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 5.2024,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 5.2154,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 5.1879,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 5.1856,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 5.1594,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 5.167,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 5.1745,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 5.1674,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.1439,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 5.1575,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 5.1534,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 5.1289,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362880913648985e-05,
+      "loss": 5.1265,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 5.1098,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 5.0966,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 5.0944,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 5.0978,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 5.0972,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 5.103,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 5.0877,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 5.0679,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 5.0684,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 5.0691,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 5.0567,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 5.049,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925392911122715e-05,
+      "loss": 5.0376,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 5.0369,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 5.039,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 5.0247,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 5.0286,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 5.015,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 5.0109,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.9984,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.996,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.9887,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.9917,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.9933,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.9709,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.9769,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.9642,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912818903498054e-05,
+      "loss": 4.9613,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911980308747002e-05,
+      "loss": 4.9545,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.9482,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.9476,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.9395,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.9365,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907787334991742e-05,
+      "loss": 4.9292,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90694874024069e-05,
+      "loss": 4.9193,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.9308,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.9186,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.9197,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.9151,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902759042246176e-05,
+      "loss": 4.916,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901920447495124e-05,
+      "loss": 4.9066,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901081852744072e-05,
+      "loss": 4.8912,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.9003,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.9049,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.8902,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977291116202375e-05,
+      "loss": 4.883,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968905168691855e-05,
+      "loss": 4.8822,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.8787,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.8699,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.8723,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.8611,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.8637,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.8533,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.871,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.8541,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.8452,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.8415,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.8438,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886830655617308e-05,
+      "loss": 4.8436,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.8439,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.8322,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884316509244525e-05,
+      "loss": 4.833,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883477914493473e-05,
+      "loss": 4.8285,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.8268,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.8145,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.8258,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.8122,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8792882164989586e-05,
+      "loss": 4.7994,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.813,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877611026996855e-05,
+      "loss": 4.8029,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876772432245803e-05,
+      "loss": 4.8033,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875933837494751e-05,
+      "loss": 4.8089,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8750952427436986e-05,
+      "loss": 4.7896,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.756977081298828,
+      "eval_runtime": 297.0801,
+      "eval_samples_per_second": 1284.472,
+      "eval_steps_per_second": 40.141,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8742566479926466e-05,
+      "loss": 4.7824,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.785,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872581096370916e-05,
+      "loss": 4.7887,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871744139500237e-05,
+      "loss": 4.7825,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870905544749185e-05,
+      "loss": 4.7872,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870066949998133e-05,
+      "loss": 4.7708,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869228355247081e-05,
+      "loss": 4.7773,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.868389760496029e-05,
+      "loss": 4.7645,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867551165744977e-05,
+      "loss": 4.773,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866712570993925e-05,
+      "loss": 4.7612,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865875614123246e-05,
+      "loss": 4.7583,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865037019372194e-05,
+      "loss": 4.7589,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642000625015155e-05,
+      "loss": 4.753,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633614677504635e-05,
+      "loss": 4.7484,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862522872999411e-05,
+      "loss": 4.7442,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861684278248359e-05,
+      "loss": 4.7312,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.7409,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860007088746255e-05,
+      "loss": 4.735,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859168493995203e-05,
+      "loss": 4.7282,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.7471,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.7299,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.733,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.7264,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549771581203164e-05,
+      "loss": 4.7431,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854140201249637e-05,
+      "loss": 4.7129,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853301606498585e-05,
+      "loss": 4.7131,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852463011747533e-05,
+      "loss": 4.7177,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851624416996481e-05,
+      "loss": 4.7192,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.7046,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.7045,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.7123,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.7097,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847433081121595e-05,
+      "loss": 4.711,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846594486370543e-05,
+      "loss": 4.7071,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845755891619491e-05,
+      "loss": 4.6983,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844917296868439e-05,
+      "loss": 4.7047,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84408033999776e-05,
+      "loss": 4.6755,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.6972,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842403150495656e-05,
+      "loss": 4.6709,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841564555744604e-05,
+      "loss": 4.6846,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840725960993552e-05,
+      "loss": 4.6764,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398873662425e-05,
+      "loss": 4.6888,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839048771491447e-05,
+      "loss": 4.6775,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838210176740395e-05,
