Create README.md

This is the text-based Sequence-to-Sequence ASR Error Correction Model pre-trained on Common Voice version 13.0 and continue-trained on TED transcriptions.

Details see our paper:

Please kindly cite

README.md

@@ -0,0 +1,11 @@
+---
+language:
+- en
+metrics:
+- wer
+- bleu
+tags:
+- ASR
+- Error Correction
+- Crossmodal
+---