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": 3.868818759918213,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full2/transformer/4/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": 10.9805,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8324,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.184,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9727,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8087,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.6998,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.5979,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5283,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4424,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.3934,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3425,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3175,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.2617,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.198,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1749,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1372,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1058,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0793,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0555,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0135,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0075,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9731,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9517,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9226,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9224,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.8946,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8678,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8505,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8323,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756873037409844e-05,
+      "loss": 4.8219,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.7996,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7937,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7791,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 4.7634,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 4.7628,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7483,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7333,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 4.7069,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 4.7019,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 4.6847,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.683,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.667,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.662,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963951890867376e-05,
+      "loss": 4.6443,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 4.6513,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6347,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6373,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6264,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.5989,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95892359812181e-05,
+      "loss": 4.5917,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958085003370758e-05,
+      "loss": 4.6049,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957246408619706e-05,
+      "loss": 4.5888,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 4.5752,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5591,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5642,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5467,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5599,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5259,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5379,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5356,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5144,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948863736869932e-05,
+      "loss": 4.5277,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.4994,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.5,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4879,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945510995746097e-05,
+      "loss": 4.5047,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.475,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4804,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4644,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.4626,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.475,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940482703000532e-05,
+      "loss": 4.4673,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396441082494794e-05,
+      "loss": 4.4583,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388055134984274e-05,
+      "loss": 4.4585,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379669187473754e-05,
+      "loss": 4.4644,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 4.4384,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 4.4529,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 4.4258,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 4.4248,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 4.4258,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 4.4268,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 4.4189,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.4268,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 4.4095,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.3901,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.392,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.4038,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927068462744446e-05,
+      "loss": 4.39,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92623314375414e-05,
+      "loss": 4.3897,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925394549003088e-05,
+      "loss": 4.3922,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924555954252036e-05,
+      "loss": 4.3744,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923717359500984e-05,
+      "loss": 4.3807,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9228804026303046e-05,
+      "loss": 4.3743,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922041807879253e-05,
+      "loss": 4.3787,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921203213128201e-05,
+      "loss": 4.3763,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920364618377149e-05,
+      "loss": 4.3719,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919526023626097e-05,
+      "loss": 4.3519,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918687428875045e-05,
+      "loss": 4.3564,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917848834123993e-05,
+      "loss": 4.35,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917010239372941e-05,
+      "loss": 4.354,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.358,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.3379,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3419,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.3442,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912818903498054e-05,
+      "loss": 4.3374,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911980308747002e-05,
+      "loss": 4.3254,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911143351876323e-05,
+      "loss": 4.3287,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.3175,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3164,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.3145,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.3132,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3086,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.3128,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3046,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90443623174828e-05,
+      "loss": 4.3076,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.3148,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027606801265495e-05,
+      "loss": 4.3095,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.3039,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.2956,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.2949,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994063011223415e-05,
+      "loss": 4.2921,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985677063712895e-05,
+      "loss": 4.2807,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977307495006104e-05,
+      "loss": 4.2835,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968921547495584e-05,
+      "loss": 4.2796,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.2852,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2751,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2819,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2743,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.2689,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2643,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2798,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.271,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2631,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888509482999785e-05,
+      "loss": 4.2634,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887672526129106e-05,
+      "loss": 4.2665,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886833931378054e-05,
+      "loss": 4.2628,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885995336627002e-05,
+      "loss": 4.2654,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515674187595e-05,
+      "loss": 4.2533,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2539,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.2529,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.2442,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2401,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8809654060010626e-05,
+      "loss": 4.2406,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8801268112500106e-05,
+      "loss": 4.2432,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8792882164989586e-05,
+      "loss": 4.2335,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2442,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877612664877228e-05,
+      "loss": 4.2358,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876774070126176e-05,
