stable-diffusion-backed-image-gen

This project implements a real-time AI image generation pipeline using:

• React frontend (hosted on AWS)
• FastAPI backend hosted on Fly.io
• Modal as the GPU-backed image generation worker

---

---

System Architecture

[ React UI ] → POST /enqueue → [ FastAPI ]
                     ↳ POST /status → [ FastAPI ] 
                                ↳ generates image

Deployment

1. Deploy Modal Worker

Steps:

1. Sign up at https://modal.com

2. Create a Hugging Face token:

   • https://huggingface.co/settings/tokens
   • Make sure it has "read" access

3. Register your Hugging Face token with Modal:

modal secret create hf-token
# Paste in:
# HF_TOKEN=your_token_here

4. Deploy the worker:

modal deploy prompt_forge.py

---

FastAPI Endpoints

• POST /generate → accepts prompt, returns job ID
• POST /status → accepts job_id, returns status of job
• GET /models → returns list of available models

---

Notes

• The image generation logic uses FluxImageGenerator

---

Next Steps

• Add retry and error handling