+      "loss": 4.6735,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373732198697167e-05,
+      "loss": 4.6763,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365346251186646e-05,
+      "loss": 4.6684,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8356960303676126e-05,
+      "loss": 4.682,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8348574356165606e-05,
+      "loss": 4.6582,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8340188408655086e-05,
+      "loss": 4.6516,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331802461144566e-05,
+      "loss": 4.6727,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323416513634046e-05,
+      "loss": 4.6695,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315046944927255e-05,
+      "loss": 4.6538,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306660997416735e-05,
+      "loss": 4.6483,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298275049906215e-05,
+      "loss": 4.656,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289905481199424e-05,
+      "loss": 4.6427,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281519533688904e-05,
+      "loss": 4.6653,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273133586178384e-05,
+      "loss": 4.6398,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826474763866787e-05,
+      "loss": 4.6462,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825636169115735e-05,
+      "loss": 4.6412,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824797574364683e-05,
+      "loss": 4.634,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823960617494004e-05,
+      "loss": 4.6454,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823122022742952e-05,
+      "loss": 4.6314,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222834279919e-05,
+      "loss": 4.6268,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821444833240848e-05,
+      "loss": 4.6282,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820606238489796e-05,
+      "loss": 4.6439,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819767643738744e-05,
+      "loss": 4.6249,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818929048987692e-05,
+      "loss": 4.6211,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81809045423664e-05,
+      "loss": 4.6151,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817255135246334e-05,
+      "loss": 4.618,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816416540495282e-05,
+      "loss": 4.6373,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155779457442305e-05,
+      "loss": 4.6324,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147393509931785e-05,
+      "loss": 4.6143,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8139007562421265e-05,
+      "loss": 4.6287,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8130621614910745e-05,
+      "loss": 4.632,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8122252046203954e-05,
+      "loss": 4.6175,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113866098693434e-05,
+      "loss": 4.6175,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105480151182914e-05,
+      "loss": 4.6072,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8097094203672393e-05,
+      "loss": 4.5947,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8088708256161873e-05,
+      "loss": 4.5987,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808032230865135e-05,
+      "loss": 4.6039,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8071936361140827e-05,
+      "loss": 4.6066,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063550413630307e-05,
+      "loss": 4.6237,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805518084492352e-05,
+      "loss": 4.609,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804681127621674e-05,
+      "loss": 4.59,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803842532870622e-05,
+      "loss": 4.5987,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80300393811957e-05,
+      "loss": 4.6011,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802165343368518e-05,
+      "loss": 4.5985,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801326748617466e-05,
+      "loss": 4.5894,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800488153866413e-05,
+      "loss": 4.5868,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799651196995735e-05,
+      "loss": 4.5881,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798812602244683e-05,
+      "loss": 4.5923,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79797400749363e-05,
+      "loss": 4.5833,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797135412742578e-05,
+      "loss": 4.5894,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7963000937522725e-05,
+      "loss": 4.5841,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795461499001221e-05,
+      "loss": 4.582,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794622904250169e-05,
+      "loss": 4.5716,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793784309499117e-05,
+      "loss": 4.5778,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792945714748065e-05,
+      "loss": 4.5731,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792107119997013e-05,
+      "loss": 4.5754,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912685252459605e-05,
+      "loss": 4.5787,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7904299304949085e-05,
+      "loss": 4.5653,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895913357438565e-05,
+      "loss": 4.5691,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887527409928045e-05,
+      "loss": 4.5652,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879141462417525e-05,
+      "loss": 4.5641,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870755514907005e-05,
+      "loss": 4.5573,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786240232500395e-05,
+      "loss": 4.5585,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785401637749343e-05,
+      "loss": 4.5627,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784563042998291e-05,
+      "loss": 4.5528,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783724448247239e-05,
+      "loss": 4.5567,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782885853496187e-05,
+      "loss": 4.5498,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782047258745135e-05,
+      "loss": 4.5415,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781208663994083e-05,
+      "loss": 4.5631,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780370069243031e-05,
+      "loss": 4.5487,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779533112372352e-05,
+      "loss": 4.5546,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786945176213e-05,
+      "loss": 4.549,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777855922870248e-05,
+      "loss": 4.5603,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777017328119196e-05,
+      "loss": 4.5498,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776178733368144e-05,
+      "loss": 4.5354,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775340138617092e-05,
+      "loss": 4.548,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77450154386604e-05,
+      "loss": 4.5517,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773666224875734e-05,
+      "loss": 4.5378,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772829268005055e-05,
+      "loss": 4.5408,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771990673254003e-05,
+      "loss": 4.5351,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771152078502951e-05,
+      "loss": 4.5433,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770313483751899e-05,
+      "loss": 4.5304,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769474889000847e-05,
+      "loss": 4.5364,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768636294249795e-05,
+      "loss": 4.5291,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767797699498743e-05,
+      "loss": 4.5349,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766959104747691e-05,
+      "loss": 4.5229,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766120509996639e-05,
+      "loss": 4.5462,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765281915245587e-05,
+      "loss": 4.5354,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764443320494535e-05,
+      "loss": 4.5208,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763606363623857e-05,
+      "loss": 4.5247,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762767768872805e-05,
+      "loss": 4.5259,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761929174121753e-05,
+      "loss": 4.5286,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761090579370701e-05,
+      "loss": 4.5371,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760253622500022e-05,
+      "loss": 4.5188,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75941502774897e-05,
+      "loss": 4.5247,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7585780708782906e-05,
+      "loss": 4.5206,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577394761272386e-05,
+      "loss": 4.5246,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569008813761866e-05,
+      "loss": 4.5155,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560622866251346e-05,
+      "loss": 4.524,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552236918740826e-05,
+      "loss": 4.5126,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543850971230306e-05,
+      "loss": 4.5083,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535465023719786e-05,
+      "loss": 4.5211,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527079076209266e-05,