+      "loss": 4.2385,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875935475375124e-05,
+      "loss": 4.2409,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875098518504445e-05,
+      "loss": 4.2223,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.1935296058654785,
+      "eval_runtime": 325.8734,
+      "eval_samples_per_second": 1170.979,
+      "eval_steps_per_second": 36.594,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874259923753393e-05,
+      "loss": 4.2119,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873421329002341e-05,
+      "loss": 4.2082,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872582734251289e-05,
+      "loss": 4.2325,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871744139500237e-05,
+      "loss": 4.2148,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870905544749185e-05,
+      "loss": 4.22,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870066949998133e-05,
+      "loss": 4.2137,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869228355247081e-05,
+      "loss": 4.2064,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683913983764027e-05,
+      "loss": 4.1981,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675544415057236e-05,
+      "loss": 4.207,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667158467546716e-05,
+      "loss": 4.1993,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658772520036196e-05,
+      "loss": 4.2215,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650386572525676e-05,
+      "loss": 4.2072,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642017003818885e-05,
+      "loss": 4.1878,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633631056308365e-05,
+      "loss": 4.1899,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8625245108797845e-05,
+      "loss": 4.1915,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8616859161287324e-05,
+      "loss": 4.1789,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608473213776804e-05,
+      "loss": 4.1928,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860008726626628e-05,
+      "loss": 4.1895,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591701318755764e-05,
+      "loss": 4.1763,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583315371245244e-05,
+      "loss": 4.2004,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574929423734724e-05,
+      "loss": 4.1796,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856655985502793e-05,
+      "loss": 4.1808,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855817390751741e-05,
+      "loss": 4.1713,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854978796000689e-05,
+      "loss": 4.1927,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854140201249637e-05,
+      "loss": 4.169,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853301606498585e-05,
+      "loss": 4.17,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852463011747533e-05,
+      "loss": 4.1652,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851626054876854e-05,
+      "loss": 4.165,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850787460125802e-05,
+      "loss": 4.1628,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84994886537475e-05,
+      "loss": 4.16,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849110270623698e-05,
+      "loss": 4.164,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848271675872646e-05,
+      "loss": 4.1629,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847433081121595e-05,
+      "loss": 4.1598,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846594486370543e-05,
+      "loss": 4.1656,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845757529499864e-05,
+      "loss": 4.1593,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844918934748812e-05,
+      "loss": 4.1607,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84408033999776e-05,
+      "loss": 4.1445,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.1472,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842404788376029e-05,
+      "loss": 4.1381,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841566193624977e-05,
+      "loss": 4.1468,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840727598873925e-05,
+      "loss": 4.1372,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839889004122873e-05,
+      "loss": 4.1469,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8390520472521936e-05,
+      "loss": 4.1311,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382134525011416e-05,
+      "loss": 4.1482,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83737485775009e-05,
+      "loss": 4.1424,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836536262999038e-05,
+      "loss": 4.1412,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835697668247986e-05,
+      "loss": 4.1382,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834859073496934e-05,
+      "loss": 4.1236,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834022116626255e-05,
+      "loss": 4.124,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833183521875203e-05,
+      "loss": 4.1397,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832344927124151e-05,
+      "loss": 4.1294,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831506332373099e-05,
+      "loss": 4.1253,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830667737622047e-05,
+      "loss": 4.1164,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829830780751368e-05,
+      "loss": 4.1259,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828992186000316e-05,
+      "loss": 4.1078,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828153591249264e-05,
+      "loss": 4.1318,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827314996498212e-05,
+      "loss": 4.1029,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82647640174716e-05,
+      "loss": 4.1217,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256394448764816e-05,
+      "loss": 4.12,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248008501254296e-05,
+      "loss": 4.1052,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823962255374377e-05,
+      "loss": 4.1197,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823123660623325e-05,
+      "loss": 4.1068,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822285065872273e-05,
+      "loss": 4.1061,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214481090015945e-05,
+      "loss": 4.0938,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820609514250542e-05,
+      "loss": 4.12,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81977091949949e-05,
+      "loss": 4.0924,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818932324748438e-05,
+      "loss": 4.1048,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180953678777594e-05,
+      "loss": 4.0853,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172567731267074e-05,
+      "loss": 4.096,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164181783756554e-05,
+      "loss": 4.1067,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155795836246034e-05,
+      "loss": 4.1041,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814742626753924e-05,
+      "loss": 4.0997,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813904032002872e-05,
+      "loss": 4.1043,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81306543725182e-05,
+      "loss": 4.1083,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812226842500768e-05,
+      "loss": 4.0941,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811389885630089e-05,
+      "loss": 4.1042,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810551290879037e-05,
+      "loss": 4.0894,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809712696127985e-05,
+      "loss": 4.0883,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808874101376933e-05,
+      "loss": 4.086,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808037144506255e-05,
+      "loss": 4.0933,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807198549755203e-05,
+      "loss": 4.0942,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806359955004151e-05,
+      "loss": 4.0988,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805521360253099e-05,
+      "loss": 4.0903,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8046844033824197e-05,
+      "loss": 4.0694,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8038458086313677e-05,
+      "loss": 4.0723,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8030072138803156e-05,
+      "loss": 4.0871,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8021686191292636e-05,
+      "loss": 4.0792,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8013316622585846e-05,
+      "loss": 4.0771,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8004930675075325e-05,
+      "loss": 4.0859,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996544727564805e-05,
+      "loss": 4.0701,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988158780054285e-05,
+      "loss": 4.0743,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979772832543765e-05,
+      "loss": 4.0742,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797140326383698e-05,
+      "loss": 4.0824,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796301731632646e-05,
+      "loss": 4.0802,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795463136881594e-05,
+      "loss": 4.0854,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794624542130542e-05,
+      "loss": 4.0603,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793787585259863e-05,
+      "loss": 4.0686,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792948990508811e-05,
+      "loss": 4.069,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792110395757759e-05,
+      "loss": 4.0677,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791271801006707e-05,
+      "loss": 4.0806,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790434844136028e-05,
+      "loss": 4.0595,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789596249384976e-05,
+      "loss": 4.0641,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788757654633924e-05,
+      "loss": 4.0712,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787919059882872e-05,
+      "loss": 4.0679,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870821030121935e-05,
