diff --git "a/checkpoint-915840/trainer_state.json" "b/checkpoint-915840/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-915840/trainer_state.json"
@@ -0,0 +1,10849 @@
+{
+  "best_metric": 3.849526882171631,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/passive/transformer/0/checkpoints/checkpoint-915840",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 915840,
+  "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": 10.9439,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.817,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1835,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9785,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8168,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6117,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5357,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4533,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4043,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3518,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3117,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2639,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2131,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.175,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1275,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.112,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0788,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0485,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0198,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983231380739706e-05,
+      "loss": 5.0109,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9774,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9542,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9241,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9211,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.8846,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 4.8729,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 4.8443,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8323,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8198,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 4.8012,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 4.7916,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7801,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7648,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.762,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7413,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.738,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.6962,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7072,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6719,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6816,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6666,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6632,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6384,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 4.651,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6361,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6303,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6246,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9597605549924894e-05,
+      "loss": 4.5932,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95892359812181e-05,
+      "loss": 4.5964,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958085003370758e-05,
+      "loss": 4.5969,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957246408619706e-05,
+      "loss": 4.5867,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 4.578,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5598,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5563,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5382,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5622,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.517,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5407,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5322,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5108,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948863736869932e-05,
+      "loss": 4.5105,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.501,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4952,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4867,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945510995746097e-05,
+      "loss": 4.4952,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.4751,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4705,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.464,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9421582546222625e-05,
+      "loss": 4.4642,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413196598712105e-05,
+      "loss": 4.4746,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4659,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4451,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.459,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4597,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4304,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.4383,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.4285,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4233,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.4145,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.4187,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.4219,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.4241,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 4.401,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.379,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 4.3946,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 4.3952,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 4.3936,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92622823011302e-05,
+      "loss": 4.3807,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925391273242342e-05,
+      "loss": 4.3844,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92455267849129e-05,
+      "loss": 4.3799,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923714083740238e-05,
+      "loss": 4.3758,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 4.3658,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922038532118507e-05,
+      "loss": 4.3683,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921199937367455e-05,
+      "loss": 4.3569,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920361342616403e-05,
+      "loss": 4.3698,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919522747865351e-05,
+      "loss": 4.358,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918684153114299e-05,
+      "loss": 4.3388,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917845558363247e-05,
+      "loss": 4.3511,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917006963612195e-05,
+      "loss": 4.3506,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916168368861143e-05,
+      "loss": 4.3539,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.3387,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.3396,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136542224883596e-05,
+      "loss": 4.3245,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128156277373076e-05,
+      "loss": 4.3338,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.3159,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.323,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.3208,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.3193,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.3093,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907787334991742e-05,
+      "loss": 4.3132,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90694874024069e-05,
+      "loss": 4.3052,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.312,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905271550738586e-05,
+      "loss": 4.3035,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.2982,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.3138,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.302,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.3,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.2877,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.2917,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.285,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.2849,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.2925,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968905168691855e-05,
+      "loss": 4.2774,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.2829,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.2636,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.2674,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2745,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.265,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918605862432463e-05,
+      "loss": 4.2563,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.2813,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2625,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2565,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.2559,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.2587,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886832293497681e-05,
+      "loss": 4.259,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.2563,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.2501,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884316509244525e-05,
+      "loss": 4.2577,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.2549,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.2501,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2372,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8809654060010626e-05,
+      "loss": 4.2463,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8801268112500106e-05,
+      "loss": 4.238,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8792882164989586e-05,
+      "loss": 4.2387,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2348,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877612664877228e-05,
+      "loss": 4.2325,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876774070126176e-05,
+      "loss": 4.2384,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875935475375124e-05,
+      "loss": 4.2257,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875096880624072e-05,
+      "loss": 4.2225,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.200098991394043,
+      "eval_runtime": 313.6584,
+      "eval_samples_per_second": 1216.581,
+      "eval_steps_per_second": 38.019,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874259923753393e-05,
+      "loss": 4.209,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873421329002341e-05,
+      "loss": 4.2087,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872582734251289e-05,
+      "loss": 4.235,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871744139500237e-05,
+      "loss": 4.2207,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870907182629558e-05,
+      "loss": 4.2131,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870068587878506e-05,
+      "loss": 4.2102,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869229993127454e-05,
+      "loss": 4.2078,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683913983764027e-05,
+      "loss": 4.1962,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675544415057236e-05,
+      "loss": 4.2081,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667158467546716e-05,
+      "loss": 4.2039,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658772520036196e-05,
+      "loss": 4.2068,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8650386572525676e-05,
+      "loss": 4.2084,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642017003818885e-05,
+      "loss": 4.1921,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633631056308365e-05,
+      "loss": 4.187,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8625245108797845e-05,
+      "loss": 4.1807,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8616859161287324e-05,
+      "loss": 4.1793,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608473213776804e-05,
+      "loss": 4.1871,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860008726626628e-05,
+      "loss": 4.1775,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591701318755764e-05,
+      "loss": 4.1835,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8583315371245244e-05,
+      "loss": 4.2002,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857494580253846e-05,
+      "loss": 4.1802,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856655985502793e-05,
+      "loss": 4.1835,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855817390751741e-05,
+      "loss": 4.1725,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854980433881063e-05,
+      "loss": 4.1866,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854141839130011e-05,
+      "loss": 4.1696,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853303244378958e-05,
+      "loss": 4.1687,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852464649627906e-05,
+      "loss": 4.1618,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851627692757228e-05,
+      "loss": 4.1681,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850789098006175e-05,
+      "loss": 4.1589,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849950503255123e-05,
+      "loss": 4.1568,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849111908504072e-05,
+      "loss": 4.162,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482749516333934e-05,
+      "loss": 4.1679,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847436356882341e-05,
+      "loss": 4.1578,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846597762131289e-05,
+      "loss": 4.1609,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845759167380237e-05,
+      "loss": 4.1533,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844922210509558e-05,
+      "loss": 4.1629,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440836157585056e-05,
+      "loss": 4.1364,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432450210074536e-05,
+      "loss": 4.1528,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8424064262564016e-05,
+      "loss": 4.1299,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415694693857225e-05,
+      "loss": 4.151,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8407308746346705e-05,
+      "loss": 4.1349,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398922798836185e-05,
+      "loss": 4.1493,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839053685132567e-05,
+      "loss": 4.1303,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838216728261888e-05,
+      "loss": 4.1437,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837378133510836e-05,
+      "loss": 4.1424,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836539538759784e-05,
+      "loss": 4.1411,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835700944008732e-05,
+      "loss": 4.1438,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834863987138053e-05,
+      "loss": 4.1164,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834025392387001e-05,
+      "loss": 4.1234,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833186797635949e-05,
+      "loss": 4.1364,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832348202884897e-05,
+      "loss": 4.1331,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831511246014218e-05,
+      "loss": 4.126,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830672651263166e-05,
+      "loss": 4.1207,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298356943924875e-05,
+      "loss": 4.1163,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289970996414354e-05,
+      "loss": 4.1088,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281585048903834e-05,
+      "loss": 4.1236,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273199101393314e-05,
+      "loss": 4.1053,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826482953268653e-05,
+      "loss": 4.1157,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256443585176003e-05,
+      "loss": 4.124,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248057637665483e-05,
+      "loss": 4.1044,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823967169015496e-05,
+      "loss": 4.1056,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823128574264444e-05,
+      "loss": 4.1045,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822289979513392e-05,
+      "loss": 4.0978,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82145138476234e-05,
+      "loss": 4.0969,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820612790011288e-05,
+      "loss": 4.1136,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819777471020983e-05,
+      "loss": 4.0989,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818938876269931e-05,
+      "loss": 4.0949,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818100281518879e-05,
+      "loss": 4.0921,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817261686767827e-05,
+      "loss": 4.094,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816423092016775e-05,
+      "loss": 4.1086,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815586135146096e-05,
+      "loss": 4.1011,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814747540395044e-05,
+      "loss": 4.0936,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813908945643992e-05,
+      "loss": 4.1025,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81307035089294e-05,
+      "loss": 4.1062,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812231756141888e-05,
+      "loss": 4.0848,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811393161390836e-05,
+      "loss": 4.0934,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810554566639784e-05,
+      "loss": 4.0911,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8097176097691046e-05,
+      "loss": 4.0885,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808879015018053e-05,
+      "loss": 4.0815,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808040420267001e-05,
+      "loss": 4.0848,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807201825515949e-05,
+      "loss": 4.0954,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806363230764897e-05,
+      "loss": 4.0916,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805526273894218e-05,
+      "loss": 4.0839,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804687679143166e-05,
+      "loss": 4.0595,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803849084392114e-05,
+      "loss": 4.0756,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803010489641062e-05,
+      "loss": 4.0818,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80217189489001e-05,
+      "loss": 4.0795,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801333300138958e-05,
+      "loss": 4.0757,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800494705387906e-05,
+      "loss": 4.0721,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799657748517227e-05,
+      "loss": 4.0787,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988207916465486e-05,
+      "loss": 4.0693,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979821968954966e-05,
+      "loss": 4.0639,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971436021444446e-05,
+      "loss": 4.0782,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7963050073933926e-05,
+      "loss": 4.0613,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954680505227135e-05,
+      "loss": 4.0817,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946294557716615e-05,
+      "loss": 4.0709,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937908610206095e-05,
+      "loss": 4.0516,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929522662695575e-05,
+      "loss": 4.0687,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7921136715185055e-05,
+      "loss": 4.065,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912750767674535e-05,
+      "loss": 4.0751,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7904381198967744e-05,
+      "loss": 4.0644,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895995251457224e-05,
+      "loss": 4.06,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887609303946704e-05,
+      "loss": 4.0498,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879223356436184e-05,
+      "loss": 4.066,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78708537877294e-05,
+      "loss": 4.0456,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786246784021888e-05,
+      "loss": 4.0535,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785408189270836e-05,
+      "loss": 4.0571,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784569594519783e-05,
+      "loss": 4.055,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783730999768731e-05,
+      "loss": 4.0443,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782892405017679e-05,
+      "loss": 4.0548,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782053810266627e-05,
