Add professional README with animated SVG banner, features table, and architecture diagram

Add retry logic for vision model (handles 500/504 timeouts)

Fix: replace non-existent en-US-DavisNeural with ChristopherNeural

Try Qwen2.5-VL-72B-Instruct for vision

Switch vision model to Qwen/Qwen2.5-VL-7B-Instruct (supported by inference providers)

Fix vision model: switch to Llama-3.2-11B-Vision + resize images to 1024px max

Add error handling with clear error messages for each pipeline stage

Add Chair/moderator who introduces the topic and speakers

Complete redesign: Small Talk style side-by-side robot panels with captions, embedded audio auto-play via iframe srcdoc

Fix robot sync: use polling watcher + hidden textbox for timeline data (scripts dont re-execute in gr.HTML updates)

Fix autoplay: use inline HTML (no iframe) + sync to Gradio autoplay audio element

Fix autoplay: add click-to-play overlay fallback and allow-autoplay on iframe

Auto-play audio in robot viewer with synced transcript log

Add animated robot avatars synced to speaker turns

Update edge-tts to >=7.2.5 to fix 403 token issue

Use CohereLabs/aya-vision-32b - confirmed available on HF inference

Switch to Qwen models - available on HF inference without gating

Fix asyncio: use new_event_loop instead of nest_asyncio

Remove explicit provider - let auto-router select with valid HF token

Fix InferenceClient: explicitly use hf-inference provider

Fix huggingface-hub version to be compatible with gradio 6.19

Remove gradio pin - HF Spaces provides it via sdk_version

Initial commit: GeoTalk AI Geoscience Debate Podcast

initial commit