Create README.md

# Audio Dataset
This dataset consists of audio data for the following categories:
* Coughing
* Toilet flushing
* Running water
* Other sounds

The dataset is slightly unbalanced which can be managed by data augmentation for these `.mp3` clips. The file structure follows:
\- audio/
      - coughing/
            \- *.mp3
      \- toilet_flush/
            \- *.mp3
      \- running_water/
            \- *.mp3
      \- other_1/
            \- *.mp3
      \- other_2/
            \- *.mp3

README.md

@@ -0,0 +1,10 @@
+---
+task_categories:
+- zero-shot-classification
+tags:
+- audio
+- sound
+pretty_name: audio
+size_categories:
+- 1K<n<10K
+---