+      "loss": 4.0473,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781216853395948e-05,
+      "loss": 4.0518,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780378258644896e-05,
+      "loss": 4.0501,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779539663893844e-05,
+      "loss": 4.0445,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778701069142792e-05,
+      "loss": 4.069,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777864112272114e-05,
+      "loss": 4.0548,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777025517521062e-05,
+      "loss": 4.0507,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77618692277001e-05,
+      "loss": 4.0379,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775348328018958e-05,
+      "loss": 4.0468,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774509733267906e-05,
+      "loss": 4.0441,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773671138516854e-05,
+      "loss": 4.037,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772832543765802e-05,
+      "loss": 4.0542,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77199394901475e-05,
+      "loss": 4.0412,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7711569921440707e-05,
+      "loss": 4.041,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703183973930186e-05,
+      "loss": 4.0333,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694814405223396e-05,
+      "loss": 4.0357,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686428457712876e-05,
+      "loss": 4.0376,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7678042510202355e-05,
+      "loss": 4.0367,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766965656269184e-05,
+      "loss": 4.024,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766128699398505e-05,
+      "loss": 4.0535,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765290104647453e-05,
+      "loss": 4.0416,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764451509896401e-05,
+      "loss": 4.0328,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763612915145349e-05,
+      "loss": 4.0308,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762774320394297e-05,
+      "loss": 4.0287,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761937363523618e-05,
+      "loss": 4.04,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761098768772566e-05,
+      "loss": 4.0364,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760260174021514e-05,
+      "loss": 4.0293,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759421579270462e-05,
+      "loss": 4.0408,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758584622399783e-05,
+      "loss": 4.0339,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757746027648731e-05,
+      "loss": 4.0372,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569074328976796e-05,
+      "loss": 4.0219,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560688381466276e-05,
+      "loss": 4.0364,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552318812759485e-05,
+      "loss": 4.0273,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543932865248965e-05,
+      "loss": 4.0249,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535546917738445e-05,
+      "loss": 4.0265,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527160970227925e-05,
+      "loss": 4.0275,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518775022717405e-05,
+      "loss": 4.028,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510405454010614e-05,
+      "loss": 4.0184,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7502019506500094e-05,
+      "loss": 4.0189,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.031566143035889,
+      "eval_runtime": 313.0538,
+      "eval_samples_per_second": 1218.931,
+      "eval_steps_per_second": 38.092,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493633558989574e-05,
+      "loss": 4.0065,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485247611479054e-05,
+      "loss": 4.0066,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476861663968534e-05,
+      "loss": 4.0332,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468475716458014e-05,
+      "loss": 4.0258,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460089768947494e-05,
+      "loss": 4.0166,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451703821436974e-05,
+      "loss": 4.0103,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443317873926453e-05,
+      "loss": 4.0131,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434931926415933e-05,
+      "loss": 3.9996,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426545978905413e-05,
+      "loss": 4.0192,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741817641019862e-05,
+      "loss": 4.014,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74097904626881e-05,
+      "loss": 4.013,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740140451517758e-05,
+      "loss": 4.0147,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739301856766706e-05,
+      "loss": 4.0025,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738463262015654e-05,
+      "loss": 4.0011,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737626305144975e-05,
+      "loss": 3.9971,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736787710393923e-05,
+      "loss": 3.996,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735949115642871e-05,
+      "loss": 4.0005,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73511052089182e-05,
+      "loss": 3.9938,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734273564021141e-05,
+      "loss": 4.0005,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733434969270089e-05,
+      "loss": 4.0233,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732596374519037e-05,
+      "loss": 3.9976,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731757779767985e-05,
+      "loss": 4.0022,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7309208228973056e-05,
+      "loss": 3.9992,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300822281462536e-05,
+      "loss": 4.0059,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292436333952016e-05,
+      "loss": 3.9893,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7284050386441496e-05,
+      "loss": 3.9981,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275664438930976e-05,
+      "loss": 3.989,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267294870224185e-05,
+      "loss": 3.9915,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258908922713665e-05,
+      "loss": 3.9847,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725052297520315e-05,
+      "loss": 3.9841,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724213702769263e-05,
+      "loss": 3.9907,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723376745898584e-05,
+      "loss": 3.9993,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722538151147532e-05,
+      "loss": 3.9899,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72169955639648e-05,
+      "loss": 3.9988,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720860961645428e-05,
+      "loss": 3.9811,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720022366894376e-05,
+      "loss": 4.0009,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719185410023697e-05,
+      "loss": 3.9708,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718346815272645e-05,
+      "loss": 3.9898,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717508220521593e-05,
+      "loss": 3.9669,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716669625770541e-05,
+      "loss": 3.9865,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715832668899862e-05,
+      "loss": 3.974,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149940741488105e-05,
+      "loss": 3.9917,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141554793977585e-05,
+      "loss": 3.9737,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133168846467065e-05,
+      "loss": 3.9794,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124782898956545e-05,
+      "loss": 3.9835,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116413330249754e-05,
+      "loss": 3.9859,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7108027382739234e-05,
+      "loss": 3.9831,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099641435228714e-05,
+      "loss": 3.9638,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091255487718194e-05,
+      "loss": 3.9636,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70828859190114e-05,
+      "loss": 3.9837,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707449997150088e-05,
+      "loss": 3.9787,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706611402399036e-05,
+      "loss": 3.9722,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705772807647984e-05,
+      "loss": 3.9691,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704935850777306e-05,
+      "loss": 3.9666,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704097256026254e-05,
+      "loss": 3.9562,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703258661275202e-05,
+      "loss": 3.9683,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70242006652415e-05,
+      "loss": 3.9558,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701583109653471e-05,
+      "loss": 3.9685,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700744514902419e-05,
+      "loss": 3.9758,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699905920151367e-05,
+      "loss": 3.956,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699067325400315e-05,
+      "loss": 3.957,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698228730649263e-05,
+      "loss": 3.962,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69739013589821e-05,
+      "loss": 3.9508,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696553179027532e-05,
+      "loss": 3.9467,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69571458427648e-05,
+      "loss": 3.9695,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694875989525428e-05,
+      "loss": 3.9562,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694037394774376e-05,
+      "loss": 3.9516,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693200437903697e-05,
+      "loss": 3.9481,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692361843152645e-05,
+      "loss": 3.9542,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915232484015926e-05,
+      "loss": 3.9647,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906846536505406e-05,
+      "loss": 3.9606,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898460588994886e-05,
+      "loss": 3.9505,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6890091020288095e-05,
+      "loss": 3.9654,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881705072777575e-05,
+      "loss": 3.9615,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873319125267055e-05,
+      "loss": 3.9491,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864933177756535e-05,
+      "loss": 3.9538,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685656360904975e-05,
+      "loss": 3.9572,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684817766153923e-05,
+      "loss": 3.9507,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683979171402871e-05,
+      "loss": 3.9477,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683140576651819e-05,
+      "loss": 3.9483,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68230361978114e-05,
+      "loss": 3.9577,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681465025030088e-05,
+      "loss": 3.9579,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680626430279036e-05,
+      "loss": 3.9442,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679787835527984e-05,
+      "loss": 3.9324,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678950878657305e-05,
+      "loss": 3.9385,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678112283906253e-05,
+      "loss": 3.9459,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677273689155201e-05,
+      "loss": 3.9496,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676435094404149e-05,
+      "loss": 3.9422,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755981375334704e-05,
+      "loss": 3.9386,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747595427824184e-05,
+      "loss": 3.9524,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739209480313664e-05,
+      "loss": 3.9352,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730823532803144e-05,
+      "loss": 3.9365,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672245396409635e-05,
+      "loss": 3.9473,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671406801658583e-05,
+      "loss": 3.9357,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670568206907531e-05,
+      "loss": 3.9547,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669729612156479e-05,
+      "loss": 3.9415,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688926552858e-05,
+      "loss": 3.9251,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668054060534748e-05,
+      "loss": 3.9406,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667215465783696e-05,
+      "loss": 3.9365,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666376871032644e-05,
+      "loss": 3.9517,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665539914161966e-05,
+      "loss": 3.9418,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664701319410914e-05,
+      "loss": 3.9347,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663862724659862e-05,
+      "loss": 3.9238,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66302412990881e-05,
+      "loss": 3.9405,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662187173038131e-05,
+      "loss": 3.922,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661348578287079e-05,
+      "loss": 3.9327,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660509983536027e-05,
+      "loss": 3.9349,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659671388784975e-05,
+      "loss": 3.9315,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588344319142956e-05,
+      "loss": 3.9209,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579958371632436e-05,
+      "loss": 3.9382,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571572424121916e-05,
+      "loss": 3.9242,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563186476611396e-05,
+      "loss": 3.9285,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655481690790461e-05,
+      "loss": 3.9329,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654643096039409e-05,
+      "loss": 3.9244,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653804501288357e-05,
+      "loss": 3.9455,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652965906537305e-05,
+      "loss": 3.934,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652128949666626e-05,
+      "loss": 3.9342,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651290354915574e-05,
+      "loss": 3.9213,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650451760164522e-05,
+      "loss": 3.9301,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64961316541347e-05,
+      "loss": 3.9303,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648776208542791e-05,
+      "loss": 3.9161,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647937613791739e-05,
+      "loss": 3.9328,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647099019040687e-05,
+      "loss": 3.9263,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646260424289635e-05,
+      "loss": 3.9239,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645423467418956e-05,
+      "loss": 3.9216,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445848726679045e-05,
+      "loss": 3.9232,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437462779168525e-05,
+      "loss": 3.9166,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429076831658005e-05,
+      "loss": 3.9262,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420690884147485e-05,
+      "loss": 3.9127,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412321315440694e-05,
+      "loss": 3.9354,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403935367930174e-05,
+      "loss": 3.9322,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395549420419654e-05,
+      "loss": 3.9177,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387163472909134e-05,
+      "loss": 3.9151,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637879390420234e-05,
+      "loss": 3.9194,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637040795669182e-05,
+      "loss": 3.9253,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63620220091813e-05,
+      "loss": 3.9253,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635363606167078e-05,
+      "loss": 3.9182,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345266492964e-05,
+      "loss": 3.9272,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633688054545348e-05,
+      "loss": 3.9231,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632849459794296e-05,
+      "loss": 3.9282,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632010865043244e-05,
+      "loss": 3.9123,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631173908172565e-05,
+      "loss": 3.9281,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630335313421513e-05,
+      "loss": 3.9193,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629496718670461e-05,
+      "loss": 3.9122,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628659761799782e-05,
+      "loss": 3.9207,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62782116704873e-05,
+      "loss": 3.9224,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626982572297678e-05,
+      "loss": 3.9164,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626143977546626e-05,
+      "loss": 3.9072,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6253070206759466e-05,
+      "loss": 3.9092,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9577343463897705,
+      "eval_runtime": 303.0161,
+      "eval_samples_per_second": 1259.309,
+      "eval_steps_per_second": 39.354,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624468425924895e-05,
+      "loss": 3.9092,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623629831173843e-05,
+      "loss": 3.9007,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622791236422791e-05,
+      "loss": 3.926,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621954279552112e-05,
+      "loss": 3.9197,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.62111568480106e-05,
+      "loss": 3.9192,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620277090050008e-05,
+      "loss": 3.8996,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619438495298956e-05,
+      "loss": 3.9141,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618599900547904e-05,
+      "loss": 3.8917,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617762943677225e-05,
+      "loss": 3.9183,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616924348926173e-05,
+      "loss": 3.9117,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616085754175121e-05,
+      "loss": 3.9046,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615248797304442e-05,
+      "loss": 3.9158,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6144118404337635e-05,
+      "loss": 3.899,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135732456827115e-05,
+      "loss": 3.9026,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127346509316595e-05,
+      "loss": 3.8916,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6118960561806075e-05,
+      "loss": 3.8983,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110574614295555e-05,
+      "loss": 3.8968,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6102188666785035e-05,
+      "loss": 3.8947,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093802719274515e-05,
+      "loss": 3.8983,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085416771763995e-05,
+      "loss": 3.9199,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6077047203057204e-05,
+      "loss": 3.8983,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068661255546684e-05,
+      "loss": 3.9057,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6060275308036164e-05,
+      "loss": 3.9001,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051889360525644e-05,
+      "loss": 3.9063,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604351979181886e-05,
+      "loss": 3.8915,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603513384430834e-05,
+      "loss": 3.8998,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602674789679782e-05,
+      "loss": 3.8916,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60183619492873e-05,
+      "loss": 3.8917,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600999238058051e-05,
+      "loss": 3.8866,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600160643306999e-05,
+      "loss": 3.8926,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599322048555947e-05,
