Use official HuggingFace documentation approach to override model detection

Major: Revert to full Moshi model with L4 GPU upgrade

Fix Dockerfile directory permissions - create /app as root before switching users

Fix v1.1.2: README.md - Change sdk from gradio to docker for WebSocket streaming

Migrate working code: Deploy README.md v1.0.1 to correct Space

Force rebuild: Add rebuild trigger to YAML metadata

Fix: Add YAML frontmatter to README.md for HuggingFace Space configuration

Upload README.md with huggingface_hub

initial commit