README.md

metadata

title: Anam Code
emoji: 🧑‍⚕️
colorFrom: blue
colorTo: purple
sdk: docker
app_port: 7860
pinned: false

Anam avatar demo — Hugging Face Spaces deploy package

Self-contained copy of the anam (FastAPI backend + frontend) and agent (LLM + local RAG server) apps, stripped of local caches/secrets, ready to push to a Hugging Face Space.

What's in here

• Dockerfile, start.sh — build the image and run both servers (agent internally on port 8005, anam publicly on port 7860 — the port Hugging Face Spaces expects).
• anam/ — the avatar app (frontend + backend in one FastAPI process).
• agent/ — the LLM proxy used by anam's "Agent LLM" mode.
• scripts/create_vector_store.py — one-off script to create the persistent OpenAI vector store used by anam's local RAG (upload documents → search).

Deploy steps

1. Create the vector store (once)

pip install openai
OPENAI_API_KEY=sk-... python scripts/create_vector_store.py

Copy the printed vs_... id — you'll need it in step 4.

2. Create the Space

On huggingface.co → New Space → SDK: Docker → Hardware: CPU basic (free).

3. Push this folder to it

git init
git add .
git commit -m "Initial deploy"
git remote add hf https://huggingface.co/spaces/<your-username>/<space-name>
git push hf main

4. Add repository secrets

Space → Settings → Variables and secrets → New secret:

Name	Value
ANAM_API_KEY	your Anam API key
OPENAI_API_KEY	your OpenAI API key
OPENAI_VECTOR_STORE_ID	the vs_... id from step 1

5. Done

The Space rebuilds automatically once secrets are set. Open the Space URL to verify.

Known limitation

anam/rag_state.json (native Anam-hosted RAG ids) and anam/telemetry_data.json reset on every Space rebuild/restart — the free tier has no persistent disk. The OpenAI-vector-store-backed local RAG (documents you upload via the UI) is not affected by this — it survives restarts as long as OPENAI_VECTOR_STORE_ID stays the same.