+      "loss": 3.8928,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598483453804895e-05,
+      "loss": 3.9036,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597646496934216e-05,
+      "loss": 3.8956,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596807902183164e-05,
+      "loss": 3.8964,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595969307432112e-05,
+      "loss": 3.8879,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59513071268106e-05,
+      "loss": 3.9012,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5942937558103813e-05,
+      "loss": 3.884,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5934551610593293e-05,
+      "loss": 3.8933,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5926165663082773e-05,
+      "loss": 3.8708,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5917779715572247e-05,
+      "loss": 3.8902,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590941014686546e-05,
+      "loss": 3.8828,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590102419935494e-05,
+      "loss": 3.9003,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589263825184442e-05,
+      "loss": 3.8837,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5884252304333896e-05,
+      "loss": 3.8869,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587588273562711e-05,
+      "loss": 3.8909,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586749678811659e-05,
+      "loss": 3.8971,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5859110840606065e-05,
+      "loss": 3.886,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585072489309555e-05,
+      "loss": 3.875,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584235532438877e-05,
+      "loss": 3.8689,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583396937687825e-05,
+      "loss": 3.8911,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582558342936772e-05,
+      "loss": 3.8886,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58171974818572e-05,
+      "loss": 3.8821,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808827913150416e-05,
+      "loss": 3.8753,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800441965639896e-05,
+      "loss": 3.8811,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579205601812937e-05,
+      "loss": 3.8623,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578367007061885e-05,
+      "loss": 3.8821,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5775300501912065e-05,
+      "loss": 3.8652,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576691455440154e-05,
+      "loss": 3.8788,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575852860689102e-05,
+      "loss": 3.8849,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5750142659380505e-05,
+      "loss": 3.8737,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741773090673714e-05,
+      "loss": 3.8644,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733387143163194e-05,
+      "loss": 3.8755,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5725001195652674e-05,
+      "loss": 3.8627,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571663162694589e-05,
+      "loss": 3.8637,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570824567943537e-05,
+      "loss": 3.8795,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569985973192484e-05,
+      "loss": 3.87,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569147378441432e-05,
+      "loss": 3.8641,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568310421570754e-05,
+      "loss": 3.8594,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567471826819701e-05,
+      "loss": 3.8679,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566633232068649e-05,
+      "loss": 3.8729,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565794637317597e-05,
+      "loss": 3.8745,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564957680446919e-05,
+      "loss": 3.8676,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564119085695867e-05,
+      "loss": 3.8782,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563280490944815e-05,
+      "loss": 3.8777,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562441896193763e-05,
+      "loss": 3.8659,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5616049393230843e-05,
+      "loss": 3.866,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560766344572032e-05,
+      "loss": 3.874,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55992774982098e-05,
+      "loss": 3.8693,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559090792950301e-05,
+      "loss": 3.8644,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582521981992486e-05,
+      "loss": 3.8645,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5574136034481966e-05,
+      "loss": 3.8712,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5565750086971446e-05,
+      "loss": 3.8729,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557364139460926e-05,
+      "loss": 3.8634,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548978191950405e-05,
+      "loss": 3.8512,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554059224443989e-05,
+      "loss": 3.8504,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553220629692937e-05,
+      "loss": 3.8614,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552383672822258e-05,
+      "loss": 3.8691,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551545078071206e-05,
+      "loss": 3.8623,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550706483320154e-05,
+      "loss": 3.8516,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549867888569102e-05,
+      "loss": 3.8691,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549030931698423e-05,
+      "loss": 3.8595,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548192336947371e-05,
+      "loss": 3.8561,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547353742196319e-05,
+      "loss": 3.8588,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546515147445267e-05,
+      "loss": 3.8632,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545678190574588e-05,
+      "loss": 3.8726,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544839595823536e-05,
+      "loss": 3.8554,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5440010010724846e-05,
+      "loss": 3.8443,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431624063214326e-05,
+      "loss": 3.8633,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5423254494507535e-05,
+      "loss": 3.8548,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414868546997015e-05,
+      "loss": 3.8695,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406482599486495e-05,
+      "loss": 3.8646,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5398096651975975e-05,
+      "loss": 3.8522,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389727083269184e-05,
+      "loss": 3.8463,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5381341135758664e-05,
+      "loss": 3.8617,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372955188248144e-05,
+      "loss": 3.8437,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364569240737624e-05,
+      "loss": 3.8542,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535619967203083e-05,
+      "loss": 3.8518,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534781372452031e-05,
+      "loss": 3.8567,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53394277770098e-05,
+      "loss": 3.8411,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533104182949928e-05,
+      "loss": 3.8603,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532267226079249e-05,
+      "loss": 3.8462,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531428631328197e-05,
+      "loss": 3.8494,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530590036577145e-05,
+      "loss": 3.8587,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529751441826093e-05,
+      "loss": 3.8471,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528914484955414e-05,
+      "loss": 3.8686,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528075890204362e-05,
+      "loss": 3.8584,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52723729545331e-05,
+      "loss": 3.8546,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526398700702258e-05,
+      "loss": 3.8452,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255617438315787e-05,
+      "loss": 3.8548,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5247231490805266e-05,
+      "loss": 3.8495,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523884554329475e-05,
+      "loss": 3.844,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523047597458796e-05,
+      "loss": 3.855,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522209002707744e-05,
+      "loss": 3.8503,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521370407956692e-05,
+      "loss": 3.8479,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52053181320564e-05,
+      "loss": 3.8459,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519693218454588e-05,
+      "loss": 3.8501,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518854623703536e-05,
+      "loss": 3.8432,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518016028952484e-05,
+      "loss": 3.8534,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517177434201432e-05,
+      "loss": 3.8394,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516342115211126e-05,
+      "loss": 3.8572,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515503520460074e-05,
+      "loss": 3.8535,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514664925709022e-05,
+      "loss": 3.8495,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513826330957971e-05,
+      "loss": 3.84,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512987736206919e-05,
+      "loss": 3.8464,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512149141455867e-05,
+      "loss": 3.8509,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511310546704815e-05,
+      "loss": 3.8528,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510471951953762e-05,
+      "loss": 3.8403,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096349950830836e-05,
+      "loss": 3.8523,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087964003320316e-05,
+      "loss": 3.8479,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079578055809796e-05,
+      "loss": 3.8577,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507119210829927e-05,
+      "loss": 3.8412,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062822539592485e-05,
+      "loss": 3.8535,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5054436592081965e-05,
+      "loss": 3.8484,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5046050644571445e-05,
+      "loss": 3.8403,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503768107586466e-05,
+      "loss": 3.8456,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502929512835414e-05,
+      "loss": 3.8533,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502090918084362e-05,
+      "loss": 3.8429,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012523233333094e-05,
+      "loss": 3.833,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5004137285822574e-05,
+      "loss": 3.8447,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9168317317962646,
+      "eval_runtime": 303.4989,
+      "eval_samples_per_second": 1257.306,
+      "eval_steps_per_second": 39.292,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995751338312054e-05,
+      "loss": 3.8312,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4987365390801534e-05,
+      "loss": 3.8296,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978979443291013e-05,
+      "loss": 3.8488,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970593495780493e-05,
+      "loss": 3.8502,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496220754826997e-05,
+      "loss": 3.8489,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495382160075945e-05,
+      "loss": 3.8357,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494543565324893e-05,
+      "loss": 3.8396,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493704970573841e-05,
+      "loss": 3.8207,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492868013703162e-05,
+      "loss": 3.8474,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492029418952111e-05,
+      "loss": 3.8434,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491190824201059e-05,
+      "loss": 3.8341,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490352229450007e-05,
+      "loss": 3.8454,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489516910459701e-05,
+      "loss": 3.8283,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488678315708649e-05,
+      "loss": 3.8335,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487839720957597e-05,
+      "loss": 3.8246,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487001126206545e-05,
+      "loss": 3.8239,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486162531455493e-05,
+      "loss": 3.8292,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485323936704441e-05,
+      "loss": 3.8267,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484485341953389e-05,
+      "loss": 3.8313,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483646747202337e-05,
+      "loss": 3.8478,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4828097903316576e-05,
+      "loss": 3.8322,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481971195580606e-05,
+      "loss": 3.8394,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481132600829554e-05,
+      "loss": 3.8308,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480294006078502e-05,
+      "loss": 3.8365,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479457049207823e-05,
+      "loss": 3.8254,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478618454456771e-05,
+      "loss": 3.8315,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477779859705719e-05,
+      "loss": 3.8289,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476941264954667e-05,
+      "loss": 3.8224,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476104308083988e-05,
+      "loss": 3.8192,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475265713332936e-05,
+      "loss": 3.8254,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474427118581884e-05,
+      "loss": 3.8259,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473590161711205e-05,
+      "loss": 3.8356,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472751566960153e-05,
+      "loss": 3.8311,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4719129722091016e-05,
+      "loss": 3.8308,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710743774580496e-05,
+      "loss": 3.8197,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4702374205873705e-05,
+      "loss": 3.8363,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693988258363185e-05,
+      "loss": 3.8205,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685602310852665e-05,
+      "loss": 3.8234,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4677216363342145e-05,
+      "loss": 3.8087,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668846794635354e-05,
+      "loss": 3.8225,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660460847124834e-05,
+      "loss": 3.8201,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4652074899614314e-05,
+      "loss": 3.8329,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643688952103794e-05,
+      "loss": 3.8186,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4635319383397003e-05,
+      "loss": 3.8243,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462693343588648e-05,
+      "loss": 3.8249,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461854748837597e-05,
+      "loss": 3.8314,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461016154086545e-05,
+      "loss": 3.8234,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460179197215866e-05,
+      "loss": 3.8108,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459340602464814e-05,
+      "loss": 3.8025,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458502007713762e-05,
+      "loss": 3.8308,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45766341296271e-05,
+      "loss": 3.8236,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456826456092031e-05,
+      "loss": 3.8177,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455987861340979e-05,
+      "loss": 3.8143,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455149266589927e-05,
+      "loss": 3.8159,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454310671838875e-05,
+      "loss": 3.7978,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453473714968196e-05,
+      "loss": 3.8152,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452635120217144e-05,
+      "loss": 3.8075,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517965254660924e-05,
+      "loss": 3.8144,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509579307150404e-05,
+      "loss": 3.8231,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450120973844361e-05,
+      "loss": 3.8097,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449282379093309e-05,
+      "loss": 3.801,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448443784342257e-05,
+      "loss": 3.8187,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447605189591205e-05,
+      "loss": 3.7974,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446768232720526e-05,
+      "loss": 3.8041,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445929637969474e-05,
+      "loss": 3.8132,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445091043218422e-05,
+      "loss": 3.8115,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442524484673695e-05,
+      "loss": 3.8017,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443415491596691e-05,
+      "loss": 3.8036,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442576896845639e-05,
+      "loss": 3.8029,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441738302094588e-05,
+      "loss": 3.8115,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440899707343535e-05,
+      "loss": 3.8107,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400627504728566e-05,
+      "loss": 3.8065,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4392241557218046e-05,
+      "loss": 3.8177,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4383855609707526e-05,
+      "loss": 3.8184,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4375469662197e-05,
+      "loss": 3.8059,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4367100093490215e-05,
+      "loss": 3.8007,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4358714145979695e-05,
+      "loss": 3.8192,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435032819846917e-05,
+      "loss": 3.8091,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4341958629762384e-05,
+      "loss": 3.8005,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333572682251864e-05,
+      "loss": 3.8059,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4325186734741344e-05,
+      "loss": 3.81,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316800787230824e-05,
+      "loss": 3.8173,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308414839720304e-05,
+      "loss": 3.7989,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4300028892209784e-05,
+      "loss": 3.7897,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291642944699264e-05,
+      "loss": 3.798,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4283256997188744e-05,
+      "loss": 3.7965,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427488742848195e-05,
+      "loss": 3.808,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426650148097143e-05,
+      "loss": 3.8045,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425811553346091e-05,
+      "loss": 3.7916,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424972958595039e-05,
+      "loss": 3.8099,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42413600172436e-05,
+      "loss": 3.7979,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423297406973308e-05,
+      "loss": 3.801,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422458812222257e-05,
+      "loss": 3.7968,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421620217471205e-05,
+      "loss": 3.8031,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420783260600526e-05,
+      "loss": 3.816,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419944665849474e-05,
+      "loss": 3.8025,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419106071098422e-05,
+      "loss": 3.782,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41826747634737e-05,
+      "loss": 3.8027,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417430519476691e-05,
+      "loss": 3.7983,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416591924725639e-05,
+      "loss": 3.8054,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415753329974587e-05,
+      "loss": 3.8095,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414914735223535e-05,
