Add `pipeline_tag: automatic-speech-recognition`

The [base model](https://huggingface.co/openai/whisper-medium) has this tag.

README.md

@@ -1,5 +1,6 @@
 ---
 library_name: mlx
+pipeline_tag: automatic-speech-recognition
 ---
 
 # whisper-medium-mlx
@@ -13,4 +14,4 @@ pip install -r requirements.txt
 
 >> import whisper
 >> whisper.transcribe("FILE_NAME")
-```
+```