diff --git "a/checkpoint-457920/trainer_state.json" "b/checkpoint-457920/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-457920/trainer_state.json"
@@ -0,0 +1,5437 @@
+{
+  "best_metric": 3.8703999519348145,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/npi-only/transformer/2/checkpoints/checkpoint-457920",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 457920,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 11.0443,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8235,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1856,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9728,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8129,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7038,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.5988,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5319,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4434,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4027,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3513,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3122,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 5.2479,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.211,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1762,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1295,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1143,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0799,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0411,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0213,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983231380739706e-05,
+      "loss": 5.0037,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9739,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9551,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9232,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.917,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.8965,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8739,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8545,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8295,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756873037409844e-05,
+      "loss": 4.8209,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.8045,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7979,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7818,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7678,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7587,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7495,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7372,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.7059,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7073,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6806,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6867,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6658,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964790485618428e-05,
+      "loss": 4.6696,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963951890867376e-05,
+      "loss": 4.6485,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 4.6466,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.636,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.629,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6184,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.5929,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.6007,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.6025,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.5967,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.5745,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 4.5631,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5651,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5514,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5668,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 4.5275,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 4.5443,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 4.5337,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 4.5146,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 4.5243,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 4.5097,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 4.4994,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 4.494,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 4.5059,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 4.4841,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 4.4807,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 4.4712,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 4.4657,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 4.4734,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4796,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.454,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4656,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4677,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4514,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.4376,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.4421,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4377,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.4258,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.422,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.4303,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4305,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 4.4146,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.3932,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 4.4054,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 4.4099,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 4.397,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.4028,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925392911122715e-05,
+      "loss": 4.3841,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 4.3836,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 4.3774,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922878764749932e-05,
+      "loss": 4.3772,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 4.3816,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.3699,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.386,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.37,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.3526,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.3557,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.3626,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.3539,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.3516,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.3512,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.3396,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.3418,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.3208,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.3274,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.3316,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.3257,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.3213,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907787334991742e-05,
+      "loss": 4.3293,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90694874024069e-05,
+      "loss": 4.3046,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.3153,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3075,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.313,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.3215,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.3101,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.3035,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.2957,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.3031,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994063011223415e-05,
+      "loss": 4.2961,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985677063712895e-05,
+      "loss": 4.2945,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977291116202375e-05,
+      "loss": 4.2861,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968905168691855e-05,
+      "loss": 4.2845,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.2854,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.2746,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.2805,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.2776,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.2723,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918605862432463e-05,
+      "loss": 4.2648,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.2806,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.2829,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.2695,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.2591,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.2659,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.2668,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885992060866256e-05,
+      "loss": 4.265,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885153466115204e-05,
+      "loss": 4.2616,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884316509244525e-05,
+      "loss": 4.2623,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883477914493473e-05,
+      "loss": 4.2565,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.2544,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.2475,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.2501,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.2428,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.2379,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878447983867534e-05,
+      "loss": 4.2485,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.2335,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.244,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.232,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.2265,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.199948310852051,
+      "eval_runtime": 463.9943,
+      "eval_samples_per_second": 822.404,
+      "eval_steps_per_second": 25.701,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.2192,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.2179,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8725794584905426e-05,