+      "loss": 3.7955,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4140777783528556e-05,
+      "loss": 3.7904,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4132391836018036e-05,
+      "loss": 3.8041,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412402226731125e-05,
+      "loss": 3.7824,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411563631980073e-05,
+      "loss": 3.7995,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410725037229021e-05,
+      "loss": 3.7921,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409886442477969e-05,
+      "loss": 3.8018,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409047847726917e-05,
+      "loss": 3.7858,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408209252975865e-05,
+      "loss": 3.8,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407370658224813e-05,
+      "loss": 3.7869,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406532063473761e-05,
+      "loss": 3.7924,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405695106603082e-05,
+      "loss": 3.8028,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40485651185203e-05,
+      "loss": 3.7958,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404017917100978e-05,
+      "loss": 3.8078,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403180960230299e-05,
+      "loss": 3.8004,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402342365479247e-05,
+      "loss": 3.7946,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4015037707281956e-05,
+      "loss": 3.7987,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4006651759771436e-05,
+      "loss": 3.7922,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3998265812260916e-05,
+      "loss": 3.7962,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3989879864750396e-05,
+      "loss": 3.7865,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981510296043605e-05,
+      "loss": 3.7998,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3973124348533085e-05,
+      "loss": 3.7894,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3964738401022565e-05,
+      "loss": 3.7898,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3956352453512045e-05,
+      "loss": 3.7884,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947966506001525e-05,
+      "loss": 3.7972,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3939580558491005e-05,
+      "loss": 3.7872,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3931210989784214e-05,
+      "loss": 3.7969,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922825042273694e-05,
+      "loss": 3.7849,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3914439094763174e-05,
+      "loss": 3.8019,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3906053147252654e-05,
+      "loss": 3.7956,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897667199742134e-05,
+      "loss": 3.7943,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388929763103535e-05,
+      "loss": 3.7872,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388091168352483e-05,
+      "loss": 3.7895,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38725257360143e-05,
+      "loss": 3.8002,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386413978850378e-05,
+      "loss": 3.7939,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385575384099326e-05,
+      "loss": 3.7839,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384736789348274e-05,
+      "loss": 3.7994,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383898194597222e-05,
+      "loss": 3.7946,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383061237726543e-05,
+      "loss": 3.8051,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382222642975491e-05,
+      "loss": 3.7887,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381384048224439e-05,
+      "loss": 3.7985,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380545453473388e-05,
+      "loss": 3.7929,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379706858722336e-05,
+      "loss": 3.7851,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378869901851657e-05,
+      "loss": 3.7943,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378031307100605e-05,
+      "loss": 3.7954,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377192712349553e-05,
+      "loss": 3.7928,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376354117598501e-05,
+      "loss": 3.7767,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375515522847449e-05,
+      "loss": 3.7901,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8920512199401855,
+      "eval_runtime": 302.1491,
+      "eval_samples_per_second": 1262.923,
+      "eval_steps_per_second": 39.467,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374676928096397e-05,
+      "loss": 3.784,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373838333345345e-05,
+      "loss": 3.7757,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372999738594293e-05,
+      "loss": 3.7944,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3721644196039865e-05,
+      "loss": 3.7973,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3713258248529345e-05,
+      "loss": 3.7937,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370487230101883e-05,
+      "loss": 3.7846,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369648635350831e-05,
+      "loss": 3.7844,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368810040599779e-05,
+      "loss": 3.7705,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3679730837291e-05,
+      "loss": 3.7919,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367134488978048e-05,
+      "loss": 3.7921,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366295894226996e-05,
+      "loss": 3.7814,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365457299475944e-05,
+      "loss": 3.7896,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364618704724892e-05,
+      "loss": 3.7796,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363781747854213e-05,
+      "loss": 3.7815,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362943153103161e-05,
+      "loss": 3.7716,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362104558352109e-05,
+      "loss": 3.7743,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361265963601057e-05,
+      "loss": 3.7763,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360427368850005e-05,
+      "loss": 3.7756,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359588774098953e-05,
+      "loss": 3.7793,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358750179347901e-05,
+      "loss": 3.7924,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3579132224772226e-05,
+      "loss": 3.7867,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3570762656065435e-05,
+      "loss": 3.7867,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3562376708554915e-05,
+      "loss": 3.7781,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3553990761044395e-05,
+      "loss": 3.7869,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3545604813533875e-05,
+      "loss": 3.7706,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3537218866023355e-05,
+      "loss": 3.7854,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3528832918512834e-05,
+      "loss": 3.7761,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3520446971002314e-05,
+      "loss": 3.7684,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351206102349179e-05,
+      "loss": 3.7716,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3503691454785003e-05,
+      "loss": 3.7716,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3495305507274483e-05,
+      "loss": 3.7783,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486919559763963e-05,
+      "loss": 3.7812,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347853361225344e-05,
+      "loss": 3.7838,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347018042235039e-05,
+      "loss": 3.7805,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346179447483987e-05,
+      "loss": 3.77,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345340852732935e-05,
+      "loss": 3.7818,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344502257981883e-05,
+      "loss": 3.7736,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343663663230831e-05,
+      "loss": 3.7688,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342826706360152e-05,
+      "loss": 3.7647,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419881116091e-05,
+      "loss": 3.7699,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341149516858048e-05,
+      "loss": 3.7741,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340310922106996e-05,
+      "loss": 3.7803,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339472327355944e-05,
+      "loss": 3.7693,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338633732604892e-05,
+      "loss": 3.7721,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337796775734213e-05,
+      "loss": 3.7765,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336958180983161e-05,
+      "loss": 3.7793,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3361195862321086e-05,
+      "loss": 3.7765,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3352809914810566e-05,
+      "loss": 3.7616,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3344423967300046e-05,
+      "loss": 3.7474,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3336038019789526e-05,
+      "loss": 3.7824,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3327668451082735e-05,
+      "loss": 3.7757,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3319282503572215e-05,
+      "loss": 3.7702,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3310896556061695e-05,
+      "loss": 3.7682,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3302510608551175e-05,
+      "loss": 3.7632,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3294124661040655e-05,
+      "loss": 3.7525,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328573871353014e-05,
+      "loss": 3.7655,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327735276601962e-05,
+      "loss": 3.7574,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326898319731283e-05,
+      "loss": 3.7661,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326059724980231e-05,
+      "loss": 3.7747,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325222768109552e-05,
+      "loss": 3.7602,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3243841733585e-05,
+      "loss": 3.7528,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323545578607448e-05,
+      "loss": 3.7716,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322706983856396e-05,
+      "loss": 3.7526,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321868389105344e-05,
+      "loss": 3.7528,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321031432234665e-05,
+      "loss": 3.7688,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320192837483613e-05,
+      "loss": 3.763,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319354242732561e-05,
+      "loss": 3.7548,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3185156479815095e-05,
+      "loss": 3.7547,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176770532304575e-05,
+      "loss": 3.7548,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3168384584794055e-05,
+      "loss": 3.7608,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3159998637283535e-05,
+      "loss": 3.768,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151612689773015e-05,
+      "loss": 3.7573,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3143243121066224e-05,
+      "loss": 3.7674,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313487355235943e-05,
+      "loss": 3.7687,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312648760484891e-05,
+      "loss": 3.7672,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311810165733839e-05,
+      "loss": 3.7508,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310971570982787e-05,
+      "loss": 3.7696,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310132976231735e-05,
+      "loss": 3.7655,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309296019361056e-05,
+      "loss": 3.7505,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308457424610005e-05,
+      "loss": 3.7659,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307618829858953e-05,
+      "loss": 3.7584,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306780235107901e-05,
+      "loss": 3.7698,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305941640356849e-05,
+      "loss": 3.7553,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30510468348617e-05,
+      "loss": 3.7403,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304266088735118e-05,
+      "loss": 3.7538,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303427493984066e-05,
+      "loss": 3.7535,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302588899233014e-05,
+      "loss": 3.7586,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301751942362335e-05,
+      "loss": 3.7576,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300913347611283e-05,
+      "loss": 3.7483,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300074752860231e-05,
+      "loss": 3.7599,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299236158109179e-05,
+      "loss": 3.7559,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.298397563358127e-05,
+      "loss": 3.7481,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297560606487448e-05,
+      "loss": 3.7545,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296722011736396e-05,
+      "loss": 3.7599,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295883416985344e-05,
+      "loss": 3.7703,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295044822234292e-05,
+      "loss": 3.754,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2942062274832396e-05,
+      "loss": 3.7409,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2933676327321876e-05,
+      "loss": 3.7576,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2925290379811355e-05,
+      "loss": 3.7511,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2916904432300835e-05,
+      "loss": 3.7598,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2908534863594045e-05,
+      "loss": 3.7641,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2900148916083525e-05,
+      "loss": 3.745,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289177934737674e-05,
+      "loss": 3.7461,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288339339986622e-05,
+      "loss": 3.7597,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28750074523557e-05,
+      "loss": 3.7377,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286662150484518e-05,
+      "loss": 3.7552,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285823555733466e-05,
+      "loss": 3.742,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284986598862787e-05,
+      "loss": 3.7591,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284148004111735e-05,
+      "loss": 3.7393,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283309409360683e-05,
+      "loss": 3.7556,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282470814609631e-05,
+      "loss": 3.7482,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281633857738952e-05,
+      "loss": 3.7437,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807952629879e-05,
+      "loss": 3.758,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279956668236848e-05,
+      "loss": 3.749,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279118073485796e-05,
+      "loss": 3.7638,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278279478734744e-05,
+      "loss": 3.7599,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2774425218640654e-05,
+      "loss": 3.7475,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2766039271130134e-05,
+      "loss": 3.756,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757653323619614e-05,
+      "loss": 3.7476,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2749267376109094e-05,
+      "loss": 3.7507,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27408978074023e-05,
+      "loss": 3.745,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273251185989178e-05,
+      "loss": 3.754,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272412591238126e-05,
+      "loss": 3.7464,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271573996487074e-05,
+      "loss": 3.7453,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270735401736022e-05,
+      "loss": 3.7489,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26989680698497e-05,
+      "loss": 3.751,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269059850114291e-05,
+      "loss": 3.7423,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268221255363239e-05,
+      "loss": 3.7544,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267382660612187e-05,
+      "loss": 3.7388,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266544065861136e-05,
+      "loss": 3.764,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265705471110084e-05,
+      "loss": 3.7464,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264868514239405e-05,
+      "loss": 3.7556,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264029919488353e-05,
+      "loss": 3.7411,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263191324737301e-05,
+      "loss": 3.7481,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262352729986249e-05,
+      "loss": 3.7542,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2615157731155696e-05,
+      "loss": 3.748,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2606771783645176e-05,
+      "loss": 3.7418,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2598385836134656e-05,
+      "loss": 3.7563,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2589999888624136e-05,
+      "loss": 3.7513,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2581630319917345e-05,
+      "loss": 3.7608,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2573244372406825e-05,
+      "loss": 3.7417,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256485842489631e-05,
+      "loss": 3.7539,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255647247738579e-05,
+      "loss": 3.7483,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254808652987527e-05,
+      "loss": 3.7444,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253971696116848e-05,
+      "loss": 3.751,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253133101365796e-05,
+      "loss": 3.7516,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252294506614744e-05,
+      "loss": 3.7547,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251455911863692e-05,
+      "loss": 3.7278,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250618954993013e-05,
+      "loss": 3.7531,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.87595272064209,
+      "eval_runtime": 302.116,
+      "eval_samples_per_second": 1263.061,
+      "eval_steps_per_second": 39.472,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249780360241961e-05,
+      "loss": 3.7403,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248941765490909e-05,
+      "loss": 3.7314,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248103170739857e-05,
+      "loss": 3.7498,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247264575988805e-05,
+      "loss": 3.758,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2464276191181266e-05,
+      "loss": 3.7514,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455890243670746e-05,
+      "loss": 3.7392,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2447504296160226e-05,
+      "loss": 3.7435,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24391183486497e-05,
+      "loss": 3.7305,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2430748779942915e-05,
+      "loss": 3.747,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2422362832432395e-05,
+      "loss": 3.7473,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2413976884921875e-05,
+      "loss": 3.7465,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240559093741135e-05,
+      "loss": 3.7452,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239720498990083e-05,
+      "loss": 3.7399,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2388835421194044e-05,
+      "loss": 3.7387,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2380449473683524e-05,
+      "loss": 3.7305,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2372063526173004e-05,
+      "loss": 3.7275,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363677578662484e-05,
+      "loss": 3.7366,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23553080099557e-05,
+      "loss": 3.7332,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234692206244517e-05,
+      "loss": 3.7371,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233853611493465e-05,
+      "loss": 3.7508,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233015016742413e-05,
+      "loss": 3.7495,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232178059871735e-05,