+      "loss": 4.5121,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518693128698745e-05,
+      "loss": 4.513,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751032355999196e-05,
+      "loss": 4.5134,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7501937612481435e-05,
+      "loss": 4.5056,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.477574825286865,
+      "eval_runtime": 297.0621,
+      "eval_samples_per_second": 1284.549,
+      "eval_steps_per_second": 40.143,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493551664970914e-05,
+      "loss": 4.4977,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485165717460394e-05,
+      "loss": 4.4976,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476779769949874e-05,
+      "loss": 4.5111,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468393822439354e-05,
+      "loss": 4.5006,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460007874928834e-05,
+      "loss": 4.5118,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451621927418314e-05,
+      "loss": 4.4969,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443235979907794e-05,
+      "loss": 4.501,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434850032397274e-05,
+      "loss": 4.4915,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426464084886754e-05,
+      "loss": 4.5066,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418078137376234e-05,
+      "loss": 4.4929,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740969218986572e-05,
+      "loss": 4.4928,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74013062423552e-05,
+      "loss": 4.4918,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739293667364841e-05,
+      "loss": 4.4886,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738455072613789e-05,
+      "loss": 4.4846,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737616477862737e-05,
+      "loss": 4.4863,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736777883111685e-05,
+      "loss": 4.4744,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735939288360632e-05,
+      "loss": 4.4831,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73510069360958e-05,
+      "loss": 4.4739,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734262098858528e-05,
+      "loss": 4.4726,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733423504107476e-05,
+      "loss": 4.4999,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732584909356424e-05,
+      "loss": 4.4791,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731746314605372e-05,
+      "loss": 4.4753,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73090771985432e-05,
+      "loss": 4.4806,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730070762983642e-05,
+      "loss": 4.4943,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.72923216823259e-05,
+      "loss": 4.4687,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728393573481538e-05,
+      "loss": 4.4713,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727554978730486e-05,
+      "loss": 4.4727,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726716383979434e-05,
+      "loss": 4.4781,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725877789228382e-05,
+      "loss": 4.4625,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72503919447733e-05,
+      "loss": 4.467,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724200599726278e-05,
+      "loss": 4.4717,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723363642855599e-05,
+      "loss": 4.4739,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225266859849196e-05,
+      "loss": 4.4803,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216880912338676e-05,
+      "loss": 4.4712,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208494964828156e-05,
+      "loss": 4.4633,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720010901731764e-05,
+      "loss": 4.471,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719172306980712e-05,
+      "loss": 4.4517,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71833371222966e-05,
+      "loss": 4.4649,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717495117478608e-05,
+      "loss": 4.4457,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716656522727556e-05,
+      "loss": 4.4546,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715819565856877e-05,
+      "loss": 4.4561,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714980971105825e-05,
+      "loss": 4.4646,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714142376354773e-05,
+      "loss": 4.4558,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713303781603721e-05,
+      "loss": 4.4519,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712466824733042e-05,
+      "loss": 4.4541,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71162822998199e-05,
+      "loss": 4.4497,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710789635230938e-05,
+      "loss": 4.4615,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709951040479886e-05,
+      "loss": 4.4431,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709114083609208e-05,
+      "loss": 4.432,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708275488858156e-05,
+      "loss": 4.4573,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074368941071037e-05,
+      "loss": 4.4522,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7065982993560517e-05,
+      "loss": 4.4412,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7057613424853726e-05,
+      "loss": 4.4388,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7049227477343206e-05,
+      "loss": 4.4437,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040841529832686e-05,
+      "loss": 4.4328,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032471961125895e-05,
+      "loss": 4.4518,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024086013615375e-05,
+      "loss": 4.4335,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015700066104855e-05,
+      "loss": 4.4405,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007314118594335e-05,
+      "loss": 4.4331,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998928171083814e-05,
+      "loss": 4.4343,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990542223573294e-05,
+      "loss": 4.4365,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982156276062774e-05,
+      "loss": 4.4315,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973786707355983e-05,
+      "loss": 4.4264,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965400759845463e-05,
+      "loss": 4.4266,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695703119113868e-05,
+      "loss": 4.4399,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694864524362816e-05,
+      "loss": 4.4287,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694025929611763e-05,
+      "loss": 4.4238,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693187334860711e-05,
+      "loss": 4.4194,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692348740109659e-05,
+      "loss": 4.4221,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691510145358607e-05,
+      "loss": 4.4433,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690671550607555e-05,
+      "loss": 4.4348,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689832955856503e-05,
+      "loss": 4.4235,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688995998985825e-05,
+      "loss": 4.4357,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688157404234773e-05,
+      "loss": 4.4385,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687318809483721e-05,
+      "loss": 4.4213,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686480214732669e-05,
+      "loss": 4.4258,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68564325786199e-05,
+      "loss": 4.4208,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684804663110938e-05,
+      "loss": 4.4086,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683966068359886e-05,
+      "loss": 4.4177,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683127473608834e-05,
+      "loss": 4.4128,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682288878857782e-05,
+      "loss": 4.4205,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6814519219871026e-05,
+      "loss": 4.4343,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806133272360506e-05,
+      "loss": 4.4284,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6797747324849986e-05,
+      "loss": 4.4035,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789361377339466e-05,
+      "loss": 4.42,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678097542982895e-05,
+      "loss": 4.4154,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677258948231843e-05,
+      "loss": 4.4206,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676420353480791e-05,
+      "loss": 4.4114,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675583396610112e-05,
+      "loss": 4.404,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.67474480185906e-05,
+      "loss": 4.4118,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673906207108008e-05,
+      "loss": 4.4095,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673067612356956e-05,
+      "loss": 4.4063,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672229017605904e-05,
+      "loss": 4.4138,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671392060735225e-05,
+      "loss": 4.4065,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670553465984173e-05,
+      "loss": 4.4101,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669714871233121e-05,
+      "loss": 4.3996,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668876276482069e-05,