+      "loss": 4.2312,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.2208,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.2245,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700636742373866e-05,
+      "loss": 4.2138,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692267173667075e-05,
+      "loss": 4.2107,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.868389760496029e-05,
+      "loss": 4.2019,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867551165744977e-05,
+      "loss": 4.2165,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866712570993925e-05,
+      "loss": 4.2092,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865873976242873e-05,
+      "loss": 4.2194,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865037019372194e-05,
+      "loss": 4.2023,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642000625015155e-05,
+      "loss": 4.1994,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633614677504635e-05,
+      "loss": 4.1934,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862522872999411e-05,
+      "loss": 4.1901,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861684278248359e-05,
+      "loss": 4.1874,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.1952,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860007088746255e-05,
+      "loss": 4.1826,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859168493995203e-05,
+      "loss": 4.1819,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.2017,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574929423734724e-05,
+      "loss": 4.1871,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566543476224204e-05,
+      "loss": 4.1891,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558157528713684e-05,
+      "loss": 4.1763,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549771581203164e-05,
+      "loss": 4.1897,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.1742,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532999686182124e-05,
+      "loss": 4.175,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524613738671604e-05,
+      "loss": 4.1738,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851624416996481e-05,
+      "loss": 4.1652,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.1607,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.1618,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.1691,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.1689,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847433081121595e-05,
+      "loss": 4.1619,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846594486370543e-05,
+      "loss": 4.1685,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845755891619491e-05,
+      "loss": 4.165,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844917296868439e-05,
+      "loss": 4.1635,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84408033999776e-05,
+      "loss": 4.1505,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.1532,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842403150495656e-05,
+      "loss": 4.1397,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841564555744604e-05,
+      "loss": 4.1531,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840725960993552e-05,
+      "loss": 4.1377,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839889004122873e-05,
+      "loss": 4.1512,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839050409371821e-05,
+      "loss": 4.1394,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382118146207687e-05,
+      "loss": 4.1407,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373732198697167e-05,
+      "loss": 4.143,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836536262999038e-05,
+      "loss": 4.139,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835697668247986e-05,
+      "loss": 4.135,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834859073496934e-05,
+      "loss": 4.1233,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834020478745882e-05,
+      "loss": 4.1282,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331818839948295e-05,
+      "loss": 4.1395,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832344927124151e-05,
+      "loss": 4.1349,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831506332373099e-05,
+      "loss": 4.1261,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830667737622047e-05,
+      "loss": 4.1165,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298291428709944e-05,
+      "loss": 4.1263,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828992186000316e-05,
+      "loss": 4.1137,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828153591249264e-05,
+      "loss": 4.1359,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827314996498212e-05,
+      "loss": 4.1081,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82647640174716e-05,
+      "loss": 4.1277,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825637806996108e-05,
+      "loss": 4.1204,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248008501254296e-05,
+      "loss": 4.1059,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823962255374377e-05,
+      "loss": 4.118,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823123660623325e-05,
+      "loss": 4.1124,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822285065872273e-05,
+      "loss": 4.1026,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214481090015945e-05,
+      "loss": 4.1043,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820609514250542e-05,
+      "loss": 4.1184,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81977091949949e-05,
+      "loss": 4.1057,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818932324748438e-05,
+      "loss": 4.1028,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818093729997386e-05,
+      "loss": 4.0981,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817255135246334e-05,
+      "loss": 4.0969,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164181783756554e-05,
+      "loss": 4.1054,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155795836246034e-05,
+      "loss": 4.1183,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147409888735514e-05,
+      "loss": 4.0989,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8139023941224994e-05,
+      "loss": 4.1126,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81306543725182e-05,
+      "loss": 4.1125,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812226842500768e-05,
+      "loss": 4.101,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811388247749716e-05,
+      "loss": 4.0918,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810549652998664e-05,
+      "loss": 4.1031,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809712696127985e-05,
+      "loss": 4.101,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808874101376933e-05,
+      "loss": 4.0917,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808035506625881e-05,
+      "loss": 4.0928,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807196911874829e-05,
+      "loss": 4.0972,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806359955004151e-05,
+      "loss": 4.1035,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805521360253099e-05,
+      "loss": 4.0927,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804682765502047e-05,
+      "loss": 4.0772,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803844170750995e-05,
+      "loss": 4.0854,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8030072138803156e-05,
+      "loss": 4.0909,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8021686191292636e-05,
+      "loss": 4.0871,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8013300243782116e-05,
+      "loss": 4.0913,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8004930675075325e-05,
+      "loss": 4.0803,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996544727564805e-05,
+      "loss": 4.0757,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988158780054285e-05,
+      "loss": 4.074,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979772832543765e-05,
+      "loss": 4.0797,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971386885033245e-05,
+      "loss": 4.0807,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7963000937522725e-05,
+      "loss": 4.077,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795463136881594e-05,
+      "loss": 4.0955,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794624542130542e-05,
+      "loss": 4.0819,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79378594737949e-05,
+      "loss": 4.0641,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792947352628438e-05,
+      "loss": 4.0737,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792108757877386e-05,
+      "loss": 4.0758,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791270163126334e-05,
+      "loss": 4.0761,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790431568375282e-05,
+      "loss": 4.073,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78959297362423e-05,
+      "loss": 4.0711,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788756016753551e-05,
+      "loss": 4.0637,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787917422002499e-05,
+      "loss": 4.0735,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787078827251447e-05,
+      "loss": 4.0482,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786240232500395e-05,
+      "loss": 4.0628,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854032756297166e-05,
+      "loss": 4.0625,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845646808786646e-05,
+      "loss": 4.0634,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837277240079855e-05,
+      "loss": 4.0535,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828891292569335e-05,
+      "loss": 4.066,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820505345058815e-05,
+      "loss": 4.0509,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812119397548295e-05,
+      "loss": 4.0555,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803733450037775e-05,
+      "loss": 4.0567,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779534750252725e-05,
