OdysseyXL Family

Powered by Stability Ai

Welcome to the OdysseyXL family! This repository showcases the cutting-edge advancements in the OdysseyXL series, a collection of finetuned Stable Diffusion XL models designed to push the boundaries of creativity and realism in generative AI. Whether you're an artist, researcher, or developer, OdysseyXL has something for you. Feel free to join our discord server

---

🚀 Models in the OdysseyXL Family

OdysseyXL 4.0

The most advanced SOTA generative model in the OdysseyXL Family, OdysseyXL 4.0 is perfect for:

• Ultra realism
• Full potential of OdysseyXL
• Cinematic and photographic images

OdysseyXL 3.0

Optimized for ultra-realism and vibrant creativity, OdysseyXL 3.0 is perfect for:

• High-detail art generation
• Photorealistic renders
• Excellent prompt adherence

OdysseyXL 2.0

A balanced model combining realism with stylistic flexibility. Ideal for:

• Concept art
• Digital painting
• Experimenting with styles

OdysseyXL 1.0

The original model that started it all. Great for:

• General-purpose image generation
• Prototyping creative ideas

---

🛠️ Usage

Option 1:

To get started with any OdysseyXL model, install the diffusers library and load the model using the following example:

pip install diffusers

Now we can the generation script. For this example we will use OdysseyXL-3.0

git clone https://github.com/Aayan-Mishra/OdysseyXL.git
cd OdysseyXL
python3 download-and-infrence.py

Option 2:

In this option you can use the new OdysseyXL API (Part of spestly package. Only supports 3.0+)!

pip install spestly

Now that we have the API system installed, we can now use it! A demo is shown below:

from spestly import OdysseyXL

odysseyxl = OdysseyXL()

# Define the prompt and negative prompt
prompt = (
    "An amateur cellphone photography of a black Ferrari. "
    "f8.0, Samsung Galaxy, noise, jpeg artefacts, poor lighting, low light, "
    "underexposed, high contrast"
)
negative_prompt = (
    "(octane render, render, drawing, anime, bad photo, bad photography:1.3), "
    "(worst quality, low quality, blurry:1.2), "
    "(bad teeth, deformed teeth, deformed lips), "
    "(bad anatomy, bad proportions:1.1), "
    "(deformed iris, deformed pupils), "
    "(deformed eyes, bad eyes), "
    "(deformed face, ugly face, bad face), "
    "(deformed hands, bad hands, fused fingers), "
    "morbid, mutilated, mutation, disfigured"
)

# Generate the image using the OdysseyXL model version 3.0
image = odysseyxl.generate(
    model_version="3.0",
    prompt=prompt,
    negative_prompt=negative_prompt,
)

image.save("output.png")
image.show()

---

🌟 Features

• Ultra-Realism: OdysseyXL 3.0 excels in generating images that blur the lines between reality and imagination.
• Fine Detail: Tailored for intricate and high-resolution artwork.
• Creative Flexibility: Adaptable to a wide range of artistic and technical applications.

---

📂 Available Models

Model	Size	API Access	Hardware
OdysseyXL 4.0	Large	HuggingFace, DiffuseCraftMod, OdysseyXL API	1xA100
OdysseyXL 3.0	Large	Civitai, HuggingFace, DiffuseCraft, OdysseyXL API	1xA100
OdysseyXL 2.0	Medium	Civitai, HuggingFace	1xL40S
OdysseyXL 1.0	Small	Civitai, HuggingFace	1xP100

---

📖 License

All models in the OdysseyXL family are shared under the Odyssey Labs Community license. Please ensure you adhere to the terms and conditions outlined in the license.

---

📢 Community and Support

Join the OdysseyXL community to share your creations, report issues, and stay updated:

• Discord: The Odyssey
• Twitter: @Spestly
• Submit your generated images!: Submit

For any questions or support, feel free to reach out via email at [email protected].