+      "loss": 4.4007,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668037681731017e-05,
+      "loss": 4.4011,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6672007248603386e-05,
+      "loss": 4.4081,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663621301092866e-05,
+      "loss": 4.408,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655235353582346e-05,
+      "loss": 4.3932,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664684940607182e-05,
+      "loss": 4.3984,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66384634585613e-05,
+      "loss": 4.3962,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663007751105078e-05,
+      "loss": 4.3986,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621707942343995e-05,
+      "loss": 4.3861,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661332199483347e-05,
+      "loss": 4.3883,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660493604732295e-05,
+      "loss": 4.4003,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659655009981243e-05,
+      "loss": 4.3908,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658816415230191e-05,
+      "loss": 4.3896,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579794583595124e-05,
+      "loss": 4.3874,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571408636084604e-05,
+      "loss": 4.3808,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563022688574084e-05,
+      "loss": 4.3965,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554636741063564e-05,
+      "loss": 4.3927,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546250793553044e-05,
+      "loss": 4.3884,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6537864846042524e-05,
+      "loss": 4.3887,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529478898532004e-05,
+      "loss": 4.4013,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6521092951021484e-05,
+      "loss": 4.3911,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651272338231469e-05,
+      "loss": 4.3778,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650433743480417e-05,
+      "loss": 4.3874,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649595148729365e-05,
+      "loss": 4.3921,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648756553978313e-05,
+      "loss": 4.382,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647919597107634e-05,
+      "loss": 4.3833,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647081002356582e-05,
+      "loss": 4.3802,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646242407605531e-05,
+      "loss": 4.3885,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645403812854479e-05,
+      "loss": 4.3743,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644565218103427e-05,
+      "loss": 4.3876,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643728261232748e-05,
+      "loss": 4.368,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642889666481696e-05,
+      "loss": 4.3854,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642051071730644e-05,
+      "loss": 4.369,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412141148599646e-05,
+      "loss": 4.3929,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403755201089126e-05,
+      "loss": 4.3848,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395369253578606e-05,
+      "loss": 4.3755,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6386983306068086e-05,
+      "loss": 4.3691,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378597358557566e-05,
+      "loss": 4.3782,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370211411047046e-05,
+      "loss": 4.3766,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361825463536526e-05,
+      "loss": 4.3883,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353439516026006e-05,
+      "loss": 4.3685,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345053568515486e-05,
+      "loss": 4.3799,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336667621004966e-05,
+      "loss": 4.3754,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632829805229818e-05,
+      "loss": 4.3824,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6319912104787655e-05,
+      "loss": 4.3649,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311526157277135e-05,
+      "loss": 4.3765,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303140209766615e-05,
+      "loss": 4.3704,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629477064105983e-05,
+      "loss": 4.3645,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286384693549304e-05,
+      "loss": 4.3775,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277998746038784e-05,
+      "loss": 4.3721,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269612798528264e-05,
+      "loss": 4.3629,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6261226851017744e-05,
+      "loss": 4.3735,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625284090350723e-05,
+      "loss": 4.3639,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.34278678894043,
+      "eval_runtime": 293.9914,
+      "eval_samples_per_second": 1297.967,
+      "eval_steps_per_second": 40.562,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624447133480044e-05,
+      "loss": 4.3606,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623608538728992e-05,
+      "loss": 4.3562,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622771581858313e-05,
+      "loss": 4.3664,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621932987107261e-05,
+      "loss": 4.3634,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621094392356209e-05,
+      "loss": 4.3755,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620255797605157e-05,
+      "loss": 4.3545,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619417202854105e-05,
+      "loss": 4.3625,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618578608103053e-05,
+      "loss": 4.355,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617740013352001e-05,
+      "loss": 4.3747,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616901418600949e-05,
+      "loss": 4.3541,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616062823849897e-05,
+      "loss": 4.3546,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615224229098845e-05,
+      "loss": 4.3608,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6143872722281664e-05,
+      "loss": 4.3503,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135486774771144e-05,
+      "loss": 4.3503,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127100827260624e-05,
+      "loss": 4.3515,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6118714879750104e-05,
+      "loss": 4.3433,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110328932239584e-05,
+      "loss": 4.3439,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6101942984729064e-05,
+      "loss": 4.3447,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093557037218544e-05,
+      "loss": 4.3406,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608517108970802e-05,
+      "loss": 4.3663,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607680152100123e-05,
+      "loss": 4.3456,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606841557349071e-05,
+      "loss": 4.3468,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606002962598019e-05,
+      "loss": 4.3495,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60516600572734e-05,
+      "loss": 4.3596,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604327410976288e-05,
+      "loss": 4.3416,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60349045410561e-05,
+      "loss": 4.3435,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602651859354558e-05,
+      "loss": 4.3422,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601813264603506e-05,
+      "loss": 4.3471,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600974669852454e-05,
+      "loss": 4.3354,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600136075101402e-05,
+      "loss": 4.3398,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599297480350349e-05,
+      "loss": 4.345,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598460523479671e-05,
+      "loss": 4.3458,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597621928728619e-05,
+      "loss": 4.3532,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596783333977567e-05,
+      "loss": 4.3467,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595944739226514e-05,
+      "loss": 4.3389,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595106144475462e-05,
+      "loss": 4.3428,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594270825485157e-05,
+      "loss": 4.3313,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593432230734105e-05,
+      "loss": 4.3363,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592593635983053e-05,
+      "loss": 4.3254,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591755041232001e-05,
+      "loss": 4.3286,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590916446480949e-05,
+      "loss": 4.3308,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5900778517298965e-05,
+      "loss": 4.3453,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5892392569788445e-05,
+      "loss": 4.3379,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5884006622277925e-05,
+      "loss": 4.3278,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5875620674767404e-05,
+      "loss": 4.332,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867251106060614e-05,
+      "loss": 4.3281,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858865158550094e-05,
+      "loss": 4.3396,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5850479211039573e-05,
