Fix dictation live streaming (async server fns), stop-status indicator, remove meta copy

Dictation: custom HTML/JS component streaming partial transcripts live via server_functions

Fix dictation tab: stream partial transcript per chunk during recording

README: say CPU only, drop cpu-basic naming

Stream transcription, add dictation tab, drop hardware/RTF mentions

Rebuild as Gradio SDK Space: compile VibeASR.cpp at startup, real speech examples

Delete files Dockerfile examples/sample.wav with huggingface_hub

Upload app.py with huggingface_hub

Upload folder using huggingface_hub

initial commit