New model focused on realistic images, based on 8Buff_Gen model: https://civarchive.com/models/41928/8buffgen
Notes and details:
This model does require a good deal of prompt crafting, the trade off is you have a lot of control on the images you create. I would recommend finding a prompt that generates the style and quality to your liking and saving it as a style.
Use 512x512 size for most consistent results. Does Well with 512x768, 768x512, and 768x768.
Recommended sampling methods: DPM++ 2S a Karras, Euler a, DDIM, DPM2 a Karras, DPM++2S a,
Known issues:
Face distortion is still an issue but better than previous versions. Restore face and positive prompts address.
Tends to draw topless women about 10% of the time. This can be addressed by adding negative prompts nude and NSFw for image generation.
People tend to lean towards asian traits for eyes and hair. This can be addressed by adding nationality to the positive prompt. Nationalities that work with high consistency are: American, Indian, Nigerian, Japanese, and Ethiopian. Others such as European, Swiss, Mexican, work but not as consistently.
When using the Indian nationality prompt, it tends to give everyone a bindi. This can be addressed with adding (((bindi))) to your negative prompt.
Ages can be inconsistent sometimes with very young and very old people. Try positive prompts like
Improvements:
Requires less sampling steps to get quality images. Start at 30 and adjust from there.
Age consistency better with wider age range.
General purpose realistic model with a fine level of control. Merged with a lot of different models, but don’t remember all of them. Links below to those that were remembered or provided inspiration for direct training.
Trained on images and Mixed with:
8Buff_Gen - https://civarchive.com/models/41928?modelVersionId=46665
526Mix V1.4.5 - https://civarchive.com/models/15022/526mix-v145
Great with deform and Automatic1111 Animator extension for GIFs or videos.
See more here for results: https://youtube.com/shorts/KwgcCbbQ35k?feature=share
Check out our How-tos here:
https://eightbuff.blogspot.com/p/my-text-to-img-journey-list-of-how-tos.html
Buy us a cup of coffee:
https://www.buymeacoffee.com/eightbuffalo
Follow up on Facebook:
Description
New model focused on realistic images, based on 8Buff_Gen model: https://civitai.com/models/41928/8buffgen
Pruned version!
Notes and details:
This model does require a good deal of prompt crafting, the trade off is you have a lot of control on the images you create. I would recommend finding a prompt that generates the style and quality to your liking and saving it as a style.
Use 512x512 size for most consistent results. Does Well with 512x768, 768x512, and 768x768.
Recommended sampling methods: DPM++ 2S a Karras, Euler a, DDIM, DPM2 a Karras, DPM++2S a,
Known issues:
Face distortion is still an issue but better than previous versions. Restore face and positive prompts address.
Tends to draw topless women about 10% of the time. This can be addressed by adding negative prompts nude and NSFw for image generation.
People tend to lean towards asian traits for eyes and hair. This can be addressed by adding nationality to the positive prompt. Nationalities that work with high consistency are: American, Indian, Nigerian, Japanese, and Ethiopian. Others such as European, Swiss, Mexican, work but not as consistently.
When using the Indian nationality prompt, it tends to give everyone a bindi. This can be addressed with adding (((bindi))) to your negative prompt.
Ages can be inconsistent sometimes with very young and very old people. Try positive prompts like
Improvements:
Requires less sampling steps to get quality images. Start at 30 and adjust from there.
Age consistency better with wider age range.
General purpose realistic model with a fine level of control. Merged with a lot of different models, but don’t remember all of them. Links below to those that were remembered or provided inspiration for direct training.
Trained on images and Mixed with:
8Buff_Gen - https://civitai.com/models/41928?modelVersionId=46665
526Mix V1.4.5 - https://civitai.com/models/15022/526mix-v145
Great with deform and Automatic1111 Animator extension for GIFs or videos.
See here for results: https://youtube.com/shorts/KwgcCbbQ35k?feature=share
Check out our How-tos here:
https://eightbuff.blogspot.com/p/my-text-to-img-journey-list-of-how-tos.html
Buy us a cup of coffee:
https://www.buymeacoffee.com/eightbuffalo
Follow up on Facebook:
FAQ
Comments (5)
Dang I do like my Swiss Miss.
What I'm finding interesting is, the first image in a batch I do is just what I'm looking for... but if I have batch count or batch size set to anything higher than a 1, I get furniture and knick-knacks for everything after the first image.
That's odd, we almost always run batches usually 4 or more. Haven't see that issue. If you find out this issue I would love to know what causes it.
turns out it's not just your checkpoint... it's happening for some reason across many of my models. I'm about to say screw it and re-install everything, if I can't figure it out. Almost seems like A1111 is ignoring my promts after it runs the first image...
@techken1 Interesting. Before complete reinstall, I would try deleting your venv folder under your stable-diffusion-webui directory. When you run A1111 again it should rebuild it and re-pull any requirements for you. If that doesn't work, I would probably reinstall.