+      "loss": 4.3229,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5842093263529053e-05,
+      "loss": 4.313,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583372369482227e-05,
+      "loss": 4.3367,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582533774731175e-05,
+      "loss": 4.3339,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581695179980123e-05,
+      "loss": 4.3187,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580856585229071e-05,
+      "loss": 4.3264,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580017990478019e-05,
+      "loss": 4.3276,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579179395726967e-05,
+      "loss": 4.3125,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578342438856288e-05,
+      "loss": 4.333,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577503844105236e-05,
+      "loss": 4.3142,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576665249354184e-05,
+      "loss": 4.3265,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575828292483505e-05,
+      "loss": 4.314,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574989697732453e-05,
+      "loss": 4.3251,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574151102981401e-05,
+      "loss": 4.3143,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733125082303494e-05,
+      "loss": 4.3182,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724739134792974e-05,
+      "loss": 4.3065,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716353187282454e-05,
+      "loss": 4.315,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5707967239771934e-05,
+      "loss": 4.3301,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699581292261414e-05,
+      "loss": 4.3111,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691195344750894e-05,
+      "loss": 4.3131,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5682809397240374e-05,
+      "loss": 4.3055,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567443982853358e-05,
+      "loss": 4.3104,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566605388102306e-05,
+      "loss": 4.3303,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565766793351254e-05,
+      "loss": 4.322,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564928198600202e-05,
+      "loss": 4.316,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5640896038491496e-05,
+      "loss": 4.3243,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632510090980976e-05,
+      "loss": 4.3252,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562412414347046e-05,
+      "loss": 4.3132,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561575457476368e-05,
+      "loss": 4.3141,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560736862725315e-05,
+      "loss": 4.3133,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559898267974263e-05,
+      "loss": 4.3018,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559059673223211e-05,
+      "loss": 4.3052,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558221078472159e-05,
+      "loss": 4.3035,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55738412160148e-05,
+      "loss": 4.3097,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556545526850428e-05,
+      "loss": 4.3244,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555706932099376e-05,
+      "loss": 4.3187,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554868337348324e-05,
+      "loss": 4.2947,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554029742597272e-05,
+      "loss": 4.3188,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55319114784622e-05,
+      "loss": 4.3082,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523541909755416e-05,
+      "loss": 4.3111,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5515155962244896e-05,
+      "loss": 4.3056,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5506786393538105e-05,
+      "loss": 4.2984,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498400446027585e-05,
+      "loss": 4.3066,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5490014498517065e-05,
+      "loss": 4.3015,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481628551006545e-05,
+      "loss": 4.3026,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5473242603496025e-05,
+      "loss": 4.3046,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5464856655985505e-05,
+      "loss": 4.3012,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5456470708474985e-05,
+      "loss": 4.309,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5448084760964465e-05,
+      "loss": 4.2952,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5439698813453945e-05,
+      "loss": 4.2965,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431312865943425e-05,
+      "loss": 4.2947,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422943297236634e-05,
+      "loss": 4.3073,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414557349726114e-05,
+      "loss": 4.3007,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406171402215594e-05,
+      "loss": 4.2935,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539778545470508e-05,
+      "loss": 4.2928,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538939950719456e-05,
+      "loss": 4.2966,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538101355968404e-05,
+      "loss": 4.2994,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537264399097725e-05,
+      "loss": 4.2884,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536425804346673e-05,
+      "loss": 4.284,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535587209595621e-05,
+      "loss": 4.2998,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534748614844569e-05,
+      "loss": 4.2875,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533910020093516e-05,
+      "loss": 4.2874,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533071425342464e-05,
+      "loss": 4.2844,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532232830591412e-05,
+      "loss": 4.2846,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53139423584036e-05,
+      "loss": 4.2968,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530557278969682e-05,
+      "loss": 4.2851,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52971868421863e-05,
+      "loss": 4.2879,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528880089467578e-05,
+      "loss": 4.2931,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528041494716526e-05,
+      "loss": 4.3019,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527204537845847e-05,
+      "loss": 4.2996,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526365943094795e-05,
+      "loss": 4.2769,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525527348343743e-05,
+      "loss": 4.2887,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524688753592691e-05,
+      "loss": 4.2888,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523853434602385e-05,
+      "loss": 4.2879,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523014839851333e-05,
+      "loss": 4.2894,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221762451002805e-05,
+      "loss": 4.2802,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213376503492285e-05,
+      "loss": 4.2908,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520499055598177e-05,
+      "loss": 4.2795,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519662098727499e-05,
+      "loss": 4.288,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518823503976446e-05,
+      "loss": 4.2747,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517984909225394e-05,
+      "loss": 4.29,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517147952354716e-05,
+      "loss": 4.2732,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516309357603663e-05,
+      "loss": 4.3037,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515470762852611e-05,
+      "loss": 4.2818,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514632168101559e-05,
+      "loss": 4.2846,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513793573350507e-05,
+      "loss": 4.2726,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512954978599455e-05,
+      "loss": 4.2862,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512116383848403e-05,
+      "loss": 4.2836,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511277789097351e-05,
+      "loss": 4.2922,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510439194346299e-05,
+      "loss": 4.2752,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509600599595247e-05,
+      "loss": 4.2851,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087620048441956e-05,
+      "loss": 4.2822,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079234100931436e-05,
+      "loss": 4.2913,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070864532224645e-05,
+      "loss": 4.2686,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062478584714125e-05,
+      "loss": 4.2838,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5054092637203605e-05,
+      "loss": 4.2768,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5045723068496814e-05,
+      "loss": 4.2761,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5037337120986294e-05,
+      "loss": 4.2833,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5028951173475774e-05,
+      "loss": 4.2803,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020565225965254e-05,
+      "loss": 4.2731,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012179278454734e-05,
+      "loss": 4.28,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500380970974794e-05,
