Stable Diffusion Image Generator

📌 Overview

Harness the power of Hugging Face's "Stable Diffusion" model with this image generator. Stable Diffusion, an avant-garde image synthesis model, seamlessly melds the capabilities of both CLIP and Diffusion techniques, churning out visually arresting images from textual descriptions. Explore the mechanics behind this innovative model here.

This system elegantly bifurcates into two agents:

1. User Agent: Sends forth image description requests with precision.
2. AI Model Agent: Skillfully interfaces with the Hugging Face API, bringing images to life.

🛠 Requirements

• Python (v3.10+ preferred for optimal performance)
• Poetry (A dynamic Python packaging and dependency management tool)

Key Libraries: Ensure your environment is configured with HUGGING_FACE_ACCESS_TOKEN.

✨ Features

• User Agent gracefully interacts with the AI Model Agent, firing periodic image synthesis requests.
• AI Model Agent efficiently retrieves images using the Stable Diffusion model, ensuring vividness and accuracy.
• Embedded error handlers diligently log and manage discrepancies, ensuring smooth operations.

Getting Started

1. Acquire the HuggingFaceAPI Token:

   1. Visit HuggingFace.
   2. Register or enter your credentials to log in.
   3. Chart a course to Profile -> Settings -> Access Tokens.
   4. Copy an existing token or mint a new one.

2. Environment Setup: Within the stable_diffusion-v1-4/src directory, forge a .env file and inscribe your HuggingFaceAPI Token:

   export HUGGING_FACE_ACCESS_TOKEN="{Your HuggingFaceAPI Token}"

3. Install Dependencies:

   Venture into the source directory, load the environment variables from .env, and install necessary packages:

   cd src
   source .env
   poetry install

4. Ignite the Main Script:

   Kickstart the project using:

   poetry run python main.py

This sequence dispatches requests to the agent at 10-minute intervals.

In your logs, scout for:

Adding agent to Bureau: {agent_address}

Snatch the {agent_address} value and transpose AI_MODEL_AGENT_ADDRESS in src/stable_diffusion_agent.py with this value.

Wish to tailor your request? Navigate to src/stable_diffusion_user.py and tweak the IMAGE_DESC variable to customize the textual prompt.

Your freshly minted images reside in the generated-image folder. Named with a timestamp, for instance: 2023-09-03 11:00:59.jpeg.

Dive in and conjure some mesmerizing visuals!