+      "loss": 3.7434,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231339465120682e-05,
+      "loss": 3.7387,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23050087036963e-05,
+      "loss": 3.7432,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229662275618578e-05,
+      "loss": 3.7283,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2288253187479e-05,
+      "loss": 3.7463,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227986723996847e-05,
+      "loss": 3.7348,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227148129245796e-05,
+      "loss": 3.7245,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226309534494744e-05,
+      "loss": 3.7317,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225470939743692e-05,
+      "loss": 3.7323,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2246339828730126e-05,
+      "loss": 3.7344,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237953881219606e-05,
+      "loss": 3.7429,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2229567933709086e-05,
+      "loss": 3.7409,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2221181986198566e-05,
+      "loss": 3.7402,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2212812417491775e-05,
+      "loss": 3.728,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2204426469981255e-05,
+      "loss": 3.749,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2196040522470735e-05,
+      "loss": 3.7314,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2187654574960215e-05,
+      "loss": 3.7241,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2179285006253424e-05,
+      "loss": 3.73,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217089905874291e-05,
+      "loss": 3.7257,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216251311123239e-05,
+      "loss": 3.7348,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215412716372187e-05,
+      "loss": 3.7426,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214575759501508e-05,
+      "loss": 3.7285,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213737164750456e-05,
+      "loss": 3.7308,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212898569999404e-05,
+      "loss": 3.7385,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212059975248352e-05,
+      "loss": 3.7421,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2112213804973e-05,
+      "loss": 3.7379,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210382785746248e-05,
+      "loss": 3.724,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209545828875569e-05,
+      "loss": 3.7013,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208707234124517e-05,
+      "loss": 3.7447,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207868639373465e-05,
+      "loss": 3.7369,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207030044622413e-05,
+      "loss": 3.7338,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061930877517345e-05,
+      "loss": 3.7267,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2053544930006824e-05,
+      "loss": 3.7233,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2045158982496304e-05,
+      "loss": 3.7148,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036773034985784e-05,
+      "loss": 3.7238,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2028387087475264e-05,
+      "loss": 3.7213,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2020017518768473e-05,
+      "loss": 3.7262,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011631571257953e-05,
+      "loss": 3.7372,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200324562374743e-05,
+      "loss": 3.7178,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199485967623691e-05,
+      "loss": 3.7147,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198649010753012e-05,
+      "loss": 3.7326,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19781041600196e-05,
+      "loss": 3.7165,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196971821250908e-05,
+      "loss": 3.7143,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196133226499856e-05,
+      "loss": 3.7275,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195296269629178e-05,
+      "loss": 3.7238,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194457674878126e-05,
+      "loss": 3.7155,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193619080127074e-05,
+      "loss": 3.7147,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192780485376022e-05,
+      "loss": 3.7153,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19194189062497e-05,
+      "loss": 3.7271,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191104933754291e-05,
+      "loss": 3.7239,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190266339003239e-05,
+      "loss": 3.7216,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189427744252187e-05,
+      "loss": 3.7257,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188589149501135e-05,
+      "loss": 3.7321,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1877521926304556e-05,
+      "loss": 3.7291,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1869135978794036e-05,
+      "loss": 3.714,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860750031283516e-05,
+      "loss": 3.7286,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1852364083772996e-05,
+      "loss": 3.7283,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184399451506621e-05,
+      "loss": 3.7069,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183560856755569e-05,
+      "loss": 3.7276,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182722262004517e-05,
+      "loss": 3.723,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181883667253465e-05,
+      "loss": 3.7276,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181045072502413e-05,
+      "loss": 3.7193,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180208115631734e-05,
+      "loss": 3.7042,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179369520880682e-05,
+      "loss": 3.7136,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17853092612963e-05,
+      "loss": 3.7157,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776923313785774e-05,
+      "loss": 3.7248,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176855374507899e-05,
+      "loss": 3.7155,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176016779756847e-05,
+      "loss": 3.7129,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175178185005795e-05,
+      "loss": 3.7213,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174339590254743e-05,
+      "loss": 3.7238,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1735026333840645e-05,
+      "loss": 3.7083,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1726640386330125e-05,
+      "loss": 3.7157,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1718254438819605e-05,
+      "loss": 3.7234,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170986849130908e-05,
+      "loss": 3.7298,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1701498922602294e-05,
+      "loss": 3.7168,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1693112975091774e-05,
+      "loss": 3.7055,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168472702758125e-05,
+      "loss": 3.7204,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167634108007073e-05,
+      "loss": 3.7107,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166797151136394e-05,
+      "loss": 3.7244,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165958556385342e-05,
+      "loss": 3.7281,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16511996163429e-05,
+      "loss": 3.7127,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164281366883238e-05,
+      "loss": 3.7106,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163442772132186e-05,
+      "loss": 3.7176,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162605815261508e-05,
+      "loss": 3.7011,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161767220510455e-05,
+      "loss": 3.7197,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160928625759403e-05,
+      "loss": 3.7014,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160090031008351e-05,
+      "loss": 3.7224,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159251436257299e-05,
+      "loss": 3.7052,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15841447938662e-05,
+      "loss": 3.7157,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157575884635568e-05,
+      "loss": 3.7113,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156737289884516e-05,
+      "loss": 3.7109,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155898695133464e-05,
+      "loss": 3.7179,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155061738262786e-05,
+      "loss": 3.7104,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154223143511734e-05,
+      "loss": 3.7283,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153384548760682e-05,
+      "loss": 3.7213,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15254595400963e-05,
+      "loss": 3.7154,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1517089971389506e-05,
+      "loss": 3.7182,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1508704023878986e-05,
+      "loss": 3.7103,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1500318076368466e-05,
+      "loss": 3.7153,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1491932128857946e-05,
+      "loss": 3.7113,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1483562560151155e-05,
+      "loss": 3.7115,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1475176612640635e-05,
+      "loss": 3.7131,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1466790665130115e-05,
+      "loss": 3.712,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1458404717619595e-05,
+      "loss": 3.7122,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145003514891281e-05,
+      "loss": 3.7135,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144164920140229e-05,
+      "loss": 3.71,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143326325389177e-05,
+      "loss": 3.715,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142487730638125e-05,
+      "loss": 3.7024,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141650773767446e-05,
+      "loss": 3.7247,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140812179016394e-05,
+      "loss": 3.7132,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139973584265342e-05,
+      "loss": 3.7212,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13913498951429e-05,
+      "loss": 3.7005,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138296394763238e-05,
+      "loss": 3.7163,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137459437892559e-05,
+      "loss": 3.7175,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136620843141507e-05,
+      "loss": 3.7178,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135782248390455e-05,
+      "loss": 3.7023,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1349436536394035e-05,
+      "loss": 3.7212,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1341066967687244e-05,
+      "loss": 3.7129,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332681020176724e-05,
+      "loss": 3.7271,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1324295072666204e-05,
+      "loss": 3.7041,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315909125155684e-05,
+      "loss": 3.7206,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130753955644889e-05,
+      "loss": 3.713,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129915360893837e-05,
+      "loss": 3.7105,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129076766142785e-05,
+      "loss": 3.7131,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128238171391733e-05,
+      "loss": 3.7174,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127401214521054e-05,
+      "loss": 3.7167,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126562619770002e-05,
+      "loss": 3.6957,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12572402501895e-05,
+      "loss": 3.7174,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8659157752990723,
+      "eval_runtime": 305.0055,
+      "eval_samples_per_second": 1251.095,
+      "eval_steps_per_second": 39.098,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124885430267898e-05,
+      "loss": 3.7047,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12404847339722e-05,
+      "loss": 3.697,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123209878646168e-05,
+      "loss": 3.7118,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122371283895116e-05,
+      "loss": 3.7218,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121532689144064e-05,
+      "loss": 3.7162,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120695732273385e-05,
+      "loss": 3.7048,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119857137522333e-05,
+      "loss": 3.7101,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119018542771281e-05,
+      "loss": 3.6964,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118179948020229e-05,
+      "loss": 3.7127,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1173429911495496e-05,
+      "loss": 3.7121,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1165043963984976e-05,
+      "loss": 3.7105,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1156658016474456e-05,
+      "loss": 3.7121,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1148272068963936e-05,
+      "loss": 3.7,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113988612145342e-05,
+      "loss": 3.7019,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113151655274663e-05,
+      "loss": 3.7008,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112313060523611e-05,
+      "loss": 3.6929,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111474465772559e-05,
+      "loss": 3.703,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110635871021507e-05,
+      "loss": 3.6989,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109798914150828e-05,
+      "loss": 3.7028,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108960319399776e-05,
+      "loss": 3.714,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108121724648724e-05,
+      "loss": 3.7158,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107283129897672e-05,
+      "loss": 3.7109,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106446173026993e-05,
+      "loss": 3.7009,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105607578275941e-05,
+      "loss": 3.7082,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104768983524889e-05,
+      "loss": 3.6976,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1039303887738376e-05,
+      "loss": 3.7112,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1030934319031585e-05,
+      "loss": 3.703,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1022548371521065e-05,
+      "loss": 3.6937,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1014162424010545e-05,
+      "loss": 3.7012,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005776476500025e-05,
+      "loss": 3.6932,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0997406907793234e-05,
+      "loss": 3.699,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0989020960282714e-05,
+      "loss": 3.7112,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0980635012772194e-05,
+      "loss": 3.7051,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0972249065261674e-05,
+      "loss": 3.7049,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096387949655488e-05,
+      "loss": 3.6936,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095549354904436e-05,
+      "loss": 3.7158,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094710760153384e-05,
+      "loss": 3.7013,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093872165402333e-05,
+      "loss": 3.6904,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093035208531654e-05,
+      "loss": 3.6988,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092196613780602e-05,
+      "loss": 3.6906,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09135801902955e-05,
+      "loss": 3.6991,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090519424278498e-05,
+      "loss": 3.7051,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089682467407819e-05,
+      "loss": 3.7011,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088843872656767e-05,
+      "loss": 3.6957,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088005277905715e-05,
+      "loss": 3.7072,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087166683154663e-05,
+      "loss": 3.7066,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08632808840361e-05,
+      "loss": 3.7077,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085491131532932e-05,
+      "loss": 3.6893,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08465253678188e-05,
+      "loss": 3.6661,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0838139420308277e-05,
+      "loss": 3.7144,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0829753472797757e-05,
+      "loss": 3.6994,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082138390409097e-05,
+      "loss": 3.7026,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081299795658045e-05,
+      "loss": 3.6968,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0804612009069926e-05,
+      "loss": 3.6849,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0796226061559406e-05,
+      "loss": 3.6856,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078785649285262e-05,
+      "loss": 3.6915,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07794705453421e-05,
+      "loss": 3.6861,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0771084597831575e-05,
+      "loss": 3.6903,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0762698650321054e-05,
+      "loss": 3.7091,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075432908161427e-05,
+      "loss": 3.6847,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074594313410375e-05,
+      "loss": 3.6801,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073755718659323e-05,
+      "loss": 3.7046,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072917123908271e-05,
+      "loss": 3.6799,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0720801670375926e-05,
+      "loss": 3.6813,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07124157228654e-05,
+      "loss": 3.6948,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070402977535488e-05,
+      "loss": 3.6916,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069564382784436e-05,
+      "loss": 3.68,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0687274259137575e-05,
+      "loss": 3.6896,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067888831162705e-05,
+      "loss": 3.6807,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067050236411653e-05,
+      "loss": 3.692,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066211641660601e-05,
+      "loss": 3.6934,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065373046909549e-05,
+      "loss": 3.6877,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0645360900388704e-05,
+      "loss": 3.695,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636974952878184e-05,
+      "loss": 3.6955,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0628589005367664e-05,
+      "loss": 3.7008,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0620203057857144e-05,
+      "loss": 3.6813,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061183348915035e-05,
+      "loss": 3.6943,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060344754163983e-05,
+      "loss": 3.6975,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059506159412931e-05,
+      "loss": 3.6736,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058667564661879e-05,
+      "loss": 3.699,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0578306077912e-05,
+      "loss": 3.6886,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056992013040148e-05,
+      "loss": 3.6958,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056153418289096e-05,
+      "loss": 3.6901,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055314823538044e-05,
+      "loss": 3.6749,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054477866667366e-05,
+      "loss": 3.6748,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053639271916314e-05,
+      "loss": 3.6846,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052800677165262e-05,
+      "loss": 3.6945,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05196208241421e-05,
+      "loss": 3.6824,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0511251255435307e-05,
+      "loss": 3.6819,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0502865307924787e-05,
+      "loss": 3.6888,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0494479360414267e-05,
+      "loss": 3.6899,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0486093412903746e-05,