+      "loss": 4.0515,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862451461415144e-05,
+      "loss": 4.0635,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854065513904624e-05,
+      "loss": 4.0544,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845679566394104e-05,
+      "loss": 4.0515,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837293618883584e-05,
+      "loss": 4.0512,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828907671373064e-05,
+      "loss": 4.0513,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820521723862544e-05,
+      "loss": 4.0502,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812135776352024e-05,
+      "loss": 4.0604,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803749828841504e-05,
+      "loss": 4.0456,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779538026013471e-05,
+      "loss": 4.0545,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778699431262419e-05,
+      "loss": 4.0663,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777860836511367e-05,
+      "loss": 4.0581,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777022241760315e-05,
+      "loss": 4.0562,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776185284889637e-05,
+      "loss": 4.0516,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775346690138585e-05,
+      "loss": 4.0446,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774508095387533e-05,
+      "loss": 4.0473,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773669500636481e-05,
+      "loss": 4.0414,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772832543765802e-05,
+      "loss": 4.0413,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77199394901475e-05,
+      "loss": 4.042,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771155354263698e-05,
+      "loss": 4.0492,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770316759512646e-05,
+      "loss": 4.0449,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769478164761594e-05,
+      "loss": 4.0435,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768639570010542e-05,
+      "loss": 4.0439,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76780097525949e-05,
+      "loss": 4.033,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7669640183888106e-05,
+      "loss": 4.0355,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7661254236377586e-05,
+      "loss": 4.0512,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765286828886707e-05,
+      "loss": 4.0462,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764448234135655e-05,
+      "loss": 4.0421,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763611277264976e-05,
+      "loss": 4.0359,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762772682513924e-05,
+      "loss": 4.0423,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761934087762872e-05,
+      "loss": 4.0389,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610954930118195e-05,
+      "loss": 4.0481,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760258536141141e-05,
+      "loss": 4.0365,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759419941390089e-05,
+      "loss": 4.0341,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758581346639037e-05,
+      "loss": 4.034,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577427518879844e-05,
+      "loss": 4.0301,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756905795017306e-05,
+      "loss": 4.023,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756067200266254e-05,
+      "loss": 4.0306,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755228605515202e-05,
+      "loss": 4.0336,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75439001076415e-05,
+      "loss": 4.0192,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753551416013098e-05,
+      "loss": 4.038,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527144591424196e-05,
+      "loss": 4.028,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751875864391367e-05,
+      "loss": 4.0275,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751037269640315e-05,
+      "loss": 4.0324,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750198674889263e-05,
+      "loss": 4.0184,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.025674343109131,
+      "eval_runtime": 314.4245,
+      "eval_samples_per_second": 1213.617,
+      "eval_steps_per_second": 37.926,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493617180185845e-05,
+      "loss": 4.0076,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748523123267532e-05,
+      "loss": 4.0067,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74768452851648e-05,
+      "loss": 4.0314,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746845933765428e-05,
+      "loss": 4.0168,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460073390143764e-05,
+      "loss": 4.0273,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451687442633244e-05,
+      "loss": 4.0127,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443301495122724e-05,
+      "loss": 4.0154,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434915547612204e-05,
+      "loss": 4.0023,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426529600101684e-05,
+      "loss": 4.0144,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418143652591164e-05,
+      "loss": 4.009,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409757705080644e-05,
+      "loss": 4.0272,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401371757570124e-05,
+      "loss": 4.0176,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739303494647079e-05,
+      "loss": 3.9975,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738464899896027e-05,
+      "loss": 4.0011,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737626305144975e-05,
+      "loss": 4.0078,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736787710393923e-05,
+      "loss": 3.9917,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735949115642871e-05,
+      "loss": 4.0058,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73511052089182e-05,
+      "loss": 4.0042,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734271926140768e-05,
+      "loss": 3.9967,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733433331389716e-05,
+      "loss": 4.0227,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732594736638664e-05,
+      "loss": 3.9985,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731756141887612e-05,
+      "loss": 4.0031,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73091754713656e-05,
+      "loss": 3.9938,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730080590265881e-05,
+      "loss": 4.0128,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729241995514829e-05,
+      "loss": 3.9887,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728403400763777e-05,
+      "loss": 3.9965,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727564806012725e-05,
+      "loss": 3.9891,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267278491420456e-05,
+      "loss": 3.9941,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258892543909936e-05,
+      "loss": 3.9883,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250506596399416e-05,
+      "loss": 3.9898,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724213702769263e-05,
+      "loss": 3.9939,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723375108018211e-05,
+      "loss": 3.9945,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722536513267159e-05,
+      "loss": 3.99,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721697918516107e-05,
+      "loss": 4.0017,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720859323765055e-05,
+      "loss": 3.9882,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720020729014003e-05,
+      "loss": 3.9949,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191821342629505e-05,
+      "loss": 3.9806,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183435395118985e-05,
+      "loss": 3.985,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71750658264122e-05,
+      "loss": 3.9729,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716667987890168e-05,
+      "loss": 3.9872,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158293931391154e-05,
+      "loss": 3.9739,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149907983880634e-05,
+      "loss": 3.9855,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714153841517385e-05,
+      "loss": 3.9734,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713315246766333e-05,
+      "loss": 3.9851,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712476652015281e-05,
+      "loss": 3.9889,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711638057264229e-05,
+      "loss": 3.9807,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7108011003935505e-05,
+      "loss": 3.9829,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709962505642498e-05,
+      "loss": 3.9672,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709123910891446e-05,
+      "loss": 3.9703,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708285316140394e-05,
+      "loss": 3.985,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074483592697154e-05,
+      "loss": 3.9743,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706609764518663e-05,
+      "loss": 3.9736,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705771169767611e-05,
+      "loss": 3.9624,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704932575016559e-05,
+      "loss": 3.9748,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70409561814588e-05,
+      "loss": 3.952,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703257023394828e-05,
+      "loss": 3.9849,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702418428643776e-05,
+      "loss": 3.9522,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701579833892724e-05,
+      "loss": 3.9724,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700742877022045e-05,
+      "loss": 3.9732,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699904282270993e-05,
+      "loss": 3.9574,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699065687519941e-05,
+      "loss": 3.9755,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698227092768889e-05,
+      "loss": 3.9582,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69739013589821e-05,
