Add pipeline tag, link to paper and code

This PR adds the `audio-to-audio` pipeline tag to the model card, ensuring people can find the model at https://huggingface.co/models?pipeline_tag=audio-to-audio.
It also includes a link to the paper and to the Github repository.

README.md

@@ -1,3 +1,8 @@
 ---
 license: apache-2.0
+pipeline_tag: audio-to-audio
 ---
+
+This repository contains the model for [Universal Source Separation with Weakly Labelled Data](https://huggingface.co/papers/2305.07447).
+
+Code: https://github.com/bytedance/uss