+      "loss": 3.6803,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0477723844196956e-05,
+      "loss": 3.681,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0469337896686436e-05,
+      "loss": 3.6929,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0460951949175915e-05,
+      "loss": 3.6949,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0452566001665395e-05,
+      "loss": 3.6894,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044419643295861e-05,
+      "loss": 3.6729,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043581048544809e-05,
+      "loss": 3.6886,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042742453793757e-05,
+      "loss": 3.6796,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041903859042705e-05,
+      "loss": 3.6933,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041066902172026e-05,
+      "loss": 3.6896,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040228307420974e-05,
+      "loss": 3.6842,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039389712669922e-05,
+      "loss": 3.6818,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03855111791887e-05,
+      "loss": 3.6806,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037714161048191e-05,
+      "loss": 3.675,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036875566297139e-05,
+      "loss": 3.6856,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036036971546087e-05,
+      "loss": 3.6735,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035198376795035e-05,
+      "loss": 3.6912,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0343614199243565e-05,
+      "loss": 3.6774,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0335228251733045e-05,
+      "loss": 3.6806,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0326842304222525e-05,
+      "loss": 3.6786,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0318456356712005e-05,
+      "loss": 3.6753,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0310086788005214e-05,
+      "loss": 3.6905,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0301700840494694e-05,
+      "loss": 3.681,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0293314892984174e-05,
+      "loss": 3.697,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284928945473654e-05,
+      "loss": 3.6906,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027655937676686e-05,
+      "loss": 3.6841,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026817342925634e-05,
+      "loss": 3.6841,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025978748174582e-05,
+      "loss": 3.6817,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02514015342353e-05,
+      "loss": 3.6819,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024303196552852e-05,
+      "loss": 3.6842,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0234646018018e-05,
+      "loss": 3.6783,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022626007050748e-05,
+      "loss": 3.6819,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021787412299696e-05,
+      "loss": 3.6823,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020950455429017e-05,
+      "loss": 3.6812,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020111860677965e-05,
+      "loss": 3.6838,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019273265926913e-05,
+      "loss": 3.6757,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018434671175861e-05,
+      "loss": 3.6827,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0175977143051817e-05,
+      "loss": 3.6759,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0167591195541297e-05,
+      "loss": 3.6942,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0159205248030776e-05,
+      "loss": 3.6826,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0150819300520256e-05,
+      "loss": 3.6884,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014244973181347e-05,
+      "loss": 3.673,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013406378430295e-05,
+      "loss": 3.6843,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012567783679243e-05,
+      "loss": 3.6844,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011729188928191e-05,
+      "loss": 3.6851,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010892232057512e-05,
+      "loss": 3.6746,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01005363730646e-05,
+      "loss": 3.6881,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009215042555408e-05,
+      "loss": 3.6794,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008376447804356e-05,
+      "loss": 3.7008,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007539490933677e-05,
+      "loss": 3.6735,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006700896182625e-05,
+      "loss": 3.6899,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005862301431573e-05,
+      "loss": 3.682,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005023706680521e-05,
+      "loss": 3.6808,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0041867498098426e-05,
+      "loss": 3.6824,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0033481550587906e-05,
+      "loss": 3.6864,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0025095603077386e-05,
+      "loss": 3.6858,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0016709655566866e-05,
+      "loss": 3.6661,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0008340086860075e-05,
+      "loss": 3.6872,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.85886812210083,
+      "eval_runtime": 303.7232,
+      "eval_samples_per_second": 1256.377,
+      "eval_steps_per_second": 39.263,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9999954139349555e-05,
+      "loss": 3.6706,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9991568191839035e-05,
+      "loss": 3.6672,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9983182244328515e-05,
+      "loss": 3.6789,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9974796296817995e-05,
+      "loss": 3.6933,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9966426728111204e-05,
+      "loss": 3.6888,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9958040780600684e-05,
+      "loss": 3.673,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9949654833090164e-05,
+      "loss": 3.6847,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9941268885579644e-05,
+      "loss": 3.6653,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993289931687286e-05,
+      "loss": 3.6803,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992451336936234e-05,
+      "loss": 3.6857,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991612742185182e-05,
+      "loss": 3.6777,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99077414743413e-05,
+      "loss": 3.6793,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989937190563451e-05,
+      "loss": 3.6781,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989098595812399e-05,
+      "loss": 3.6707,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988260001061347e-05,
+      "loss": 3.6669,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987421406310295e-05,
+      "loss": 3.6646,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986584449439616e-05,
+      "loss": 3.6704,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985745854688564e-05,
+      "loss": 3.6744,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984907259937512e-05,
+      "loss": 3.6681,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98406866518646e-05,
+      "loss": 3.6841,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983231708315781e-05,
+      "loss": 3.6869,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982393113564729e-05,
+      "loss": 3.6837,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981554518813677e-05,
+      "loss": 3.6731,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9807159240626246e-05,
+      "loss": 3.6749,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979878967191946e-05,
+      "loss": 3.6746,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979040372440894e-05,
+      "loss": 3.6747,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978201777689842e-05,
+      "loss": 3.676,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9773631829387895e-05,
+      "loss": 3.6668,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976526226068111e-05,
+      "loss": 3.6692,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975687631317059e-05,
+      "loss": 3.6672,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974849036566007e-05,
+      "loss": 3.6663,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974010441814955e-05,
+      "loss": 3.6815,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973173484944277e-05,
+      "loss": 3.6786,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972334890193225e-05,
+      "loss": 3.6759,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971496295442172e-05,
+      "loss": 3.6673,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.97065770069112e-05,
+      "loss": 3.6794,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9698207438204416e-05,
+      "loss": 3.673,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9689821490693896e-05,
+      "loss": 3.6645,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968143554318337e-05,
+      "loss": 3.6696,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967304959567285e-05,
+      "loss": 3.6604,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9664680026966065e-05,
+      "loss": 3.6679,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9656294079455545e-05,
+      "loss": 3.6764,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9647908131945025e-05,
+      "loss": 3.67,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9639522184434505e-05,
+      "loss": 3.6706,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963115261572772e-05,
+      "loss": 3.6794,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9622766668217194e-05,
+      "loss": 3.6752,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9614380720706674e-05,
+      "loss": 3.6745,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9605994773196154e-05,
+      "loss": 3.6661,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959762520448937e-05,
+      "loss": 3.6389,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958923925697884e-05,
+      "loss": 3.682,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958085330946832e-05,
+      "loss": 3.6685,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95724673619578e-05,
+      "loss": 3.6741,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956409779325101e-05,
+      "loss": 3.6704,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95557118457405e-05,
+      "loss": 3.6586,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954732589822998e-05,
+      "loss": 3.6532,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953893995071946e-05,
+      "loss": 3.6632,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953057038201267e-05,
+      "loss": 3.659,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952218443450215e-05,
+      "loss": 3.6599,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951379848699163e-05,
+      "loss": 3.6798,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950541253948111e-05,
+      "loss": 3.6575,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9497042970774316e-05,
+      "loss": 3.6514,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9488657023263796e-05,
+      "loss": 3.6766,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9480271075753276e-05,
+      "loss": 3.6496,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9471885128242756e-05,
+      "loss": 3.6526,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9463515559535965e-05,
+      "loss": 3.6698,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945512961202545e-05,
+      "loss": 3.6649,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944674366451493e-05,
+      "loss": 3.653,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943835771700441e-05,
+      "loss": 3.6574,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942998814829762e-05,
+      "loss": 3.6536,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94216022007871e-05,
+      "loss": 3.6642,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941321625327658e-05,
+      "loss": 3.6603,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940483030576606e-05,
+      "loss": 3.6599,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939646073705927e-05,
+      "loss": 3.6657,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938807478954875e-05,
+      "loss": 3.6678,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937968884203823e-05,
+      "loss": 3.6733,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937130289452771e-05,
+      "loss": 3.6527,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936293332582092e-05,
+      "loss": 3.6678,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93545473783104e-05,
+      "loss": 3.669,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9346161430799886e-05,
+      "loss": 3.6441,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9337775483289366e-05,
+      "loss": 3.6684,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9329405914582575e-05,
+      "loss": 3.6654,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9321019967072055e-05,
+      "loss": 3.6701,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9312634019561535e-05,
+      "loss": 3.6611,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9304248072051015e-05,
+      "loss": 3.6506,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9295878503344224e-05,
+      "loss": 3.6423,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9287492555833704e-05,
+      "loss": 3.6585,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9279106608323184e-05,
+      "loss": 3.6669,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9270720660812664e-05,
+      "loss": 3.6555,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926235109210587e-05,
+      "loss": 3.652,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925396514459535e-05,
+      "loss": 3.6593,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924557919708484e-05,
+      "loss": 3.6673,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923719324957432e-05,
+      "loss": 3.6537,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922882368086753e-05,
+      "loss": 3.6556,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922043773335701e-05,
+      "loss": 3.6621,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921205178584649e-05,
+      "loss": 3.6702,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920366583833597e-05,
+      "loss": 3.6595,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919529626962918e-05,
+      "loss": 3.6505,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918691032211866e-05,
+      "loss": 3.6543,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917852437460814e-05,
+      "loss": 3.6538,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917013842709762e-05,
+      "loss": 3.6651,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9161768858390826e-05,
+      "loss": 3.6642,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9153382910880306e-05,
+      "loss": 3.6585,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914499696336979e-05,
+      "loss": 3.652,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913661101585927e-05,
+      "loss": 3.6508,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912824144715248e-05,
+      "loss": 3.6521,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911985549964196e-05,
+      "loss": 3.657,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911146955213144e-05,
+      "loss": 3.6455,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910308360462092e-05,
+      "loss": 3.6678,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909471403591413e-05,
+      "loss": 3.6501,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908632808840361e-05,
+      "loss": 3.6532,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907794214089309e-05,
+      "loss": 3.6496,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906955619338257e-05,
+      "loss": 3.6463,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906118662467578e-05,
+      "loss": 3.6661,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905280067716526e-05,
+      "loss": 3.6491,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904441472965475e-05,
+      "loss": 3.6716,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903602878214423e-05,
+      "loss": 3.6643,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9027659213437436e-05,
+      "loss": 3.657,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9019273265926916e-05,
+      "loss": 3.655,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9010887318416396e-05,
+      "loss": 3.6606,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9002501370905876e-05,
+      "loss": 3.6534,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8994131802199085e-05,
+      "loss": 3.6587,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8985745854688565e-05,
+      "loss": 3.6519,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8977359907178045e-05,
+      "loss": 3.6581,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8968973959667525e-05,
+      "loss": 3.6522,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8960604390960734e-05,
+      "loss": 3.6565,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8952218443450214e-05,
+      "loss": 3.6576,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89438324959397e-05,
+      "loss": 3.647,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893544654842918e-05,
+      "loss": 3.6598,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892707697972239e-05,
+      "loss": 3.6479,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891869103221187e-05,
+      "loss": 3.6655,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891030508470135e-05,
+      "loss": 3.6575,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890191913719083e-05,
+      "loss": 3.6596,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889354956848404e-05,
+      "loss": 3.6453,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888516362097352e-05,
+      "loss": 3.6632,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8876777673463e-05,
+      "loss": 3.6534,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886840810475621e-05,
+      "loss": 3.6584,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886002215724569e-05,
+      "loss": 3.6492,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885163620973517e-05,
+      "loss": 3.6592,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8843250262224654e-05,
+      "loss": 3.6558,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883488069351786e-05,
+      "loss": 3.6744,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882649474600734e-05,
+      "loss": 3.6467,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881810879849682e-05,
+      "loss": 3.6615,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88097228509863e-05,
+      "loss": 3.6566,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880135328227951e-05,
+      "loss": 3.6585,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879298371357272e-05,
+      "loss": 3.6521,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.87845977660622e-05,
+      "loss": 3.6607,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877621181855168e-05,
+      "loss": 3.6577,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876782587104116e-05,
+      "loss": 3.6461,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875943992353064e-05,
+      "loss": 3.6568,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8543734550476074,
+      "eval_runtime": 305.9458,
+      "eval_samples_per_second": 1247.25,
+      "eval_steps_per_second": 38.977,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.875105397602012e-05,
+      "loss": 3.6469,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874266802850961e-05,
+      "loss": 3.6401,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.873428208099909e-05,
+      "loss": 3.652,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.87259125122923e-05,
+      "loss": 3.6699,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871752656478178e-05,
+      "loss": 3.6639,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870914061727126e-05,
+      "loss": 3.6425,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870075466976074e-05,
+      "loss": 3.6586,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8692385101053946e-05,
+      "loss": 3.6423,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8683999153543426e-05,
+      "loss": 3.6521,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8675629584836635e-05,
+      "loss": 3.6626,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8667243637326115e-05,
+      "loss": 3.6498,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8658857689815595e-05,
