Upload app.py with huggingface_hub

Upload app.py with huggingface_hub

Upload app.py with huggingface_hub

Upload app.py with huggingface_hub

Upload app.py with huggingface_hub

Upload requirements.txt with huggingface_hub

Upload app.py with huggingface_hub

Use ZeroGPU hardware

Enable ZeroGPU inference API

Use CPU Space for protected API

Fix Space startup without ZeroGPU

Add API runtime dependencies

Add protected inference API

Delete index.html with huggingface_hub

Full-feature Gradio UI: 15,856 characters, artwork, RAG, presets, no history

Mount UI routes via Gradio app factory, preserving paths

Gradio launches (ZeroGPU detection); Flask inserted ahead of its catch-all

Register a GPU probe so ZeroGPU detects a spaces.GPU function at startup

FastAPI at root with Flask mounted; Gradio parked at /gradio for ZeroGPU

Gradio owns the server; Flask UI mounted onto it

Replace with the real PersonaMini UI: full character library, CharacterCodex, RAG, no history panel

Delete app.py with huggingface_hub

PersonaMini Chat: three models, character library, session-only memory

initial commit