+      "loss": 4.0564,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778696155501673e-05,
+      "loss": 4.0711,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777857560750621e-05,
+      "loss": 4.0615,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777022241760315e-05,
+      "loss": 4.0567,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776183647009263e-05,
+      "loss": 4.0504,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775345052258212e-05,
+      "loss": 4.0534,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77450645750716e-05,
+      "loss": 4.053,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773667862756108e-05,
+      "loss": 4.0485,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772829268005055e-05,
+      "loss": 4.0468,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771990673254003e-05,
+      "loss": 4.045,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771152078502951e-05,
+      "loss": 4.0494,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770315121632272e-05,
+      "loss": 4.0369,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76947652688122e-05,
+      "loss": 4.0451,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768637932130168e-05,
+      "loss": 4.0434,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76780097525949e-05,
+      "loss": 4.0422,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766962380508437e-05,
+      "loss": 4.0326,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766123785757386e-05,
+      "loss": 4.0539,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765285191006334e-05,
+      "loss": 4.0566,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764446596255282e-05,
+      "loss": 4.0415,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76360800150423e-05,
+      "loss": 4.0333,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762769406753178e-05,
+      "loss": 4.0421,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761930812002126e-05,
+      "loss": 4.0438,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610938551314466e-05,
+      "loss": 4.0453,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7602552603803946e-05,
+      "loss": 4.0403,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7594166656293426e-05,
+      "loss": 4.0422,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7585780708782906e-05,
+      "loss": 4.0361,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577411140076115e-05,
+      "loss": 4.0414,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569025192565595e-05,
+      "loss": 4.031,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560639245055075e-05,
+      "loss": 4.0382,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755226967634829e-05,
+      "loss": 4.0344,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754388372883777e-05,
+      "loss": 4.0226,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753549778132725e-05,
+      "loss": 4.0398,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752711183381673e-05,
+      "loss": 4.027,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751872588630621e-05,
+      "loss": 4.0354,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751033993879569e-05,
+      "loss": 4.0235,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750195399128517e-05,
+      "loss": 4.0215,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.029198169708252,
+      "eval_runtime": 309.4125,
+      "eval_samples_per_second": 1233.276,
+      "eval_steps_per_second": 38.541,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749356804377465e-05,
+      "loss": 4.0196,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748518209626413e-05,
+      "loss": 4.0113,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747679614875361e-05,
+      "loss": 4.0325,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468410201243084e-05,
+      "loss": 4.022,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460024253732563e-05,
+      "loss": 4.0294,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451638306222043e-05,
+      "loss": 4.0134,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744325235871152e-05,
+      "loss": 4.0174,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434866411201e-05,
+      "loss": 4.0037,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742648046369049e-05,
+      "loss": 4.0199,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741809451617997e-05,
+      "loss": 4.0229,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740970856866945e-05,
+      "loss": 4.0234,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740133899996266e-05,
+      "loss": 4.0143,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739296943125587e-05,
+      "loss": 4.0081,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738458348374535e-05,
+      "loss": 4.0067,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737619753623483e-05,
+      "loss": 4.0021,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736781158872431e-05,
+      "loss": 4.0024,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735942564121379e-05,
+      "loss": 4.0071,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735103969370327e-05,
+      "loss": 3.9982,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734265374619275e-05,
+      "loss": 4.0008,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733426779868223e-05,
+      "loss": 4.0206,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732588185117171e-05,
+      "loss": 4.0064,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731749590366119e-05,
+      "loss": 4.0109,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7309109956150674e-05,
+      "loss": 3.9984,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300740387443884e-05,
+      "loss": 4.0136,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729237081873709e-05,
+      "loss": 3.9895,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728398487122657e-05,
+      "loss": 4.0029,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727559892371605e-05,
+      "loss": 3.9986,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726721297620553e-05,
+      "loss": 3.9959,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725882702869501e-05,
+      "loss": 3.983,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725044108118449e-05,
+      "loss": 3.99,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724205513367397e-05,
+      "loss": 3.9978,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233669186163446e-05,
+      "loss": 4.0001,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72253159962604e-05,
+      "loss": 3.9956,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721693004874988e-05,
+      "loss": 3.9961,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720854410123936e-05,
+      "loss": 3.9948,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720015815372884e-05,
+      "loss": 3.9958,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719177220621832e-05,
+      "loss": 3.9861,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71833862587078e-05,
+      "loss": 3.9878,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717500031119727e-05,
+      "loss": 3.9763,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7166630742490486e-05,
+      "loss": 3.9918,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158244794979966e-05,
+      "loss": 3.978,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149858847469446e-05,
+      "loss": 3.9907,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714147289995892e-05,
+      "loss": 3.9815,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133103331252135e-05,
+      "loss": 3.9768,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124717383741615e-05,
+      "loss": 3.9899,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116331436231095e-05,
+      "loss": 3.9768,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7107945488720575e-05,
+      "loss": 3.9797,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709957592001379e-05,
+      "loss": 3.9627,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709118997250327e-05,
+      "loss": 3.9755,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082804024992744e-05,
+      "loss": 3.9828,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074418077482224e-05,
+      "loss": 3.9805,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706606488757917e-05,
+      "loss": 3.977,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705767894006865e-05,
+      "loss": 3.9612,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704929299255813e-05,
+      "loss": 3.9749,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704090704504761e-05,
+      "loss": 3.9611,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703252109753709e-05,
+      "loss": 3.9834,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702413515002657e-05,
+      "loss": 3.9602,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701574920251605e-05,
+      "loss": 3.9773,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700736325500553e-05,
+      "loss": 3.9753,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6999010065102474e-05,
+      "loss": 3.9545,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990624117591954e-05,
+      "loss": 3.965,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982238170081434e-05,
+      "loss": 3.9697,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973852222570914e-05,
+      "loss": 3.9564,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965466275060394e-05,
+      "loss": 3.9596,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69570967063536e-05,
+      "loss": 3.9709,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694871075884308e-05,
+      "loss": 3.9665,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694032481133256e-05,
+      "loss": 3.9543,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693193886382204e-05,
+      "loss": 3.9588,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692355291631152e-05,
+      "loss": 3.9512,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691518334760474e-05,