+      "loss": 3.9595,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696551541147158e-05,
+      "loss": 3.9485,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695712946396106e-05,
+      "loss": 3.9739,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694874351645054e-05,
+      "loss": 3.9518,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694037394774376e-05,
+      "loss": 3.9615,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693198800023324e-05,
+      "loss": 3.9465,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692360205272272e-05,
+      "loss": 3.95,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69152161052122e-05,
+      "loss": 3.9638,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906846536505406e-05,
+      "loss": 3.9657,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898460588994886e-05,
+      "loss": 3.9547,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6890074641484366e-05,
+      "loss": 3.9697,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881688693973846e-05,
+      "loss": 3.9667,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873319125267055e-05,
+      "loss": 3.9585,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686494955656027e-05,
+      "loss": 3.9597,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685656360904975e-05,
+      "loss": 3.958,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684817766153923e-05,
+      "loss": 3.9521,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683979171402871e-05,
+      "loss": 3.9509,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683140576651819e-05,
+      "loss": 3.9555,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682301981900767e-05,
+      "loss": 3.9599,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681463387149715e-05,
+      "loss": 3.9677,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680624792398663e-05,
+      "loss": 3.9531,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679787835527984e-05,
+      "loss": 3.9325,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678949240776932e-05,
+      "loss": 3.9413,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67811064602588e-05,
+      "loss": 3.9539,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677272051274828e-05,
+      "loss": 3.9474,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676435094404149e-05,
+      "loss": 3.9442,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675596499653097e-05,
+      "loss": 3.9528,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674757904902045e-05,
+      "loss": 3.9448,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739193101509935e-05,
+      "loss": 3.9385,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730823532803144e-05,
+      "loss": 3.9473,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722437585292624e-05,
+      "loss": 3.9538,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6714051637782104e-05,
+      "loss": 3.948,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705665690271584e-05,
+      "loss": 3.9552,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669729612156479e-05,
+      "loss": 3.938,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668891017405427e-05,
+      "loss": 3.9412,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668052422654375e-05,
+      "loss": 3.9413,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667213827903323e-05,
+      "loss": 3.9456,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666376871032644e-05,
+      "loss": 3.9527,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665538276281592e-05,
+      "loss": 3.9359,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66469968153054e-05,
+      "loss": 3.9366,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663861086779488e-05,
+      "loss": 3.9436,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66302412990881e-05,
+      "loss": 3.9462,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662185535157758e-05,
+      "loss": 3.9283,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661348578287079e-05,
+      "loss": 3.9375,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660509983536027e-05,
+      "loss": 3.9363,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659671388784975e-05,
+      "loss": 3.93,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658832794033923e-05,
+      "loss": 3.9281,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579941992828707e-05,
+      "loss": 3.9303,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571556045318187e-05,
+      "loss": 3.9333,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563170097807667e-05,
+      "loss": 3.9376,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554784150297146e-05,
+      "loss": 3.9268,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546414581590356e-05,
+      "loss": 3.9346,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6538028634079836e-05,
+      "loss": 3.9443,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652964268656932e-05,
+      "loss": 3.9446,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65212567390588e-05,
+      "loss": 3.9347,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651288717035201e-05,
+      "loss": 3.9353,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650450122284149e-05,
+      "loss": 3.9294,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649611527533097e-05,
+      "loss": 3.9322,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648772932782045e-05,
+      "loss": 3.9214,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647935975911366e-05,
+      "loss": 3.9234,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647097381160314e-05,
+      "loss": 3.9266,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646258786409262e-05,
+      "loss": 3.9359,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64542019165821e-05,
+      "loss": 3.9263,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644581596907158e-05,
+      "loss": 3.9287,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643743002156106e-05,
+      "loss": 3.9282,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642904407405054e-05,
+      "loss": 3.919,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420674505343756e-05,
+      "loss": 3.9205,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412288557833236e-05,
+      "loss": 3.9386,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640390261032271e-05,
+      "loss": 3.9383,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639551666281219e-05,
+      "loss": 3.9273,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387147094105405e-05,
+      "loss": 3.9233,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637876114659488e-05,
+      "loss": 3.9269,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637037519908436e-05,
+      "loss": 3.9302,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636198925157384e-05,
+      "loss": 3.94,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353619682867054e-05,
+      "loss": 3.9247,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634523373535653e-05,
+      "loss": 3.9206,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633686416664974e-05,
+      "loss": 3.9223,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632847821913923e-05,
+      "loss": 3.9215,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632009227162871e-05,
+      "loss": 3.9168,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631170632411818e-05,
+      "loss": 3.9208,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630332037660766e-05,
+      "loss": 3.9246,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629493442909714e-05,
+      "loss": 3.914,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628654848158662e-05,
+      "loss": 3.9259,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62781625340761e-05,
+      "loss": 3.9252,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626979296536931e-05,
+      "loss": 3.9181,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626140701785879e-05,
+      "loss": 3.923,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625302107034827e-05,
+      "loss": 3.9091,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9514849185943604,
+      "eval_runtime": 305.0841,
+      "eval_samples_per_second": 1250.773,
+      "eval_steps_per_second": 39.088,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624463512283775e-05,
+      "loss": 3.9116,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623626555413097e-05,
+      "loss": 3.9003,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622787960662045e-05,
+      "loss": 3.9269,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621949365910993e-05,
+      "loss": 3.9105,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621110771159941e-05,
+      "loss": 3.9257,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202738142892616e-05,
+      "loss": 3.9058,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194352195382096e-05,
+      "loss": 3.9148,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185966247871576e-05,
+      "loss": 3.8954,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177596679164785e-05,
+      "loss": 3.9147,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6169210731654265e-05,
+      "loss": 3.9029,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160824784143745e-05,
+      "loss": 3.9237,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152438836633225e-05,
+      "loss": 3.9155,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6144069267926434e-05,
+      "loss": 3.8947,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613568332041592e-05,
+      "loss": 3.9,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61272973729054e-05,
+      "loss": 3.9071,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611891142539488e-05,
+      "loss": 3.8929,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611052547788436e-05,
+      "loss": 3.9022,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610213953037384e-05,
+      "loss": 3.9041,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609375358286332e-05,
+      "loss": 3.8968,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60853676353528e-05,
+      "loss": 3.9244,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607698168784228e-05,