+      "loss": 4.2748,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.259811878204346,
+      "eval_runtime": 295.959,
+      "eval_samples_per_second": 1289.338,
+      "eval_steps_per_second": 40.293,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499542376223742e-05,
+      "loss": 4.2695,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498703781472691e-05,
+      "loss": 4.2631,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497865186721639e-05,
+      "loss": 4.2769,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497026591970587e-05,
+      "loss": 4.2729,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496187997219535e-05,
+      "loss": 4.2866,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495349402468482e-05,
+      "loss": 4.2692,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49451080771743e-05,
+      "loss": 4.2696,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493672212966378e-05,
+      "loss": 4.2665,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492833618215326e-05,
+      "loss": 4.2832,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491995023464274e-05,
+      "loss": 4.2692,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491156428713222e-05,
+      "loss": 4.2649,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49031783396217e-05,
+      "loss": 4.2684,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489480877091491e-05,
+      "loss": 4.2629,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488642282340439e-05,
+      "loss": 4.2646,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487803687589387e-05,
+      "loss": 4.2679,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486965092838336e-05,
+      "loss": 4.2511,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486126498087284e-05,
+      "loss": 4.2584,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485287903336232e-05,
+      "loss": 4.2572,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48444930858518e-05,
+      "loss": 4.2543,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483610713834128e-05,
+      "loss": 4.2753,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482772119083076e-05,
+      "loss": 4.2657,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481935162212397e-05,
+      "loss": 4.2582,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481096567461345e-05,
+      "loss": 4.2653,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480257972710293e-05,
+      "loss": 4.2749,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479419377959241e-05,
+      "loss": 4.2556,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478580783208188e-05,
+      "loss": 4.2593,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477742188457136e-05,
+      "loss": 4.2567,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476903593706084e-05,
+      "loss": 4.2593,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476064998955033e-05,
+      "loss": 4.2493,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752280420843536e-05,
+      "loss": 4.2581,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4743894473333016e-05,
+      "loss": 4.2608,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4735508525822496e-05,
+      "loss": 4.2625,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4727122578311976e-05,
+      "loss": 4.2684,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4718753009605185e-05,
+      "loss": 4.2622,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710367062094665e-05,
+      "loss": 4.2524,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4701981114584145e-05,
+      "loss": 4.259,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693611545877354e-05,
+      "loss": 4.253,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685225598366834e-05,
+      "loss": 4.2519,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676839650856314e-05,
+      "loss": 4.2425,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466847008214953e-05,
+      "loss": 4.2463,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466008413463901e-05,
+      "loss": 4.2521,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465169818712849e-05,
+      "loss": 4.2617,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464331223961797e-05,
+      "loss": 4.254,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463492629210745e-05,
+      "loss": 4.2484,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462654034459693e-05,
+      "loss": 4.2509,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461815439708641e-05,
+      "loss": 4.2488,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460976844957589e-05,
+      "loss": 4.2587,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460138250206537e-05,
+      "loss": 4.2449,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459299655455485e-05,
+      "loss": 4.2306,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458461060704433e-05,
+      "loss": 4.2557,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457622465953381e-05,
+      "loss": 4.2543,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456785509082702e-05,
+      "loss": 4.2372,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45594691433165e-05,
+      "loss": 4.2448,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455108319580598e-05,
+      "loss": 4.2511,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542697248295465e-05,
+      "loss": 4.2328,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534327679588674e-05,
+      "loss": 4.2538,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4525941732078154e-05,
+      "loss": 4.2373,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517555784567634e-05,
+      "loss": 4.2464,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509169837057114e-05,
+      "loss": 4.2357,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450080026835032e-05,
+      "loss": 4.2444,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44924143208398e-05,
+      "loss": 4.239,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448402837332928e-05,
+      "loss": 4.2428,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447564242581876e-05,
+      "loss": 4.2284,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446725647830824e-05,
+      "loss": 4.24,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4458870530797716e-05,
+      "loss": 4.2494,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445050096209093e-05,
+      "loss": 4.237,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444211501458041e-05,
+      "loss": 4.2342,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44337290670699e-05,
+      "loss": 4.2298,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442534311955937e-05,
+      "loss": 4.2324,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441695717204885e-05,
+      "loss": 4.2502,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440857122453833e-05,
+      "loss": 4.2478,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440018527702781e-05,
+      "loss": 4.2406,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439179932951729e-05,
+      "loss": 4.2484,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438344613961424e-05,
+      "loss": 4.2508,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437506019210372e-05,
+      "loss": 4.2361,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436667424459319e-05,
+      "loss": 4.2375,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435828829708267e-05,
+      "loss": 4.2409,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434990234957216e-05,
+      "loss": 4.2277,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434151640206164e-05,
+      "loss": 4.2282,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333130454551117e-05,
+      "loss": 4.2332,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4324744507040597e-05,
+      "loss": 4.2334,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316374938333806e-05,
+      "loss": 4.2498,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4307988990823286e-05,
+      "loss": 4.2448,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299603043312766e-05,
+      "loss": 4.2206,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291217095802245e-05,
+      "loss": 4.2398,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4282847527095455e-05,
+      "loss": 4.2363,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274461579584935e-05,
+      "loss": 4.2385,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266075632074415e-05,
+      "loss": 4.2328,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257706063367624e-05,
+      "loss": 4.2242,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249320115857104e-05,
+      "loss": 4.2326,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424093416834659e-05,
+      "loss": 4.229,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423254822083607e-05,
+      "loss": 4.2293,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422416227332555e-05,
+      "loss": 4.2274,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421577632581503e-05,
+      "loss": 4.2307,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420739037830451e-05,
+      "loss": 4.2327,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419900443079399e-05,
+      "loss": 4.2306,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41906348620872e-05,
+      "loss": 4.22,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418224891457668e-05,
+      "loss": 4.2254,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417387934586989e-05,
+      "loss": 4.2313,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416549339835937e-05,