+      "loss": 3.9627,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690679740009422e-05,
+      "loss": 3.9766,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689841145258369e-05,
+      "loss": 3.9604,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689002550507317e-05,
+      "loss": 3.9735,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688163955756265e-05,
+      "loss": 3.9698,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687325361005213e-05,
+      "loss": 3.9643,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686488404134534e-05,
+      "loss": 3.9506,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685649809383482e-05,
+      "loss": 3.9689,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68481121463243e-05,
+      "loss": 3.9609,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683972619881378e-05,
+      "loss": 3.9582,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683134025130326e-05,
+      "loss": 3.9554,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682295430379274e-05,
+      "loss": 3.9613,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681456835628222e-05,
+      "loss": 3.9726,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806198787575436e-05,
+      "loss": 3.9557,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6797812840064916e-05,
+      "loss": 3.942,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789426892554396e-05,
+      "loss": 3.9487,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6781040945043876e-05,
+      "loss": 3.9584,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772654997533356e-05,
+      "loss": 3.9573,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764269050022836e-05,
+      "loss": 3.9596,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755899481316045e-05,
+      "loss": 3.9447,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747513533805525e-05,
+      "loss": 3.9496,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739127586295005e-05,
+      "loss": 3.9385,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730741638784485e-05,
+      "loss": 3.9477,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722372070077694e-05,
+      "loss": 3.9522,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6713986122567174e-05,
+      "loss": 3.9515,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670560017505666e-05,
+      "loss": 3.9702,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669721422754614e-05,
+      "loss": 3.9484,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668882828003562e-05,
+      "loss": 3.9403,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66804423325251e-05,
+      "loss": 3.946,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667205638501458e-05,
+      "loss": 3.9463,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663670437504054e-05,
+      "loss": 3.9536,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665530086879727e-05,
+      "loss": 3.9483,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664693130009048e-05,
+      "loss": 3.9423,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663854535257996e-05,
+      "loss": 3.9412,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663015940506944e-05,
+      "loss": 3.9492,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662177345755892e-05,
+      "loss": 3.9247,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66133875100484e-05,
+      "loss": 3.9407,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6605017941341614e-05,
+      "loss": 3.9379,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596631993831094e-05,
+      "loss": 3.9416,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588246046320574e-05,
+      "loss": 3.9333,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579860098810054e-05,
+      "loss": 3.9434,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657147415129953e-05,
+      "loss": 3.9305,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656308820378901e-05,
+      "loss": 3.9321,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655471863508222e-05,
+      "loss": 3.9412,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65463326875717e-05,
+      "loss": 3.9329,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6537946740061176e-05,
+      "loss": 3.9519,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529560792550656e-05,
+      "loss": 3.944,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652119122384387e-05,
+      "loss": 3.9384,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651280527633335e-05,
+      "loss": 3.9302,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650441932882283e-05,
+      "loss": 3.9335,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649603338131231e-05,
+      "loss": 3.9361,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648766381260553e-05,
+      "loss": 3.9341,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479277865095e-05,
+      "loss": 3.931,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647089191758448e-05,
+      "loss": 3.9238,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646250597007396e-05,
+      "loss": 3.9347,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645412002256344e-05,
+      "loss": 3.9207,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644575045385665e-05,
+      "loss": 3.9298,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643736450634613e-05,
+      "loss": 3.9279,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642897855883561e-05,
+      "loss": 3.9261,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642059261132509e-05,
+      "loss": 3.9176,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412223042618306e-05,
+      "loss": 3.9417,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403837095107786e-05,
+      "loss": 3.9424,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395451147597266e-05,
+      "loss": 3.9329,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387065200086746e-05,
+      "loss": 3.9147,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378679252576226e-05,
+      "loss": 3.9312,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370309683869435e-05,
+      "loss": 3.9286,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361923736358915e-05,
+      "loss": 3.9354,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353537788848395e-05,
+      "loss": 3.928,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345151841337874e-05,
+      "loss": 3.9269,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336765893827354e-05,
+      "loss": 3.9252,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328396325120564e-05,
+      "loss": 3.9319,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6320010377610043e-05,
+      "loss": 3.9246,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311624430099523e-05,
+      "loss": 3.9261,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303238482589003e-05,
+      "loss": 3.9257,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629486891388222e-05,
+      "loss": 3.9133,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62864829663717e-05,
+      "loss": 3.9289,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627809701886118e-05,
+      "loss": 3.9251,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626971107135066e-05,
+      "loss": 3.9205,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626132512384014e-05,
+      "loss": 3.916,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625293917632962e-05,
+      "loss": 3.9133,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.954420804977417,
+      "eval_runtime": 311.7206,
+      "eval_samples_per_second": 1224.144,
+      "eval_steps_per_second": 38.255,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624456960762283e-05,
+      "loss": 3.9169,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623618366011231e-05,
+      "loss": 3.9074,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622779771260179e-05,
+      "loss": 3.9231,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621941176509127e-05,
+      "loss": 3.9158,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621104219638448e-05,
+      "loss": 3.926,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620265624887396e-05,
+      "loss": 3.908,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619427030136344e-05,
+      "loss": 3.9129,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185884353852924e-05,
+      "loss": 3.8996,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177498406342404e-05,
+      "loss": 3.9162,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616912883763561e-05,
+      "loss": 3.9163,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616074289012509e-05,
+      "loss": 3.9183,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615235694261457e-05,
+      "loss": 3.9102,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614398737390778e-05,
+      "loss": 3.9063,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613560142639726e-05,
+      "loss": 3.9061,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612721547888674e-05,
+      "loss": 3.8998,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611882953137622e-05,
+      "loss": 3.9022,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61104435838657e-05,
+      "loss": 3.9066,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610205763635518e-05,
+      "loss": 3.8993,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609367168884466e-05,
+      "loss": 3.8994,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608528574133414e-05,
+      "loss": 3.9247,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607689979382362e-05,
+      "loss": 3.9064,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606853022511684e-05,
+      "loss": 3.9069,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606014427760631e-05,
+      "loss": 3.9029,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605175833009579e-05,
+      "loss": 3.9092,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604337238258527e-05,
+      "loss": 3.8936,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603498643507475e-05,