+      "loss": 3.9,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606859574033176e-05,
+      "loss": 3.9021,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606020979282124e-05,
+      "loss": 3.9,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605184022411445e-05,
+      "loss": 3.9098,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604345427660393e-05,
+      "loss": 3.8946,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603506832909341e-05,
+      "loss": 3.8997,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602669876038662e-05,
+      "loss": 3.8932,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60183128128761e-05,
+      "loss": 3.8971,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009926865365585e-05,
+      "loss": 3.8923,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001540917855065e-05,
+      "loss": 3.8925,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599315497034454e-05,
+      "loss": 3.8951,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5984785401637754e-05,
+      "loss": 3.9004,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976399454127234e-05,
+      "loss": 3.8997,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5968013506616714e-05,
+      "loss": 3.9016,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595962755910619e-05,
+      "loss": 3.8923,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5951257990399403e-05,
+      "loss": 3.8998,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594287204288888e-05,
+      "loss": 3.8826,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593448609537836e-05,
+      "loss": 3.8959,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5926100147867836e-05,
+      "loss": 3.8786,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591773057916105e-05,
+      "loss": 3.8901,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590934463165054e-05,
+      "loss": 3.8811,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590095868414001e-05,
+      "loss": 3.8956,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589257273662949e-05,
+      "loss": 3.8822,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588418678911897e-05,
+      "loss": 3.8889,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587580084160845e-05,
+      "loss": 3.8989,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586741489409793e-05,
+      "loss": 3.8918,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585902894658741e-05,
+      "loss": 3.8941,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585065937788062e-05,
+      "loss": 3.8753,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58422734303701e-05,
+      "loss": 3.8772,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583388748285958e-05,
+      "loss": 3.8916,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582551791415279e-05,
+      "loss": 3.8872,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581713196664228e-05,
+      "loss": 3.8809,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580874601913176e-05,
+      "loss": 3.8746,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800376450424966e-05,
+      "loss": 3.8852,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791990502914446e-05,
+      "loss": 3.8685,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783604555403926e-05,
+      "loss": 3.8853,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5775218607893406e-05,
+      "loss": 3.8686,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766832660382886e-05,
+      "loss": 3.8797,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758446712872366e-05,
+      "loss": 3.888,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5750060765361846e-05,
+      "loss": 3.8678,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741691196655055e-05,
+      "loss": 3.8831,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733305249144535e-05,
+      "loss": 3.8735,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724919301634015e-05,
+      "loss": 3.8711,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716533354123495e-05,
+      "loss": 3.8674,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5708147406612975e-05,
+      "loss": 3.8811,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569977783790619e-05,
+      "loss": 3.866,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569139189039567e-05,
+      "loss": 3.875,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568300594288515e-05,
+      "loss": 3.8623,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567461999537463e-05,
+      "loss": 3.8624,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566623404786411e-05,
+      "loss": 3.8753,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565784810035359e-05,
+      "loss": 3.8828,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564946215284307e-05,
+      "loss": 3.8678,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564107620533255e-05,
+      "loss": 3.8848,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563270663662576e-05,
+      "loss": 3.88,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562432068911524e-05,
+      "loss": 3.8723,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561595112040845e-05,
+      "loss": 3.877,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560756517289793e-05,
+      "loss": 3.8728,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5599179225387415e-05,
+      "loss": 3.8682,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5590793277876895e-05,
+      "loss": 3.8696,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582423709170104e-05,
+      "loss": 3.8732,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5574037761659584e-05,
+      "loss": 3.8794,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5565651814149064e-05,
+      "loss": 3.8738,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557265866638544e-05,
+      "loss": 3.8725,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548879919128024e-05,
+      "loss": 3.8526,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554051035042123e-05,
+      "loss": 3.8562,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553212440291071e-05,
+      "loss": 3.8694,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552373845540019e-05,
+      "loss": 3.8679,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5515352507889666e-05,
+      "loss": 3.8632,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550696656037915e-05,
+      "loss": 3.8683,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549859699167237e-05,
+      "loss": 3.8638,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549021104416185e-05,
+      "loss": 3.8582,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548182509665132e-05,
+      "loss": 3.8644,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54734391491408e-05,
+      "loss": 3.8703,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546506958043402e-05,
+      "loss": 3.8683,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54566836329235e-05,
+      "loss": 3.8776,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544829768541297e-05,
+      "loss": 3.8607,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543991173790245e-05,
+      "loss": 3.8593,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543152579039193e-05,
+      "loss": 3.8634,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542313984288141e-05,
+      "loss": 3.8655,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541475389537089e-05,
+      "loss": 3.8731,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540636794786037e-05,
+      "loss": 3.8583,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397998379153586e-05,
+      "loss": 3.86,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389628810446796e-05,
+      "loss": 3.8631,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5381242862936275e-05,
+      "loss": 3.8672,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372856915425755e-05,
+      "loss": 3.8513,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364470967915235e-05,
+      "loss": 3.86,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5356085020404715e-05,
+      "loss": 3.8586,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347699072894195e-05,
+      "loss": 3.8504,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339329504187404e-05,
+      "loss": 3.8463,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5330943556676884e-05,
+      "loss": 3.855,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5322557609166364e-05,
+      "loss": 3.8582,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314171661655844e-05,
+      "loss": 3.8568,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530580209294906e-05,
+      "loss": 3.8514,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529741614543854e-05,
+      "loss": 3.8526,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528903019792802e-05,
+      "loss": 3.8694,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52806442504175e-05,
+      "loss": 3.866,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527227468171071e-05,
+      "loss": 3.8619,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526388873420019e-05,
+      "loss": 3.8586,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525550278668967e-05,
+      "loss": 3.8531,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524711683917915e-05,
+      "loss": 3.8544,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523873089166863e-05,
+      "loss": 3.8461,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523036132296184e-05,
+      "loss": 3.8486,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522197537545132e-05,
+      "loss": 3.8507,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52135894279408e-05,
+      "loss": 3.8593,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520520348043028e-05,
+      "loss": 3.8494,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519681753291976e-05,
+      "loss": 3.8573,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518843158540924e-05,
+      "loss": 3.849,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5180045637898725e-05,
+      "loss": 3.8467,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5171676069191934e-05,
