Watching my second Anime ever "JoJo's Bizarre Adventure: Stone Ocean" (The first was "One punch man"). Enjoying it a lot, so obviously fine-tuned a Stable-Diffusion model upon its images (After trying with hypernetwork a few times but failed, I chose full finetuning).
This is a full fine-tuning upon this blip-captioned dataset and there is no specific trigger word. You can try out some prompts online using the Inference Widget on Huggingface.
For some reason, limbs do not come out so well.
Description
This is a full fine-tuning upon this blip-captioned dataset and there is no specific trigger word. You can try out some prompts online using the Inference Widget on Huggingface.
Interestingly, in some of these generated images, the "Negative prompt" part was missing a ',' so it went into the positive and yet, came out okay
FAQ
Comments (6)
Best model, that's the coolest thing I've seen today
Thank you 🙏
Any tipps on what promts to use here, exemples and what on the non list ?
This is a full fine-tuning upon this blip-captioned dataset so I'm afraid there are no specific trigger words. It's supposed to apply the style on (almost?) everything you give it.
Any plans for a safetensor version?
Well, this model is more than a year old and I'm not so active on this site lately. I did make an SDXL LoRA for "Stone Ocean" at some point, if it helps: https://civitai.com/models/129551/jojo-stone-ocean-style-xl
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.