+      "loss": 3.9097,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602661686636796e-05,
+      "loss": 3.8984,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601823091885744e-05,
+      "loss": 3.8963,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600984497134692e-05,
+      "loss": 3.8863,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.60014590238364e-05,
+      "loss": 3.8949,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5993089455129615e-05,
+      "loss": 3.8994,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5984703507619095e-05,
+      "loss": 3.9043,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976317560108575e-05,
+      "loss": 3.9037,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967931612598055e-05,
+      "loss": 3.9006,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959562043891264e-05,
+      "loss": 3.8936,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5951176096380744e-05,
+      "loss": 3.9039,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5942790148870224e-05,
+      "loss": 3.8883,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5934404201359704e-05,
+      "loss": 3.897,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592603463265291e-05,
+      "loss": 3.8809,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591764868514239e-05,
+      "loss": 3.8958,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590926273763187e-05,
+      "loss": 3.8847,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590087679012135e-05,
+      "loss": 3.8971,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589250722141457e-05,
+      "loss": 3.8907,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588412127390405e-05,
+      "loss": 3.8826,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587573532639353e-05,
+      "loss": 3.8965,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586734937888301e-05,
+      "loss": 3.8863,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585896343137249e-05,
+      "loss": 3.8853,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58505938626657e-05,
+      "loss": 3.873,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584220791515518e-05,
+      "loss": 3.8815,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583382196764466e-05,
+      "loss": 3.889,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582543602013414e-05,
+      "loss": 3.8891,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581706645142735e-05,
+      "loss": 3.8853,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580868050391683e-05,
+      "loss": 3.8714,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580029455640631e-05,
+      "loss": 3.8843,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579190860889579e-05,
+      "loss": 3.8708,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783539040189e-05,
+      "loss": 3.8899,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577515309267848e-05,
+      "loss": 3.8774,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576676714516796e-05,
+      "loss": 3.8844,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575838119765744e-05,
+      "loss": 3.8878,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575001162895065e-05,
+      "loss": 3.8705,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574162568144013e-05,
+      "loss": 3.874,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573323973392961e-05,
+      "loss": 3.8831,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572485378641909e-05,
+      "loss": 3.8678,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57164842177123e-05,
+      "loss": 3.8761,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570809827020178e-05,
+      "loss": 3.8792,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569971232269126e-05,
+      "loss": 3.8822,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569132637518074e-05,
+      "loss": 3.8671,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5682956806473956e-05,
+      "loss": 3.8738,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674570858963436e-05,
+      "loss": 3.8636,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5666184911452916e-05,
+      "loss": 3.8745,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657798963942396e-05,
+      "loss": 3.8863,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649413016431876e-05,
+      "loss": 3.8766,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5641043447725085e-05,
+      "loss": 3.8864,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632657500214565e-05,
+      "loss": 3.8844,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624271552704045e-05,
+      "loss": 3.8838,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615885605193525e-05,
+      "loss": 3.8622,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5607516036486734e-05,
+      "loss": 3.8817,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5599130088976214e-05,
+      "loss": 3.8809,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5590744141465694e-05,
+      "loss": 3.8693,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582358193955174e-05,
+      "loss": 3.8749,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557398862524839e-05,
+      "loss": 3.8765,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556560267773787e-05,
+      "loss": 3.8897,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555721673022735e-05,
+      "loss": 3.8724,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554883078271683e-05,
+      "loss": 3.8575,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554046121401004e-05,
+      "loss": 3.8694,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553207526649952e-05,
+      "loss": 3.8721,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523689318989e-05,
+      "loss": 3.8734,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551530337147848e-05,
+      "loss": 3.8749,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550693380277169e-05,
+      "loss": 3.8666,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549854785526117e-05,
+      "loss": 3.8689,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549016190775065e-05,
+      "loss": 3.8575,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548177596024013e-05,
+      "loss": 3.8635,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547339001272961e-05,
+      "loss": 3.8669,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5465020444022823e-05,
+      "loss": 3.876,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54566344965123e-05,
+      "loss": 3.8902,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544824854900178e-05,
+      "loss": 3.8655,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543986260149126e-05,
+      "loss": 3.856,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543149303278447e-05,
+      "loss": 3.8669,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542310708527395e-05,
+      "loss": 3.8646,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541472113776343e-05,
+      "loss": 3.8746,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540633519025291e-05,
+      "loss": 3.865,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539796562154612e-05,
+      "loss": 3.8642,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53895796740356e-05,
+      "loss": 3.8622,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538119372652508e-05,
+      "loss": 3.8717,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537280777901456e-05,
+      "loss": 3.8461,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536443821030778e-05,
+      "loss": 3.8607,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535605226279726e-05,
+      "loss": 3.8549,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534766631528674e-05,
+      "loss": 3.8676,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533928036777621e-05,
+      "loss": 3.8522,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533089442026569e-05,
+      "loss": 3.8623,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5322524851558906e-05,
+      "loss": 3.852,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314138904048386e-05,
+      "loss": 3.8564,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530575295653786e-05,
+      "loss": 3.8614,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529736700902734e-05,
+      "loss": 3.8594,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5288997440320555e-05,
+      "loss": 3.8708,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5280611492810035e-05,
+      "loss": 3.8686,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527224192410325e-05,
+      "loss": 3.8609,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526385597659273e-05,
+      "loss": 3.8542,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525547002908221e-05,
+      "loss": 3.8559,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5247084081571684e-05,
+      "loss": 3.861,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238698134061164e-05,
+      "loss": 3.8521,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230312186550644e-05,
+      "loss": 3.8566,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221926239040124e-05,
+      "loss": 3.8544,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213540291529604e-05,
+      "loss": 3.8537,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520517072282281e-05,
+      "loss": 3.8497,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519678477531229e-05,
+      "loss": 3.8529,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518839882780177e-05,
+      "loss": 3.8507,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518001288029125e-05,
+      "loss": 3.8546,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517164331158447e-05,
+      "loss": 3.8407,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516325736407395e-05,
+      "loss": 3.868,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515487141656343e-05,
+      "loss": 3.8632,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514648546905291e-05,
+      "loss": 3.8642,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513811590034612e-05,