+      "loss": 3.8456,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5163290121681414e-05,
+      "loss": 3.8631,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154904174170894e-05,
+      "loss": 3.866,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146518226660373e-05,
+      "loss": 3.8523,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5138132279149853e-05,
+      "loss": 3.8479,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129746331639333e-05,
+      "loss": 3.854,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121360384128807e-05,
+      "loss": 3.8507,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112974436618287e-05,
+      "loss": 3.8655,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51046048679115e-05,
+      "loss": 3.8462,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096218920400976e-05,
+      "loss": 3.8514,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508783297289046e-05,
+      "loss": 3.8467,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507944702537994e-05,
+      "loss": 3.8555,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507106107786942e-05,
+      "loss": 3.8423,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50626751303589e-05,
+      "loss": 3.8469,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505428918284838e-05,
+      "loss": 3.8519,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504591961414159e-05,
+      "loss": 3.8434,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503753366663107e-05,
+      "loss": 3.8516,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502914771912055e-05,
+      "loss": 3.8544,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502076177161003e-05,
+      "loss": 3.8453,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501237582409951e-05,
+      "loss": 3.8487,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500400625539272e-05,
+      "loss": 3.845,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9100821018218994,
+      "eval_runtime": 304.4947,
+      "eval_samples_per_second": 1253.194,
+      "eval_steps_per_second": 39.163,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49956203078822e-05,
+      "loss": 3.8318,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498723436037168e-05,
+      "loss": 3.8295,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497884841286116e-05,
+      "loss": 3.8518,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497046246535065e-05,
+      "loss": 3.8431,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496207651784013e-05,
+      "loss": 3.8567,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495369057032961e-05,
+      "loss": 3.8342,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945304622819087e-05,
+      "loss": 3.8454,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936918675308567e-05,
+      "loss": 3.8256,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928549106601776e-05,
+      "loss": 3.8479,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4920163159091256e-05,
+      "loss": 3.8327,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911777211580736e-05,
+      "loss": 3.8498,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903391264070216e-05,
+      "loss": 3.8478,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4895021695363425e-05,
+      "loss": 3.8221,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886635747852905e-05,
+      "loss": 3.8362,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878249800342385e-05,
+      "loss": 3.8358,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869863852831864e-05,
+      "loss": 3.8237,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861477905321344e-05,
+      "loss": 3.8339,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4853091957810824e-05,
+      "loss": 3.8337,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844706010300304e-05,
+      "loss": 3.8321,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4836320062789784e-05,
+      "loss": 3.8522,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4827934115279264e-05,
+      "loss": 3.8304,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481956454657247e-05,
+      "loss": 3.8342,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481117859906195e-05,
+      "loss": 3.8341,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480279265155143e-05,
+      "loss": 3.8406,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479440670404091e-05,
+      "loss": 3.8268,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478603713533412e-05,
+      "loss": 3.8397,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47776511878236e-05,
+      "loss": 3.8226,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476926524031308e-05,
+      "loss": 3.8294,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476087929280256e-05,
+      "loss": 3.8231,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475249334529205e-05,
+      "loss": 3.8281,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474410739778153e-05,
+      "loss": 3.8266,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473573782907474e-05,
+      "loss": 3.8366,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472735188156422e-05,
+      "loss": 3.833,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47189659340537e-05,
+      "loss": 3.8373,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471057998654318e-05,
+      "loss": 3.825,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470221041783639e-05,
+      "loss": 3.8353,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469382447032587e-05,
+      "loss": 3.8221,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468543852281535e-05,
+      "loss": 3.8238,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467705257530483e-05,
+      "loss": 3.8154,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668683006598036e-05,
+      "loss": 3.8247,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660297059087516e-05,
+      "loss": 3.818,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651911111577e-05,
+      "loss": 3.8309,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464352516406648e-05,
+      "loss": 3.8187,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463513921655596e-05,
+      "loss": 3.8219,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462675326904544e-05,
+      "loss": 3.8371,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461838370033865e-05,
+      "loss": 3.8269,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460999775282813e-05,
+      "loss": 3.826,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460161180531761e-05,
+      "loss": 3.8161,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459322585780709e-05,
+      "loss": 3.8076,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45848562891003e-05,
+      "loss": 3.8292,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457647034158978e-05,
+      "loss": 3.8236,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456808439407926e-05,
+      "loss": 3.8166,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455969844656874e-05,
+      "loss": 3.8156,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551328877861956e-05,
+      "loss": 3.8248,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542942930351436e-05,
+      "loss": 3.8026,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534556982840916e-05,
+      "loss": 3.8248,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4526171035330396e-05,
+      "loss": 3.8039,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517785087819876e-05,
+      "loss": 3.8198,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509415519113085e-05,
+      "loss": 3.8226,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4501029571602565e-05,
+      "loss": 3.8095,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492643624092045e-05,
+      "loss": 3.8159,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4484257676581525e-05,
+      "loss": 3.8134,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475871729071e-05,
+      "loss": 3.8065,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467502160364214e-05,
+      "loss": 3.8109,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4459116212853694e-05,
+      "loss": 3.8153,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4450730265343174e-05,
+      "loss": 3.8086,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442344317832654e-05,
+      "loss": 3.8117,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4433958370322134e-05,
+      "loss": 3.8028,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442558880161535e-05,
+      "loss": 3.7995,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441720285410482e-05,
+      "loss": 3.814,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44088169065943e-05,
+      "loss": 3.8232,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440043095908378e-05,
+      "loss": 3.8063,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439204501157326e-05,
+      "loss": 3.8257,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438367544286647e-05,
+      "loss": 3.8208,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437528949535595e-05,
+      "loss": 3.8101,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436690354784543e-05,
+      "loss": 3.814,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435851760033491e-05,
+      "loss": 3.8174,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435014803162813e-05,
+      "loss": 3.8082,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434176208411761e-05,
+      "loss": 3.8079,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433337613660709e-05,
+      "loss": 3.8156,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432499018909657e-05,
+      "loss": 3.8139,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431660424158605e-05,
+      "loss": 3.814,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308234672879257e-05,
+      "loss": 3.8106,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299848725368737e-05,
+      "loss": 3.7936,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291462777858217e-05,
+      "loss": 3.7957,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4283076830347696e-05,
+      "loss": 3.8087,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274707261640906e-05,