+      "loss": 4.2343,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415710745084885e-05,
+      "loss": 4.2198,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414872150333833e-05,
+      "loss": 4.2223,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414033555582781e-05,
+      "loss": 4.2244,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413194960831729e-05,
+      "loss": 4.2269,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123563660806775e-05,
+      "loss": 4.2186,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115194092099984e-05,
+      "loss": 4.2163,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4106808144589464e-05,
+      "loss": 4.2239,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409843857588267e-05,
+      "loss": 4.2204,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409005262837215e-05,
+      "loss": 4.2168,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408166668086163e-05,
+      "loss": 4.215,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407328073335111e-05,
+      "loss": 4.2147,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406489478584059e-05,
+      "loss": 4.2265,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405650883833007e-05,
+      "loss": 4.217,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404812289081955e-05,
+      "loss": 4.2162,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4039736943309026e-05,
+      "loss": 4.2241,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403135099579851e-05,
+      "loss": 4.2344,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402298142709173e-05,
+      "loss": 4.2296,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40145954795812e-05,
+      "loss": 4.209,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400620953207068e-05,
+      "loss": 4.2177,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399782358456016e-05,
+      "loss": 4.2209,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398943763704964e-05,
+      "loss": 4.2181,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398106806834285e-05,
+      "loss": 4.2232,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397268212083233e-05,
+      "loss": 4.2115,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396429617332181e-05,
+      "loss": 4.2208,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395591022581129e-05,
+      "loss": 4.2171,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39475406571045e-05,
+      "loss": 4.2198,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393915470959398e-05,
+      "loss": 4.2059,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3930768762083466e-05,
+      "loss": 4.2186,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922382814572946e-05,
+      "loss": 4.2136,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3914013245866155e-05,
+      "loss": 4.2286,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905627298355635e-05,
+      "loss": 4.2149,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897241350845115e-05,
+      "loss": 4.2207,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3888855403334595e-05,
+      "loss": 4.2056,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880469455824075e-05,
+      "loss": 4.2222,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872099887117284e-05,
+      "loss": 4.2192,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863713939606764e-05,
+      "loss": 4.2237,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855327992096244e-05,
+      "loss": 4.2062,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846942044585724e-05,
+      "loss": 4.2242,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383857247587893e-05,
+      "loss": 4.2124,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383018652836842e-05,
+      "loss": 4.227,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38218005808579e-05,
+      "loss": 4.2063,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381341463334738e-05,
+      "loss": 4.2189,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380502868583686e-05,
+      "loss": 4.2106,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379664273832634e-05,
+      "loss": 4.2073,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378825679081582e-05,
+      "loss": 4.2184,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37798708433053e-05,
+      "loss": 4.2159,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377150127459851e-05,
+      "loss": 4.2083,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376311532708799e-05,
+      "loss": 4.2124,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37547457583812e-05,
+      "loss": 4.2152,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.203825950622559,
+      "eval_runtime": 300.5419,
+      "eval_samples_per_second": 1269.677,
+      "eval_steps_per_second": 39.678,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374635981087068e-05,
+      "loss": 4.2061,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373797386336016e-05,
+      "loss": 4.1995,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372958791584964e-05,
+      "loss": 4.2126,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372120196833912e-05,
+      "loss": 4.2085,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.37128160208286e-05,
+      "loss": 4.2191,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3704430073318084e-05,
+      "loss": 4.2068,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3696044125807564e-05,
+      "loss": 4.2048,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368765817829704e-05,
+      "loss": 4.2061,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367927223078652e-05,
+      "loss": 4.2176,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3670886283276e-05,
+      "loss": 4.2047,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366250033576548e-05,
+      "loss": 4.2049,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365411438825496e-05,
+      "loss": 4.2044,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3645744819548166e-05,
+      "loss": 4.2046,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3637358872037646e-05,
+      "loss": 4.2023,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3628972924527126e-05,
+      "loss": 4.2065,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3620586977016606e-05,
+      "loss": 4.189,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3612201029506086e-05,
+      "loss": 4.1943,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3603815081995566e-05,
+      "loss": 4.1985,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359542913448505e-05,
+      "loss": 4.1897,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358704318697453e-05,
+      "loss": 4.2113,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357865723946401e-05,
+      "loss": 4.2048,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357027129195349e-05,
+      "loss": 4.1998,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35619017232467e-05,
+      "loss": 4.1977,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355351577573618e-05,
+      "loss": 4.2111,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354512982822566e-05,
+      "loss": 4.1979,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353674388071514e-05,
+      "loss": 4.2008,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352835793320462e-05,
+      "loss": 4.1939,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351998836449783e-05,
+      "loss": 4.198,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351160241698731e-05,
+      "loss": 4.1934,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350321646947679e-05,
+      "loss": 4.1949,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349483052196627e-05,
+      "loss": 4.1987,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348644457445575e-05,
+      "loss": 4.2017,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347805862694523e-05,
+      "loss": 4.2058,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346967267943471e-05,
+      "loss": 4.2058,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346128673192419e-05,
+      "loss": 4.1913,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34529171632174e-05,
+      "loss": 4.2018,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3444563973314344e-05,
+      "loss": 4.1952,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3436178025803824e-05,
+      "loss": 4.185,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3427792078293304e-05,
+      "loss": 4.1885,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3419406130782784e-05,
+      "loss": 4.1852,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3411020183272264e-05,
+      "loss": 4.1923,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402634235761744e-05,
+      "loss": 4.2054,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3394248288251224e-05,
+      "loss": 4.1899,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3385862340740704e-05,
+      "loss": 4.1911,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377476393230184e-05,
+      "loss": 4.1981,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3369090445719664e-05,
+      "loss": 4.1892,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3360704498209144e-05,
+      "loss": 4.1949,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3352318550698624e-05,
+      "loss": 4.1865,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334396536079557e-05,
+      "loss": 4.1713,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333557941328505e-05,
+      "loss": 4.1994,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332719346577452e-05,