+      "loss": 3.8383,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51297299528356e-05,
+      "loss": 3.8569,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512134400532508e-05,
+      "loss": 3.8586,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511295805781456e-05,
+      "loss": 3.8574,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104588489107766e-05,
+      "loss": 3.8562,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096202541597246e-05,
+      "loss": 3.8532,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087816594086726e-05,
+      "loss": 3.8527,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079430646576206e-05,
+      "loss": 3.8597,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507104469906569e-05,
+      "loss": 3.8518,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50626751303589e-05,
+      "loss": 3.8522,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505428918284838e-05,
+      "loss": 3.8523,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504590323533786e-05,
+      "loss": 3.8403,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503751728782734e-05,
+      "loss": 3.8584,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502914771912055e-05,
+      "loss": 3.8539,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502076177161003e-05,
+      "loss": 3.851,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501237582409951e-05,
+      "loss": 3.8413,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500398987658899e-05,
+      "loss": 3.8411,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9124677181243896,
+      "eval_runtime": 305.0369,
+      "eval_samples_per_second": 1250.967,
+      "eval_steps_per_second": 39.094,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499560392907847e-05,
+      "loss": 3.843,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498721798156795e-05,
+      "loss": 3.8357,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497883203405743e-05,
+      "loss": 3.8483,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497044608654691e-05,
+      "loss": 3.846,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496206013903639e-05,
+      "loss": 3.8601,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495367419152587e-05,
+      "loss": 3.8359,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494528824401535e-05,
+      "loss": 3.8404,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493690229650483e-05,
+      "loss": 3.828,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928532727798047e-05,
+      "loss": 3.848,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492014678028752e-05,
+      "loss": 3.8467,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911760832777e-05,
+      "loss": 3.8435,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490337488526648e-05,
+      "loss": 3.8439,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489500531655969e-05,
+      "loss": 3.8329,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488661936904917e-05,
+      "loss": 3.8413,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487823342153865e-05,
+      "loss": 3.8267,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486984747402813e-05,
+      "loss": 3.8364,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486146152651761e-05,
+      "loss": 3.8344,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4853075579007095e-05,
+      "loss": 3.8318,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844689631496575e-05,
+      "loss": 3.8355,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4836303683986055e-05,
+      "loss": 3.8526,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4827934115279264e-05,
+      "loss": 3.8335,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4819548167768744e-05,
+      "loss": 3.8399,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4811162220258224e-05,
+      "loss": 3.8331,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4802776272747704e-05,
+      "loss": 3.8419,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479440670404091e-05,
+      "loss": 3.8287,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478602075653039e-05,
+      "loss": 3.8415,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477763480901987e-05,
+      "loss": 3.8311,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476924886150935e-05,
+      "loss": 3.8287,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476087929280256e-05,
+      "loss": 3.8214,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475249334529205e-05,
+      "loss": 3.826,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474410739778153e-05,
+      "loss": 3.8311,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473572145027101e-05,
+      "loss": 3.8372,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472735188156422e-05,
+      "loss": 3.8376,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47189659340537e-05,
+      "loss": 3.8332,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471057998654318e-05,
+      "loss": 3.8277,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470219403903266e-05,
+      "loss": 3.8387,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469382447032587e-05,
+      "loss": 3.8292,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468543852281535e-05,
+      "loss": 3.8282,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467705257530483e-05,
+      "loss": 3.8131,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466866662779431e-05,
+      "loss": 3.8272,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660297059087516e-05,
+      "loss": 3.8225,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651911111577e-05,
+      "loss": 3.8341,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464352516406648e-05,
+      "loss": 3.8263,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463513921655596e-05,
+      "loss": 3.8163,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462676964784917e-05,
+      "loss": 3.834,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461838370033865e-05,
+      "loss": 3.8237,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460999775282813e-05,
+      "loss": 3.8196,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460161180531761e-05,
+      "loss": 3.8111,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459324223661082e-05,
+      "loss": 3.8131,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45848562891003e-05,
+      "loss": 3.8237,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457647034158978e-05,
+      "loss": 3.8246,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456808439407926e-05,
+      "loss": 3.8197,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455971482537247e-05,
+      "loss": 3.8123,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551328877861956e-05,
+      "loss": 3.8241,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542942930351436e-05,
+      "loss": 3.8072,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534556982840916e-05,
+      "loss": 3.8236,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4526187414134125e-05,
+      "loss": 3.8109,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517801466623605e-05,
+      "loss": 3.8236,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509415519113085e-05,
+      "loss": 3.82,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4501029571602565e-05,
+      "loss": 3.8146,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492660002895774e-05,
+      "loss": 3.8098,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4484274055385254e-05,
+      "loss": 3.8207,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475888107874734e-05,
+      "loss": 3.8054,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467502160364214e-05,
+      "loss": 3.8128,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445913259165742e-05,
+      "loss": 3.8204,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445074664414691e-05,
+      "loss": 3.82,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444236069663639e-05,
+      "loss": 3.807,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443397474912587e-05,
+      "loss": 3.8091,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442560518041908e-05,
+      "loss": 3.8068,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441721923290856e-05,
+      "loss": 3.8083,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440883328539804e-05,
+      "loss": 3.8227,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440044733788752e-05,
+      "loss": 3.8166,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439207776918073e-05,
+      "loss": 3.8276,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438369182167021e-05,
+      "loss": 3.8237,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437530587415969e-05,
+      "loss": 3.8241,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436691992664917e-05,
+      "loss": 3.8058,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435855035794238e-05,
+      "loss": 3.8157,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4350164410431864e-05,
+      "loss": 3.824,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4341778462921344e-05,
+      "loss": 3.8088,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333392515410824e-05,
+      "loss": 3.812,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432502294670403e-05,
+      "loss": 3.8151,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431663699919351e-05,
+      "loss": 3.8301,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430825105168299e-05,
+      "loss": 3.8119,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429986510417247e-05,
+      "loss": 3.7964,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429149553546568e-05,
+      "loss": 3.8089,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428310958795516e-05,
+      "loss": 3.8126,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427472364044464e-05,
+      "loss": 3.8161,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426633769293412e-05,
+      "loss": 3.8147,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425796812422733e-05,
+      "loss": 3.8048,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424958217671682e-05,