+      "loss": 3.6515,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8650471742305075e-05,
+      "loss": 3.6543,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8642085794794555e-05,
+      "loss": 3.6446,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.863369984728404e-05,
+      "loss": 3.6431,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862533027857725e-05,
+      "loss": 3.6379,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861694433106673e-05,
+      "loss": 3.6436,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.860855838355621e-05,
+      "loss": 3.6503,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.860017243604569e-05,
+      "loss": 3.642,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8591786488535164e-05,
+      "loss": 3.6582,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8583400541024643e-05,
+      "loss": 3.6603,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8575014593514123e-05,
+      "loss": 3.6577,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.856664502480733e-05,
+      "loss": 3.6495,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855825907729681e-05,
+      "loss": 3.645,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854987312978629e-05,
+      "loss": 3.6517,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854148718227578e-05,
+      "loss": 3.6496,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853311761356899e-05,
+      "loss": 3.6472,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852473166605847e-05,
+      "loss": 3.6453,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851634571854795e-05,
+      "loss": 3.6415,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.850795977103743e-05,
+      "loss": 3.6436,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849959020233064e-05,
+      "loss": 3.6426,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849120425482012e-05,
+      "loss": 3.653,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84828183073096e-05,
+      "loss": 3.6598,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847443235979908e-05,
+      "loss": 3.6476,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8466062791092286e-05,
+      "loss": 3.6452,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8457676843581766e-05,
+      "loss": 3.6438,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8449290896071246e-05,
+      "loss": 3.6538,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.844090494856073e-05,
+      "loss": 3.64,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843253537985394e-05,
+      "loss": 3.6445,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.842414943234342e-05,
+      "loss": 3.6339,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84157634848329e-05,
+      "loss": 3.6477,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840737753732238e-05,
+      "loss": 3.6463,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839900796861559e-05,
+      "loss": 3.6468,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839062202110507e-05,
+      "loss": 3.646,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838223607359455e-05,
+      "loss": 3.6505,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837385012608403e-05,
+      "loss": 3.6507,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836548055737724e-05,
+      "loss": 3.652,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835709460986672e-05,
+      "loss": 3.6414,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83487086623562e-05,
+      "loss": 3.6125,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8340322714845687e-05,
+      "loss": 3.6538,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8331953146138896e-05,
+      "loss": 3.6448,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8323567198628376e-05,
+      "loss": 3.6492,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8315181251117856e-05,
+      "loss": 3.6464,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8306795303607335e-05,
+      "loss": 3.6352,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8298425734900545e-05,
+      "loss": 3.6292,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8290039787390025e-05,
+      "loss": 3.6367,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8281653839879504e-05,
+      "loss": 3.6361,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8273267892368984e-05,
+      "loss": 3.6341,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8264898323662194e-05,
+      "loss": 3.6545,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8256512376151674e-05,
+      "loss": 3.6379,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8248126428641153e-05,
+      "loss": 3.6204,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823974048113064e-05,
+      "loss": 3.6538,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823137091242385e-05,
+      "loss": 3.6257,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822298496491333e-05,
+      "loss": 3.6255,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821459901740281e-05,
+      "loss": 3.6444,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820621306989229e-05,
+      "loss": 3.6369,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81978435011855e-05,
+      "loss": 3.6337,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818945755367498e-05,
+      "loss": 3.6324,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818107160616446e-05,
+      "loss": 3.6316,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817268565865394e-05,
+      "loss": 3.6391,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816431608994715e-05,
+      "loss": 3.6352,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815593014243663e-05,
+      "loss": 3.6393,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814754419492611e-05,
+      "loss": 3.6388,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8139158247415594e-05,
+      "loss": 3.6418,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81307886787088e-05,
+      "loss": 3.6511,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812240273119828e-05,
+      "loss": 3.6276,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811401678368776e-05,
+      "loss": 3.6421,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810563083617724e-05,
+      "loss": 3.6456,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.809726126747045e-05,
+      "loss": 3.62,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808887531995993e-05,
+      "loss": 3.6389,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808048937244941e-05,
+      "loss": 3.6425,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807210342493889e-05,
+      "loss": 3.6432,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80637338562321e-05,
+      "loss": 3.6426,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805534790872158e-05,
+      "loss": 3.6221,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804696196121106e-05,
+      "loss": 3.6172,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803857601370055e-05,
+      "loss": 3.6378,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803020644499376e-05,
+      "loss": 3.6408,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8021820497483237e-05,
+      "loss": 3.6271,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8013434549972717e-05,
+      "loss": 3.633,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8005048602462196e-05,
+      "loss": 3.6318,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7996679033755406e-05,
+      "loss": 3.6431,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7988293086244886e-05,
+      "loss": 3.6331,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7979907138734366e-05,
+      "loss": 3.6329,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7971521191223845e-05,
+      "loss": 3.6325,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7963151622517055e-05,
+      "loss": 3.6445,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7954765675006535e-05,
+      "loss": 3.638,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7946379727496014e-05,
+      "loss": 3.6245,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79379937799855e-05,
+      "loss": 3.6313,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792962421127871e-05,
+      "loss": 3.6309,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792123826376819e-05,
+      "loss": 3.6401,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791285231625767e-05,
+      "loss": 3.638,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.790446636874715e-05,
+      "loss": 3.6342,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789609680004036e-05,
+      "loss": 3.6297,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788771085252984e-05,
+      "loss": 3.6258,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787934128382305e-05,
+      "loss": 3.6315,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787095533631253e-05,
+      "loss": 3.6277,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786256938880201e-05,
+      "loss": 3.6242,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785418344129149e-05,
+      "loss": 3.6424,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.784579749378097e-05,
+      "loss": 3.627,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7837411546270455e-05,
+      "loss": 3.6303,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7829025598759935e-05,
+      "loss": 3.627,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7820639651249415e-05,
+      "loss": 3.6219,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7812270082542624e-05,
+      "loss": 3.6413,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7803884135032104e-05,
+      "loss": 3.6276,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7795498187521584e-05,
+      "loss": 3.6452,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7787112240011064e-05,
+      "loss": 3.6387,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777874267130427e-05,
+      "loss": 3.6349,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777035672379375e-05,
+      "loss": 3.6318,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776197077628323e-05,
+      "loss": 3.6364,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.775358482877271e-05,
+      "loss": 3.6259,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.774521526006592e-05,
+      "loss": 3.6366,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77368293125554e-05,
+      "loss": 3.6312,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772845974384862e-05,
+      "loss": 3.6302,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77200737963381e-05,
+      "loss": 3.6258,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771168784882758e-05,
+      "loss": 3.6331,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770330190131706e-05,
+      "loss": 3.6341,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769491595380654e-05,
+      "loss": 3.6247,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768653000629601e-05,
+      "loss": 3.6365,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767814405878549e-05,
+      "loss": 3.6175,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766975811127497e-05,
+      "loss": 3.6494,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7661388542568186e-05,
+      "loss": 3.6309,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765300259505766e-05,
+      "loss": 3.6367,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764461664754714e-05,
+      "loss": 3.6227,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7636230700036626e-05,
+      "loss": 3.6386,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7627861131329835e-05,
+      "loss": 3.6287,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7619475183819315e-05,
+      "loss": 3.6351,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7611089236308795e-05,
+      "loss": 3.6276,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760271966760201e-05,
+      "loss": 3.6313,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7594333720091484e-05,
+      "loss": 3.6325,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7585947772580964e-05,
+      "loss": 3.6525,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7577561825070444e-05,
+      "loss": 3.6231,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756919225636366e-05,
+      "loss": 3.6394,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756080630885313e-05,
+      "loss": 3.6346,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755242036134261e-05,
+      "loss": 3.633,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754405079263583e-05,
+      "loss": 3.626,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753566484512531e-05,
+      "loss": 3.6368,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752727889761479e-05,
+      "loss": 3.6313,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751889295010427e-05,
+      "loss": 3.6252,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751050700259375e-05,
+      "loss": 3.6319,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.851743459701538,
+      "eval_runtime": 306.1531,
+      "eval_samples_per_second": 1246.406,
+      "eval_steps_per_second": 38.951,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750212105508323e-05,
+      "loss": 3.6235,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749373510757271e-05,
+      "loss": 3.6145,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748534916006219e-05,
+      "loss": 3.6315,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747696321255167e-05,
+      "loss": 3.6447,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746859364384488e-05,
+      "loss": 3.6423,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746020769633436e-05,
+      "loss": 3.6249,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745182174882384e-05,
+      "loss": 3.6314,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744343580131332e-05,
+      "loss": 3.6227,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7435066232606534e-05,
+      "loss": 3.6254,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7426680285096014e-05,
+      "loss": 3.6383,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7418294337585494e-05,
+      "loss": 3.6292,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7409908390074973e-05,
+      "loss": 3.6302,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740153882136818e-05,
+      "loss": 3.633,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.739316925266139e-05,
+      "loss": 3.6209,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738478330515087e-05,
+      "loss": 3.618,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737639735764035e-05,
+      "loss": 3.6121,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736801141012983e-05,
+      "loss": 3.624,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735962546261931e-05,
+      "loss": 3.6276,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735123951510879e-05,
+      "loss": 3.6172,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734285356759827e-05,
+      "loss": 3.6323,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733446762008775e-05,
+      "loss": 3.6374,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732609805138097e-05,
+      "loss": 3.6326,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731771210387045e-05,
+      "loss": 3.6271,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730932615635993e-05,
+      "loss": 3.6253,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730094020884941e-05,
+      "loss": 3.6292,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7292570640142616e-05,
+      "loss": 3.6276,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7284184692632096e-05,
+      "loss": 3.6217,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7275798745121576e-05,
+      "loss": 3.62,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7267412797611056e-05,
+      "loss": 3.6237,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7259043228904265e-05,
+      "loss": 3.6193,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7250657281393745e-05,
+      "loss": 3.6208,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7242271333883225e-05,
+      "loss": 3.6283,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7233885386372705e-05,
+      "loss": 3.6316,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722551581766592e-05,
+      "loss": 3.6299,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.72171298701554e-05,
+      "loss": 3.6166,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720874392264488e-05,
+      "loss": 3.6268,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720035797513436e-05,
+      "loss": 3.6307,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719198840642757e-05,
+      "loss": 3.6128,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.718360245891705e-05,
+      "loss": 3.624,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717521651140653e-05,
+      "loss": 3.6123,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.716684694269974e-05,
+      "loss": 3.6248,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.715846099518922e-05,
+      "loss": 3.6201,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71500750476787e-05,
+      "loss": 3.6235,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714168910016818e-05,
+      "loss": 3.6258,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7133319531461395e-05,
+      "loss": 3.6283,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7124933583950875e-05,
+      "loss": 3.6296,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7116547636440355e-05,
+      "loss": 3.6312,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7108161688929834e-05,
+      "loss": 3.6186,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7099792120223044e-05,
+      "loss": 3.5937,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7091406172712524e-05,
+      "loss": 3.6312,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7083020225202003e-05,
+      "loss": 3.621,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7074634277691483e-05,
+      "loss": 3.626,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.706626470898469e-05,
+      "loss": 3.6231,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.705787876147417e-05,
+      "loss": 3.614,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704949281396365e-05,
+      "loss": 3.6039,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704110686645313e-05,
+      "loss": 3.6137,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703273729774635e-05,
+      "loss": 3.6173,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702435135023583e-05,
+      "loss": 3.6055,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.701596540272531e-05,
+      "loss": 3.6329,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.700757945521479e-05,
+      "loss": 3.6162,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6999209886508e-05,
+      "loss": 3.6006,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699082393899748e-05,
+      "loss": 3.6276,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698243799148696e-05,
+      "loss": 3.6063,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697405204397644e-05,
+      "loss": 3.605,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6965682475269646e-05,
+      "loss": 3.6199,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6957296527759126e-05,
+      "loss": 3.6147,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6948910580248606e-05,
+      "loss": 3.6098,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6940524632738086e-05,
+      "loss": 3.608,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6932155064031295e-05,
+      "loss": 3.6089,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.692376911652078e-05,
+      "loss": 3.6131,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691538316901026e-05,
+      "loss": 3.6196,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.690699722149974e-05,
+      "loss": 3.6147,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689862765279295e-05,
+      "loss": 3.6175,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689024170528243e-05,
+      "loss": 3.6122,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688185575777191e-05,
+      "loss": 3.6324,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6873469810261384e-05,
+      "loss": 3.6096,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68651002415546e-05,
+      "loss": 3.6198,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685671429404408e-05,
+      "loss": 3.6184,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684832834653356e-05,
+      "loss": 3.6053,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683994239902303e-05,
+      "loss": 3.6152,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683157283031625e-05,
+      "loss": 3.6192,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6823186882805736e-05,