+      "loss": 4.1913,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331882389706774e-05,
+      "loss": 4.1845,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331043794955722e-05,
+      "loss": 4.1846,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33020520020467e-05,
+      "loss": 4.1896,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329366605453618e-05,
+      "loss": 4.1749,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328528010702566e-05,
+      "loss": 4.1917,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327689415951514e-05,
+      "loss": 4.1836,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326850821200462e-05,
+      "loss": 4.1882,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32601222644941e-05,
+      "loss": 4.1824,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325173631698358e-05,
+      "loss": 4.1838,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324335036947306e-05,
+      "loss": 4.1799,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323496442196254e-05,
+      "loss": 4.1893,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322657847445202e-05,
+      "loss": 4.1718,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321820890574523e-05,
+      "loss": 4.1815,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3209822958234707e-05,
+      "loss": 4.1908,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3201453389527916e-05,
+      "loss": 4.1816,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3193067442017396e-05,
+      "loss": 4.1755,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3184681494506876e-05,
+      "loss": 4.177,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317629554699636e-05,
+      "loss": 4.1738,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316790959948584e-05,
+      "loss": 4.192,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315952365197532e-05,
+      "loss": 4.1916,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31511377044648e-05,
+      "loss": 4.1832,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314275175695428e-05,
+      "loss": 4.1932,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313439856705122e-05,
+      "loss": 4.1927,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31260126195407e-05,
+      "loss": 4.1862,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311762667203018e-05,
+      "loss": 4.1781,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310924072451966e-05,
+      "loss": 4.1813,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310085477700914e-05,
+      "loss": 4.1782,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309246882949862e-05,
+      "loss": 4.1701,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308409926079183e-05,
+      "loss": 4.1824,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3075729692085045e-05,
+      "loss": 4.1767,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067343744574525e-05,
+      "loss": 4.191,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3058957797064005e-05,
+      "loss": 4.194,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3050571849553485e-05,
+      "loss": 4.1642,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042185902042965e-05,
+      "loss": 4.1866,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3033799954532445e-05,
+      "loss": 4.1763,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025414007021925e-05,
+      "loss": 4.1842,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3017028059511405e-05,
+      "loss": 4.1777,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3008658490804614e-05,
+      "loss": 4.1715,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3000272543294094e-05,
+      "loss": 4.1796,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2991886595783574e-05,
+      "loss": 4.1744,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2983500648273054e-05,
+      "loss": 4.1772,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2975114700762534e-05,
+      "loss": 4.1724,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2966728753252014e-05,
+      "loss": 4.1808,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2958342805741494e-05,
+      "loss": 4.1773,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2949956858230974e-05,
+      "loss": 4.1766,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294158728952418e-05,
+      "loss": 4.1664,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293320134201366e-05,
+      "loss": 4.1732,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292481539450314e-05,
+      "loss": 4.1745,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291642944699262e-05,
+      "loss": 4.1812,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290805987828583e-05,
+      "loss": 4.1679,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289967393077531e-05,
+      "loss": 4.173,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289128798326479e-05,
+      "loss": 4.1686,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288290203575427e-05,
+      "loss": 4.1775,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287451608824375e-05,
+      "loss": 4.1608,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286614651953697e-05,
+      "loss": 4.1665,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285777695083018e-05,
+      "loss": 4.1696,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2849391003319656e-05,
+      "loss": 4.1694,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2841005055809136e-05,
+      "loss": 4.1606,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2832619108298616e-05,
+      "loss": 4.1675,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2824233160788096e-05,
+      "loss": 4.1626,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2815847213277576e-05,
+      "loss": 4.1688,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2807461265767056e-05,
+      "loss": 4.171,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2799075318256536e-05,
+      "loss": 4.1594,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2790705749549745e-05,
+      "loss": 4.1728,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278233618084296e-05,
+      "loss": 4.1844,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277395023333244e-05,
+      "loss": 4.179,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276556428582192e-05,
+      "loss": 4.1577,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275719471711513e-05,
+      "loss": 4.1656,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274880876960461e-05,
+      "loss": 4.169,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274042282209409e-05,
+      "loss": 4.1636,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273203687458357e-05,
+      "loss": 4.1723,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272365092707305e-05,
+      "loss": 4.1575,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271526497956253e-05,
+      "loss": 4.1696,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270687903205201e-05,
+      "loss": 4.1648,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269849308454149e-05,
+      "loss": 4.173,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269010713703097e-05,
+      "loss": 4.1563,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268172118952045e-05,
+      "loss": 4.1612,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267335162081366e-05,
+      "loss": 4.168,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2664982052106875e-05,
+      "loss": 4.1752,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2656596104596355e-05,
+      "loss": 4.1654,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2648210157085835e-05,
+      "loss": 4.1684,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2639824209575315e-05,
+      "loss": 4.1562,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2631438262064795e-05,
+      "loss": 4.1712,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2623052314554274e-05,
+      "loss": 4.1668,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2614666367043754e-05,
+      "loss": 4.1742,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2606280419533234e-05,
+      "loss": 4.1586,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2597894472022714e-05,
+      "loss": 4.1697,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258950852451219e-05,
+      "loss": 4.1622,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258112257700167e-05,
+      "loss": 4.1799,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257273662949115e-05,
+      "loss": 4.1531,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256436706078436e-05,
+      "loss": 4.1703,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255599749207758e-05,
+      "loss": 4.1613,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254761154456706e-05,
+      "loss": 4.1571,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253922559705654e-05,
+      "loss": 4.1669,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253083964954602e-05,
+      "loss": 4.1624,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252245370203549e-05,
+      "loss": 4.1651,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251406775452497e-05,
+      "loss": 4.1579,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250569818581819e-05,
+      "loss": 4.1683,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.1629862785339355,
+      "eval_runtime": 295.6108,
+      "eval_samples_per_second": 1290.856,
+      "eval_steps_per_second": 40.34,
+      "step": 457920
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.8787776957255702e+17,
+  "trial_name": null,
+  "trial_params": null
+}