+      "loss": 3.807,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42411962292063e-05,
+      "loss": 3.8056,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423281028169577e-05,
+      "loss": 3.8009,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224440712988986e-05,
+      "loss": 3.8099,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216054765478466e-05,
+      "loss": 3.8151,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207668817967946e-05,
+      "loss": 3.8293,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419928287045742e-05,
+      "loss": 3.8146,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190913301750635e-05,
+      "loss": 3.7945,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182527354240115e-05,
+      "loss": 3.808,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4174141406729595e-05,
+      "loss": 3.8034,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416575545921907e-05,
+      "loss": 3.8151,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157385890512284e-05,
+      "loss": 3.8113,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414899994300177e-05,
+      "loss": 3.8067,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4140613995491244e-05,
+      "loss": 3.8026,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4132228047980724e-05,
+      "loss": 3.8117,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412385847927394e-05,
+      "loss": 3.791,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411547253176342e-05,
+      "loss": 3.8048,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410708658425289e-05,
+      "loss": 3.795,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409871701554611e-05,
+      "loss": 3.8121,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409033106803559e-05,
+      "loss": 3.7979,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408194512052507e-05,
+      "loss": 3.8022,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407355917301454e-05,
+      "loss": 3.7945,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406518960430776e-05,
+      "loss": 3.8018,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405680365679724e-05,
+      "loss": 3.8047,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404841770928672e-05,
+      "loss": 3.8022,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40400317617762e-05,
+      "loss": 3.8135,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031662193069414e-05,
+      "loss": 3.8133,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402329262436262e-05,
+      "loss": 3.8039,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40149066768521e-05,
+      "loss": 3.7981,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400652072934158e-05,
+      "loss": 3.7988,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399813478183106e-05,
+      "loss": 3.8082,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398974883432054e-05,
+      "loss": 3.7903,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981362886810016e-05,
+      "loss": 3.8046,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3972976939299496e-05,
+      "loss": 3.7922,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3964590991788976e-05,
+      "loss": 3.8006,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395622142308219e-05,
+      "loss": 3.7965,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394783547557167e-05,
+      "loss": 3.7965,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393944952806115e-05,
+      "loss": 3.7899,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393106358055063e-05,
+      "loss": 3.8024,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392269401184384e-05,
+      "loss": 3.7882,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391430806433332e-05,
+      "loss": 3.8116,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39059221168228e-05,
+      "loss": 3.804,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389753616931228e-05,
+      "loss": 3.8133,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388916660060549e-05,
+      "loss": 3.7804,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388078065309497e-05,
+      "loss": 3.8036,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387239470558445e-05,
+      "loss": 3.803,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386400875807393e-05,
+      "loss": 3.8023,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855639189367145e-05,
+      "loss": 3.8013,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3847253241856625e-05,
+      "loss": 3.8021,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838867294346105e-05,
+      "loss": 3.797,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830481346835585e-05,
+      "loss": 3.8055,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3822111778128794e-05,
+      "loss": 3.7907,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813725830618274e-05,
+      "loss": 3.7978,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805339883107754e-05,
+      "loss": 3.7962,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796953935597234e-05,
+      "loss": 3.7883,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378858436689044e-05,
+      "loss": 3.7984,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378019841937992e-05,
+      "loss": 3.8039,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37718124718694e-05,
+      "loss": 3.7957,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376342652435888e-05,
+      "loss": 3.7828,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37550569556521e-05,
+      "loss": 3.7911,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8867132663726807,
+      "eval_runtime": 304.4412,
+      "eval_samples_per_second": 1253.414,
+      "eval_steps_per_second": 39.17,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374667100814158e-05,
+      "loss": 3.7984,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373828506063106e-05,
+      "loss": 3.7819,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372989911312054e-05,
+      "loss": 3.7941,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372152954441375e-05,
+      "loss": 3.7971,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371314359690323e-05,
+      "loss": 3.8031,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370475764939271e-05,
+      "loss": 3.788,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369637170188219e-05,
+      "loss": 3.7857,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36880021331754e-05,
+      "loss": 3.775,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367961618566488e-05,
+      "loss": 3.7925,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367123023815436e-05,
+      "loss": 3.7967,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366284429064384e-05,
+      "loss": 3.7885,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365447472193705e-05,
+      "loss": 3.7922,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364610515323026e-05,
+      "loss": 3.7811,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363771920571974e-05,
+      "loss": 3.7889,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362933325820922e-05,
+      "loss": 3.777,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36209473106987e-05,
+      "loss": 3.7823,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361256136318818e-05,
+      "loss": 3.7784,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360417541567766e-05,
+      "loss": 3.7857,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359578946816714e-05,
+      "loss": 3.7793,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358740352065662e-05,
+      "loss": 3.7996,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357903395194983e-05,
+      "loss": 3.786,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357064800443931e-05,
+      "loss": 3.7946,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356226205692879e-05,
+      "loss": 3.7799,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355387610941827e-05,
+      "loss": 3.7897,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3545506540711486e-05,
+      "loss": 3.7754,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3537120593200966e-05,
+      "loss": 3.79,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3528734645690446e-05,
+      "loss": 3.7825,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3520348698179926e-05,
+      "loss": 3.7721,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3511979129473135e-05,
+      "loss": 3.7709,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3503593181962615e-05,
+      "loss": 3.7738,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3495207234452095e-05,
+      "loss": 3.7807,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486837665745304e-05,
+      "loss": 3.7842,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478451718234784e-05,
+      "loss": 3.7866,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3470065770724264e-05,
+      "loss": 3.7859,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461679823213744e-05,
+      "loss": 3.7768,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345331025450696e-05,
+      "loss": 3.7867,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344492430699644e-05,
+      "loss": 3.7793,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343653835948592e-05,
+      "loss": 3.7781,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34281524119754e-05,
+      "loss": 3.7625,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341978284326861e-05,
+      "loss": 3.7777,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341139689575809e-05,
+      "loss": 3.7723,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340301094824757e-05,
+      "loss": 3.7847,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339462500073705e-05,
+      "loss": 3.777,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338625543203026e-05,
+      "loss": 3.7685,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337786948451974e-05,
+      "loss": 3.7847,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336948353700922e-05,