+      "loss": 3.8098,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266321314130386e-05,
+      "loss": 3.805,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257935366619865e-05,
+      "loss": 3.8059,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249549419109345e-05,
+      "loss": 3.8022,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424116347159883e-05,
+      "loss": 3.8015,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423279390289204e-05,
+      "loss": 3.8069,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422440795538152e-05,
+      "loss": 3.8055,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216022007871e-05,
+      "loss": 3.8131,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420763606036048e-05,
+      "loss": 3.8137,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419926649165369e-05,
+      "loss": 3.8086,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419088054414317e-05,
+      "loss": 3.7978,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418249459663265e-05,
+      "loss": 3.8097,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417410864912213e-05,
+      "loss": 3.805,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416572270161161e-05,
+      "loss": 3.8142,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415735313290482e-05,
+      "loss": 3.8022,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41489671853943e-05,
+      "loss": 3.7953,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414058123788378e-05,
+      "loss": 3.8116,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4132195290373266e-05,
+      "loss": 3.807,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123825721666475e-05,
+      "loss": 3.797,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115439774155955e-05,
+      "loss": 3.804,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4107053826645435e-05,
+      "loss": 3.8024,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4098667879134915e-05,
+      "loss": 3.7865,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4090298310428124e-05,
+      "loss": 3.7909,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4081912362917604e-05,
+      "loss": 3.8025,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4073526415407084e-05,
+      "loss": 3.799,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4065140467896564e-05,
+      "loss": 3.799,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405677089918977e-05,
+      "loss": 3.7955,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404838495167925e-05,
+      "loss": 3.7943,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403999900416873e-05,
+      "loss": 3.8155,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403161305665822e-05,
+      "loss": 3.8105,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402324348795143e-05,
+      "loss": 3.8047,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401485754044091e-05,
+      "loss": 3.8019,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400647159293039e-05,
+      "loss": 3.7947,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399808564541987e-05,
+      "loss": 3.7964,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398971607671308e-05,
+      "loss": 3.7912,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398133012920256e-05,
+      "loss": 3.7957,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397294418169204e-05,
+      "loss": 3.7941,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396455823418152e-05,
+      "loss": 3.806,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3956188665474726e-05,
+      "loss": 3.7941,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947802717964206e-05,
+      "loss": 3.7998,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3939416770453686e-05,
+      "loss": 3.7953,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393103082294317e-05,
+      "loss": 3.7921,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392266125423638e-05,
+      "loss": 3.7926,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391427530672586e-05,
+      "loss": 3.8039,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390588935921534e-05,
+      "loss": 3.8108,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389750341170482e-05,
+      "loss": 3.8013,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388913384299803e-05,
+      "loss": 3.7908,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388074789548751e-05,
+      "loss": 3.8022,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387236194797699e-05,
+      "loss": 3.7943,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386397600046647e-05,
+      "loss": 3.8074,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385559005295595e-05,
+      "loss": 3.7956,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384722048424916e-05,
+      "loss": 3.7947,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383883453673864e-05,
+      "loss": 3.7917,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383044858922813e-05,
+      "loss": 3.8031,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382206264171761e-05,
+      "loss": 3.7862,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813693073010816e-05,
+      "loss": 3.7929,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805307125500296e-05,
+      "loss": 3.7964,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796921177989776e-05,
+      "loss": 3.7902,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788535230479256e-05,
+      "loss": 3.7975,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3780165661772465e-05,
+      "loss": 3.7987,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771779714261945e-05,
+      "loss": 3.7939,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763393766751425e-05,
+      "loss": 3.7905,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37550078192409e-05,
+      "loss": 3.7943,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.885098934173584,
+      "eval_runtime": 303.3818,
+      "eval_samples_per_second": 1257.791,
+      "eval_steps_per_second": 39.307,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3746638250534114e-05,
+      "loss": 3.7821,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3738252303023594e-05,
+      "loss": 3.7754,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372986635551308e-05,
+      "loss": 3.7986,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3721480408002554e-05,
+      "loss": 3.793,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3713094460492034e-05,
+      "loss": 3.8,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3704708512981514e-05,
+      "loss": 3.7852,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3696322565470993e-05,
+      "loss": 3.789,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3687936617960473e-05,
+      "loss": 3.7773,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367956704925368e-05,
+      "loss": 3.7891,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36711974805469e-05,
+      "loss": 3.7873,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366281153303637e-05,
+      "loss": 3.7956,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365442558552585e-05,
+      "loss": 3.7947,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364603963801533e-05,
+      "loss": 3.7699,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363765369050482e-05,
+      "loss": 3.7841,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36292677429943e-05,
+      "loss": 3.7862,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362088179548378e-05,
+      "loss": 3.7736,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361249584797326e-05,
+      "loss": 3.7808,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360412627926647e-05,
+      "loss": 3.7821,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359574033175595e-05,
+      "loss": 3.7811,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358735438424543e-05,
+      "loss": 3.7967,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357896843673491e-05,
+      "loss": 3.7812,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3570598868028116e-05,
+      "loss": 3.787,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3562212920517596e-05,
+      "loss": 3.7819,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3553826973007076e-05,
+      "loss": 3.788,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3545441025496556e-05,
+      "loss": 3.777,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353707145678977e-05,
+      "loss": 3.7853,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352868550927925e-05,
+      "loss": 3.7771,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352029956176873e-05,
+      "loss": 3.7759,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351191361425821e-05,
+      "loss": 3.7743,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350352766674769e-05,
+      "loss": 3.7775,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34951580980409e-05,
+      "loss": 3.7766,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348677215053038e-05,
+      "loss": 3.7836,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347838620301986e-05,
+      "loss": 3.7861,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.347000025550934e-05,
+      "loss": 3.785,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346163068680255e-05,
+      "loss": 3.7745,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345324473929203e-05,
+      "loss": 3.7822,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344485879178151e-05,
+      "loss": 3.7746,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343647284427099e-05,
+      "loss": 3.7755,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3428103275564206e-05,
+      "loss": 3.7643,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3419717328053685e-05,
+      "loss": 3.7715,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3411331380543165e-05,
+      "loss": 3.7726,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402945433032645e-05,
+      "loss": 3.7801,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3394575864325854e-05,
+      "loss": 3.77,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3386189916815334e-05,