+      "loss": 3.6193,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6814800935295216e-05,
+      "loss": 3.6222,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680641498778469e-05,
+      "loss": 3.5969,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6798045419077905e-05,
+      "loss": 3.5993,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6789659471567385e-05,
+      "loss": 3.6138,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678127352405686e-05,
+      "loss": 3.6185,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677288757654634e-05,
+      "loss": 3.6056,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6764518007839554e-05,
+      "loss": 3.6117,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6756132060329033e-05,
+      "loss": 3.6114,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.674774611281851e-05,
+      "loss": 3.6194,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673936016530799e-05,
+      "loss": 3.6102,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.67309905966012e-05,
+      "loss": 3.6099,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672260464909069e-05,
+      "loss": 3.6139,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.671421870158016e-05,
+      "loss": 3.621,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.670584913287338e-05,
+      "loss": 3.6164,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669746318536286e-05,
+      "loss": 3.601,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668907723785233e-05,
+      "loss": 3.6069,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668069129034181e-05,
+      "loss": 3.6118,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667232172163503e-05,
+      "loss": 3.6179,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.666393577412451e-05,
+      "loss": 3.6175,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.665554982661398e-05,
+      "loss": 3.6141,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664716387910346e-05,
+      "loss": 3.6064,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6638794310396676e-05,
+      "loss": 3.6033,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6630408362886156e-05,
+      "loss": 3.6108,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6622022415375636e-05,
+      "loss": 3.6045,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6613636467865116e-05,
+      "loss": 3.6051,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.660526689915833e-05,
+      "loss": 3.6187,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6596880951647805e-05,
+      "loss": 3.606,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6588495004137285e-05,
+      "loss": 3.6065,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6580109056626765e-05,
+      "loss": 3.604,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657173948791998e-05,
+      "loss": 3.6002,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6563353540409454e-05,
+      "loss": 3.6197,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6554967592898934e-05,
+      "loss": 3.6115,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6546581645388414e-05,
+      "loss": 3.6178,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.653821207668163e-05,
+      "loss": 3.6174,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652982612917111e-05,
+      "loss": 3.6159,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652144018166059e-05,
+      "loss": 3.6104,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651305423415007e-05,
+      "loss": 3.6144,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650468466544328e-05,
+      "loss": 3.6073,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649629871793276e-05,
+      "loss": 3.6146,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648791277042224e-05,
+      "loss": 3.6069,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647952682291172e-05,
+      "loss": 3.6133,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647115725420493e-05,
+      "loss": 3.6022,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646277130669441e-05,
+      "loss": 3.6132,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.645438535918389e-05,
+      "loss": 3.6086,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644599941167337e-05,
+      "loss": 3.6067,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6437629842966584e-05,
+      "loss": 3.6109,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6429243895456064e-05,
+      "loss": 3.5964,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6420857947945543e-05,
+      "loss": 3.6284,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6412472000435023e-05,
+      "loss": 3.6078,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.640410243172823e-05,
+      "loss": 3.6184,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.639571648421771e-05,
+      "loss": 3.6013,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.638733053670719e-05,
+      "loss": 3.6129,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637894458919667e-05,
+      "loss": 3.6114,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637057502048988e-05,
+      "loss": 3.6142,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636218907297936e-05,
+      "loss": 3.6058,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635380312546884e-05,
+      "loss": 3.6079,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634541717795832e-05,
+      "loss": 3.614,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633704760925154e-05,
+      "loss": 3.6329,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632866166174102e-05,
+      "loss": 3.5988,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63202757142305e-05,
+      "loss": 3.62,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631188976671998e-05,
+      "loss": 3.6141,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6303520198013186e-05,
+      "loss": 3.6073,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.62951506293064e-05,
+      "loss": 3.6072,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6286764681795875e-05,
+      "loss": 3.6149,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6278378734285355e-05,
+      "loss": 3.613,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6269992786774835e-05,
+      "loss": 3.6099,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6261606839264315e-05,
+      "loss": 3.6089,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8503987789154053,
+      "eval_runtime": 316.5647,
+      "eval_samples_per_second": 1205.412,
+      "eval_steps_per_second": 37.67,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6253220891753795e-05,
+      "loss": 3.6042,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6244834944243275e-05,
+      "loss": 3.5935,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6236448996732755e-05,
+      "loss": 3.6098,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622807942802597e-05,
+      "loss": 3.6216,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621969348051545e-05,
+      "loss": 3.6232,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621130753300493e-05,
+      "loss": 3.6023,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620292158549441e-05,
+      "loss": 3.6143,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.619455201678762e-05,
+      "loss": 3.6027,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61861660692771e-05,
+      "loss": 3.6025,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617779650057031e-05,
+      "loss": 3.6138,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616941055305979e-05,
+      "loss": 3.6145,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616102460554927e-05,
+      "loss": 3.6128,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615263865803875e-05,
+      "loss": 3.6069,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6144269089331965e-05,
+      "loss": 3.6001,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6135883141821445e-05,
+      "loss": 3.6003,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6127497194310925e-05,
+      "loss": 3.588,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6119111246800404e-05,
+      "loss": 3.6068,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6110725299289884e-05,
+      "loss": 3.6053,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6102339351779364e-05,
+      "loss": 3.5956,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6093953404268844e-05,
+      "loss": 3.6098,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6085567456758324e-05,
+      "loss": 3.6186,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607719788805153e-05,
+      "loss": 3.6097,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606881194054101e-05,
+      "loss": 3.6084,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606042599303049e-05,
+      "loss": 3.6029,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605204004551997e-05,
+      "loss": 3.6108,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.604367047681318e-05,
+      "loss": 3.6063,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603528452930266e-05,
+      "loss": 3.6006,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602689858179214e-05,
+      "loss": 3.5981,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601851263428163e-05,
+      "loss": 3.6019,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601014306557484e-05,
+      "loss": 3.5984,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600175711806432e-05,
+      "loss": 3.6056,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59933711705538e-05,
+      "loss": 3.6049,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598498522304328e-05,
+      "loss": 3.6052,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597661565433649e-05,
+      "loss": 3.6133,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596822970682597e-05,
+      "loss": 3.5957,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595984375931545e-05,
+      "loss": 3.6058,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595145781180493e-05,
+      "loss": 3.6095,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5943088243098136e-05,
+      "loss": 3.5916,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5934702295587616e-05,
+      "loss": 3.6039,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5926316348077096e-05,
+      "loss": 3.5931,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591793040056658e-05,
+      "loss": 3.5996,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590956083185979e-05,
+      "loss": 3.6009,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590117488434927e-05,
+      "loss": 3.6035,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589278893683875e-05,
+      "loss": 3.6063,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588441936813196e-05,
+      "loss": 3.6059,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587603342062144e-05,
+      "loss": 3.6087,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586764747311092e-05,
+      "loss": 3.6086,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58592615256004e-05,
+      "loss": 3.5998,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585089195689361e-05,
+      "loss": 3.5741,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584250600938309e-05,
+      "loss": 3.6121,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583412006187257e-05,
+      "loss": 3.6007,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582573411436205e-05,
+      "loss": 3.6087,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5817364545655265e-05,
+      "loss": 3.6007,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5808978598144745e-05,
+      "loss": 3.5899,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5800592650634225e-05,
+      "loss": 3.5907,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5792206703123705e-05,
+      "loss": 3.5871,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5783837134416914e-05,
+      "loss": 3.6002,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5775451186906394e-05,
+      "loss": 3.5809,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5767065239395874e-05,
+      "loss": 3.6179,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5758679291885354e-05,
+      "loss": 3.597,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5750309723178563e-05,
+      "loss": 3.5761,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574192377566804e-05,
+      "loss": 3.6072,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573353782815752e-05,
+      "loss": 3.5852,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5725151880647e-05,
+      "loss": 3.5865,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571678231194022e-05,
+      "loss": 3.5959,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57083963644297e-05,
+      "loss": 3.5957,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570001041691918e-05,
+      "loss": 3.5899,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569162446940865e-05,
+      "loss": 3.5932,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568325490070187e-05,
+      "loss": 3.5848,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567486895319135e-05,
+      "loss": 3.5951,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566648300568083e-05,
+      "loss": 3.5989,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56580970581703e-05,
+      "loss": 3.5923,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564972748946352e-05,
+      "loss": 3.6008,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5641341541953e-05,
+      "loss": 3.5894,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563295559444248e-05,
+      "loss": 3.6158,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562456964693196e-05,
+      "loss": 3.5861,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561620007822517e-05,
+      "loss": 3.6039,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.560781413071465e-05,
+      "loss": 3.5899,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5599428183204126e-05,
+      "loss": 3.5895,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5591042235693606e-05,
+      "loss": 3.5946,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558267266698682e-05,
+      "loss": 3.5996,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55742867194763e-05,
+      "loss": 3.6018,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5565900771965775e-05,
+      "loss": 3.6011,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5557514824455255e-05,
+      "loss": 3.5776,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554914525574847e-05,
+      "loss": 3.5796,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554075930823795e-05,
+      "loss": 3.5931,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553237336072743e-05,
+      "loss": 3.6007,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.552398741321691e-05,
+      "loss": 3.5865,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5515617844510126e-05,
+      "loss": 3.5895,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55072318969996e-05,
+      "loss": 3.5924,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549884594948908e-05,
+      "loss": 3.5944,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549046000197856e-05,
+      "loss": 3.5932,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5482090433271775e-05,
+      "loss": 3.5915,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.547370448576125e-05,
+      "loss": 3.5935,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.546531853825073e-05,
+      "loss": 3.6003,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.545693259074021e-05,
+      "loss": 3.6017,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5448563022033424e-05,
+      "loss": 3.5813,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5440177074522904e-05,
+      "loss": 3.5868,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5431791127012384e-05,
+      "loss": 3.593,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54234215583056e-05,
+      "loss": 3.5961,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.541503561079507e-05,
+      "loss": 3.5997,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.540664966328455e-05,
+      "loss": 3.5928,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539826371577403e-05,
+      "loss": 3.5883,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538989414706725e-05,
+      "loss": 3.586,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538152457836046e-05,
+      "loss": 3.5894,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537313863084994e-05,
+      "loss": 3.588,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536475268333942e-05,
+      "loss": 3.5839,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53563667358289e-05,
+      "loss": 3.5961,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534798078831837e-05,
+      "loss": 3.588,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533961121961159e-05,
+      "loss": 3.5844,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5331225272101074e-05,
+      "loss": 3.5858,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532283932459055e-05,
+      "loss": 3.5788,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531445337708003e-05,
+      "loss": 3.6002,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.530606742956951e-05,
+      "loss": 3.5906,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.529768148205899e-05,
+      "loss": 3.5983,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528929553454847e-05,
+      "loss": 3.5938,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528090958703795e-05,
+      "loss": 3.6005,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5272540018331156e-05,
+      "loss": 3.5924,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5264154070820636e-05,
+      "loss": 3.5892,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5255768123310116e-05,
+      "loss": 3.5947,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5247398554603325e-05,
+      "loss": 3.5889,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523901260709281e-05,
+      "loss": 3.5855,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523064303838602e-05,
+      "loss": 3.5949,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52222570908755e-05,
+      "loss": 3.5851,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521387114336498e-05,
+      "loss": 3.5916,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520548519585446e-05,
+      "loss": 3.5881,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519709924834394e-05,
+      "loss": 3.5872,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518871330083342e-05,
+      "loss": 3.5923,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51803273533229e-05,
+      "loss": 3.5807,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517194140581238e-05,
+      "loss": 3.6059,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516357183710559e-05,
+      "loss": 3.5889,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515518588959507e-05,
+      "loss": 3.6,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.514679994208455e-05,
+      "loss": 3.5806,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.513841399457403e-05,
+      "loss": 3.5946,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5130044425867245e-05,
+      "loss": 3.5927,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5121658478356725e-05,
+      "loss": 3.5936,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5113272530846205e-05,
+      "loss": 3.5883,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5104886583335685e-05,
+      "loss": 3.5855,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5096517014628894e-05,
+      "loss": 3.5908,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5088131067118374e-05,
+      "loss": 3.6114,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5079745119607854e-05,
+      "loss": 3.5843,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507137555090106e-05,
+      "loss": 3.5996,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506298960339054e-05,
+      "loss": 3.5954,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505462003468375e-05,
+      "loss": 3.5855,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504623408717323e-05,
+      "loss": 3.595,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.503784813966271e-05,
+      "loss": 3.5903,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.50294621921522e-05,
+      "loss": 3.5902,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502107624464168e-05,
+      "loss": 3.591,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501269029713116e-05,
+      "loss": 3.5884,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.849526882171631,
+      "eval_runtime": 310.9633,
+      "eval_samples_per_second": 1227.125,
+      "eval_steps_per_second": 38.349,
+      "step": 915840
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 6.37493367108649e+17,
+  "trial_name": null,
+  "trial_params": null
+}