OdysseyXL is a personal project crafted as part of my portfolio, showcasing expertise in AI model finetuning and innovation in creative technology. Built on the SDXL framework, it delivers stunning image generation with exceptional realism, detail, and artistic flexibility. Designed to demonstrate technical skill and creative vision, OdysseyXL highlights my ability to push the boundaries of AI-driven creativity and performance.
Below we can se a comparison compared to other models:
3.0:
1.0:
2.0:
To use the model you can either access the model weight from HuggingFace and run it locally, or you can use DiffuseCraft to run the model on an A100 Space.
- Submit your generated images!: Submit!
Description
FAQ
Comments (3)
can you convert this to a checkpoint for automatic/comfy?
Sure give me a day or two an I will do it
For model v1 (v2) To use it unpack archive, place safetensor from unet folder to your unet models folder (rename it as OdysseyXL.safetensors. Place textencoders files from both folders to your textencoders folder (but also rename them to anything) and use DualClip Loader, where select both safetensors clip files. It works.