+      "loss": 3.7711,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377803969304814e-05,
+      "loss": 3.7882,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3369418021794294e-05,
+      "loss": 3.7767,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3361048453087503e-05,
+      "loss": 3.7794,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3352662505576983e-05,
+      "loss": 3.7684,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334427655806646e-05,
+      "loss": 3.7548,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333589061055594e-05,
+      "loss": 3.7794,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332752104184916e-05,
+      "loss": 3.7768,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331913509433864e-05,
+      "loss": 3.767,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331074914682812e-05,
+      "loss": 3.7674,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33023631993176e-05,
+      "loss": 3.7783,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329399363061081e-05,
+      "loss": 3.7496,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328560768310029e-05,
+      "loss": 3.776,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327722173558977e-05,
+      "loss": 3.7595,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326883578807925e-05,
+      "loss": 3.7672,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326044984056873e-05,
+      "loss": 3.7759,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325208027186194e-05,
+      "loss": 3.7655,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324369432435142e-05,
+      "loss": 3.7644,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32353083768409e-05,
+      "loss": 3.7692,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322692242933038e-05,
+      "loss": 3.7589,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321855286062359e-05,
+      "loss": 3.7586,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.321016691311307e-05,
+      "loss": 3.7684,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320178096560255e-05,
+      "loss": 3.7657,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319339501809203e-05,
+      "loss": 3.7601,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318502544938524e-05,
+      "loss": 3.7577,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317663950187472e-05,
+      "loss": 3.7527,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31682535543642e-05,
+      "loss": 3.7627,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315986760685368e-05,
+      "loss": 3.7718,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315149803814689e-05,
+      "loss": 3.7607,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314311209063637e-05,
+      "loss": 3.7786,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313472614312585e-05,
+      "loss": 3.7753,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312634019561533e-05,
+      "loss": 3.7655,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3117970626908546e-05,
+      "loss": 3.7609,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109584679398026e-05,
+      "loss": 3.7733,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3101198731887506e-05,
+      "loss": 3.7628,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309281278437698e-05,
+      "loss": 3.7598,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3084443215670195e-05,
+      "loss": 3.7688,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3076057268159675e-05,
+      "loss": 3.7651,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067671320649155e-05,
+      "loss": 3.7695,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305928537313863e-05,
+      "loss": 3.7647,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3050915804431844e-05,
+      "loss": 3.75,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042529856921324e-05,
+      "loss": 3.7472,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3034143909410804e-05,
+      "loss": 3.7618,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025757961900284e-05,
+      "loss": 3.7648,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30173883931935e-05,
+      "loss": 3.76,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300900244568298e-05,
+      "loss": 3.7582,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300061649817245e-05,
+      "loss": 3.7603,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299223055066193e-05,
+      "loss": 3.7515,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298386098195515e-05,
+      "loss": 3.7648,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297547503444463e-05,
+      "loss": 3.7562,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29670890869341e-05,
+      "loss": 3.7676,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.295870313942358e-05,
+      "loss": 3.7716,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29503335707168e-05,
+      "loss": 3.7621,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294194762320628e-05,
+      "loss": 3.7541,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293356167569576e-05,
+      "loss": 3.7604,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292517572818524e-05,
+      "loss": 3.7598,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2916806159478454e-05,
+      "loss": 3.7712,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290842021196793e-05,
+      "loss": 3.7559,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290003426445741e-05,
+      "loss": 3.7532,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289164831694689e-05,
+      "loss": 3.766,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28832787482401e-05,
+      "loss": 3.7597,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2874892800729576e-05,
+      "loss": 3.7542,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2866506853219056e-05,
+      "loss": 3.7588,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2858120905708536e-05,
+      "loss": 3.7539,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284975133700175e-05,
+      "loss": 3.7485,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2841365389491225e-05,
+      "loss": 3.7407,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283297944198071e-05,
+      "loss": 3.7578,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282459349447019e-05,
+      "loss": 3.7557,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28162239257634e-05,
+      "loss": 3.7535,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280783797825288e-05,
+      "loss": 3.7512,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279945203074236e-05,
+      "loss": 3.7468,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279106608323184e-05,
+      "loss": 3.7703,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278269651452505e-05,
+      "loss": 3.7653,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277431056701453e-05,
+      "loss": 3.7619,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276592461950401e-05,
+      "loss": 3.7593,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275753867199349e-05,
+      "loss": 3.7499,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27491691032867e-05,
+      "loss": 3.7526,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274078315577618e-05,
+      "loss": 3.7473,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2732397208265665e-05,
+      "loss": 3.7498,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2724011260755145e-05,
+      "loss": 3.7543,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715641692048354e-05,
+      "loss": 3.7571,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2707255744537834e-05,
+      "loss": 3.7541,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2698869797027314e-05,
+      "loss": 3.7518,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2690483849516794e-05,
+      "loss": 3.7531,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268211428081e-05,
+      "loss": 3.7492,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267372833329948e-05,
+      "loss": 3.7503,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266534238578896e-05,
+      "loss": 3.7541,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265695643827844e-05,
+      "loss": 3.7664,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264858686957165e-05,
+      "loss": 3.762,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264020092206113e-05,
+      "loss": 3.7463,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263181497455062e-05,
+      "loss": 3.759,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26234290270401e-05,
+      "loss": 3.7547,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261505945833331e-05,
+      "loss": 3.7598,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260667351082279e-05,
+      "loss": 3.7538,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.259828756331227e-05,
+      "loss": 3.7556,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258990161580175e-05,
+      "loss": 3.7487,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258153204709496e-05,
+      "loss": 3.7558,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257314609958444e-05,
+      "loss": 3.744,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.256476015207392e-05,
+      "loss": 3.7505,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25563742045634e-05,
+      "loss": 3.75,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2548004635856606e-05,
+      "loss": 3.7508,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2539618688346086e-05,
+      "loss": 3.7559,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253123274083557e-05,
+      "loss": 3.75,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252284679332505e-05,
+      "loss": 3.7594,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251447722461826e-05,
+      "loss": 3.7431,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250609127710774e-05,
+      "loss": 3.7537,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.868818759918213,
+      "eval_runtime": 303.073,
+      "eval_samples_per_second": 1259.073,
+      "eval_steps_per_second": 39.347,
+      "step": 457920
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 3.187007566568018e+17,
+  "trial_name": null,
+  "trial_params": null
+}