handle various input types e.g across different gradio versions

vad (#3)

normalization (#2)

dynamic sample rate (#1)

display WERs as percentage

Push latest updates

Update offline_pipeline.py

extended intro and limited file size to 5MB

ruff

summarize offline enhancement

Add Gradio info message for processing indication to beginning

Update transcribe_audio function to include stream_name parameter and adjust usage in retrieve_audio_information

for debugging in hugginface

remove wer box

fix offline pipeline