add citation

README.md

@@ -1,5 +1,21 @@
 This repo contains the fully trained ByT5 that was used to estimate per-character entropies. Using it, you can also recreate the illustration in the paper.
 
+## Citation
+
+If you use this for research, please cite:
+```bibtex
+@misc{https://doi.org/10.48550/arxiv.2206.12693,
+  doi = {10.48550/ARXIV.2206.12693},
+  url = {https://arxiv.org/abs/2206.12693},
+  author = {Krabbenhöft, Hajo Nils and Barth, Erhardt},  
+  keywords = {Computation and Language (cs.CL), Sound (cs.SD), Audio and Speech Processing (eess.AS), FOS: Computer and information sciences, FOS: Computer and information sciences, FOS: Electrical engineering, electronic engineering, information engineering, FOS: Electrical engineering, electronic engineering, information engineering, F.2.1; I.2.6; I.2.7},  
+  title = {TEVR: Improving Speech Recognition by Token Entropy Variance Reduction},  
+  publisher = {arXiv},  
+  year = {2022}, 
+  copyright = {Creative Commons Attribution 4.0 International}
+}
+```
+
 ## Generate TEVR Tokenizer from Text corpus
 (copy of `Generate TEVR Tokenizer.ipynb`)