+      "loss": 3.7698,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33610975894987e-05,
+      "loss": 3.7723,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3352728020791914e-05,
+      "loss": 3.7638,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3344342073281393e-05,
+      "loss": 3.7625,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3335956125770873e-05,
+      "loss": 3.7752,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3327570178260353e-05,
+      "loss": 3.7763,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331920060955356e-05,
+      "loss": 3.7685,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331081466204304e-05,
+      "loss": 3.7648,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330244509333625e-05,
+      "loss": 3.7757,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329405914582573e-05,
+      "loss": 3.7546,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328567319831521e-05,
+      "loss": 3.7773,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327728725080469e-05,
+      "loss": 3.7634,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326890130329417e-05,
+      "loss": 3.7728,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326051535578365e-05,
+      "loss": 3.7704,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325212940827313e-05,
+      "loss": 3.7669,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324374346076261e-05,
+      "loss": 3.7609,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323537389205583e-05,
+      "loss": 3.7703,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322698794454531e-05,
+      "loss": 3.7634,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321860199703479e-05,
+      "loss": 3.7612,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321021604952427e-05,
+      "loss": 3.7711,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201846480817476e-05,
+      "loss": 3.7702,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193460533306956e-05,
+      "loss": 3.7576,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3185074585796436e-05,
+      "loss": 3.7664,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176688638285916e-05,
+      "loss": 3.755,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3168319069579125e-05,
+      "loss": 3.7602,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3159933122068605e-05,
+      "loss": 3.7761,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151547174558085e-05,
+      "loss": 3.7713,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3143161227047565e-05,
+      "loss": 3.7756,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313479165834078e-05,
+      "loss": 3.7724,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312640571083026e-05,
+      "loss": 3.7815,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311801976331974e-05,
+      "loss": 3.761,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3109633815809214e-05,
+      "loss": 3.7653,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310126424710243e-05,
+      "loss": 3.7756,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309287829959191e-05,
+      "loss": 3.7585,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308449235208139e-05,
+      "loss": 3.7686,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307610640457086e-05,
+      "loss": 3.7714,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306773683586408e-05,
+      "loss": 3.7768,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305935088835356e-05,
+      "loss": 3.7649,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305096494084304e-05,
+      "loss": 3.7527,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304257899333252e-05,
+      "loss": 3.7605,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3034209424625734e-05,
+      "loss": 3.7648,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025823477115214e-05,
+      "loss": 3.7688,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301743752960469e-05,
+      "loss": 3.7677,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300905158209417e-05,
+      "loss": 3.762,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000682013387383e-05,
+      "loss": 3.7645,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299229606587686e-05,
+      "loss": 3.7549,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983910118366337e-05,
+      "loss": 3.7536,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975524170855817e-05,
+      "loss": 3.7606,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296715460214903e-05,
+      "loss": 3.7701,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295876865463851e-05,
+      "loss": 3.7849,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295038270712799e-05,
+      "loss": 3.7693,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294199675961747e-05,
+      "loss": 3.7474,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293362719091069e-05,
+      "loss": 3.7655,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292524124340016e-05,
+      "loss": 3.7578,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291685529588964e-05,
+      "loss": 3.7696,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290846934837912e-05,
+      "loss": 3.761,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290009977967234e-05,
+      "loss": 3.7614,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289171383216181e-05,
+      "loss": 3.7587,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288332788465129e-05,
+      "loss": 3.7629,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287494193714077e-05,
+      "loss": 3.7507,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286657236843398e-05,
+      "loss": 3.7593,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2858186420923466e-05,
+      "loss": 3.7475,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2849800473412946e-05,
+      "loss": 3.7698,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284143090470616e-05,
+      "loss": 3.7515,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2833044957195635e-05,
+      "loss": 3.755,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2824659009685115e-05,
+      "loss": 3.7464,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2816273062174595e-05,
+      "loss": 3.7609,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807887114664075e-05,
+      "loss": 3.759,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2799501167153555e-05,
+      "loss": 3.7571,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2791115219643035e-05,
+      "loss": 3.7686,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782729272132515e-05,
+      "loss": 3.7676,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277437608222945e-05,
+      "loss": 3.762,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276599013471893e-05,
+      "loss": 3.7562,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275760418720842e-05,
+      "loss": 3.7509,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27492182396979e-05,
+      "loss": 3.7645,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274083229218738e-05,
+      "loss": 3.7486,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273244634467686e-05,
+      "loss": 3.7588,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272406039716634e-05,
+      "loss": 3.7454,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271567444965582e-05,
+      "loss": 3.76,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270730488094903e-05,
+      "loss": 3.746,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269891893343851e-05,
+      "loss": 3.7544,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269053298592799e-05,
+      "loss": 3.7509,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268214703841747e-05,
+      "loss": 3.7531,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267377746971068e-05,
+      "loss": 3.7483,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266539152220016e-05,
+      "loss": 3.7645,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265700557468964e-05,
+      "loss": 3.7624,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264861962717912e-05,
+      "loss": 3.7653,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264025005847233e-05,
+      "loss": 3.7393,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263186411096181e-05,
+      "loss": 3.7606,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262347816345129e-05,
+      "loss": 3.7574,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261509221594077e-05,
+      "loss": 3.7605,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260672264723398e-05,
+      "loss": 3.7548,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259833669972346e-05,
+      "loss": 3.7594,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258995075221294e-05,
+      "loss": 3.7537,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258156480470242e-05,
+      "loss": 3.7638,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257319523599563e-05,
+      "loss": 3.7463,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256480928848511e-05,
+      "loss": 3.7586,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255642334097459e-05,
+      "loss": 3.7529,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254803739346407e-05,
+      "loss": 3.7483,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253966782475729e-05,
+      "loss": 3.7554,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253128187724677e-05,
+      "loss": 3.7549,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252289592973625e-05,
+      "loss": 3.7585,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251450998222573e-05,
+      "loss": 3.7376,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2506140413518936e-05,
+      "loss": 3.7507,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8703999519348145,
+      "eval_runtime": 303.1842,
+      "eval_samples_per_second": 1258.611,
+      "eval_steps_per_second": 39.333,
+      "step": 457920
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 3.1932166101728256e+17,
+  "trial_name": null,
+  "trial_params": null
+}