Pika's Animated Mix: a new mix for generating quality images
My mix is capable of generating quality images in Anime and other similar styles.
NSFW content and the use of simple prompts
My mix is also capable of generating NSFW content. I mixed a lot of models to achieve the highest quality result possible. It is possible to use simple prompts to get a good result, but keep in mind that the model is sensitive to the prompt.
Leave your image generation results below in the comments
I'd love it if you could leave your image generation results below in the comments. That way I can improve the model.
PS: VAE from orangemix and version without VAE
At the moment the model has VAE from orangemix sewn in, but I'll be releasing a version without VAE soon. Stay tuned for updates!
Recommendations for use:
-Sampler: DPM++ SDE Karras, UniPC
-Steps: 20-40
-CFG: 4-7
-HiresFix: 1.5-2.5
-ClipSkip: 2
-Negative: easynegative, painting by bad-artist-anime, verybadimagenegative_v1.3, (low quality:1.2), (worst quality:1.2)
------------------------------------------------------------------------------------------------------
Pika's Animated Mix: новый микс для генерации качественных изображений
Мой микс способен генерировать качественные изображения в Аниме стиле и других подобных стилях.
NSFW контент и использование простых промптов
Мой микс также умеет генерировать NSFW контент. Я смешал множество моделей, чтобы достичь максимально качественного результата. Возможно использование простых промптов для получения хорошего результата, но имейте в виду, что модель чувствительна к промпту.
Оставляйте свои результаты генерации изображений ниже в комментариях
Я буду рад, если вы оставите свои результаты генерации изображений ниже в комментариях. Таким образом я смогу совершенствовать модель.
PS: VAE от orangemix и версия без VAE
На данный момент модель имеет вшитый VAE от orangemix, но в скором времени я выпущу версию без VAE. Следите за обновлениями!
Рекомендации к использованию:
-Sampler: DPM++ SDE Karras, UniPC
-Steps: 20-40
-CFG: 4-7
-HiresFix: 1.5-2.5
-ClipSkip: 2
-Negative: easynegative, painting by bad-artist-anime, verybadimagenegative_v1.3, (low quality:1.2), (worst quality:1.2)
Available on https://tensor.art/u/597898445910444159 !
Description
Pruned version!
FAQ
Comments (15)
Updated: No Vae, pruned!
Very pretty outputs, and pretty easy to use a short prompt, but as soon as you start prompting for more specifics, there starts to be problems. Definitely better for generalized vague prompts than specific ones!
This is Witchcore model, it generates ex actually same images, maybe even slightly worse, https://civitai.com/models/26849/witchcore
Indeed the model is similar in style to witchcore, but I didn't use witchcore in my mix.
And don't worry, I'm preparing a new version that will be much less like witchcore and the like
модель хорошенькая )
many thanks!
Thank you to everyone who uses my mix, thank you for your likes, I am very happy that you liked the model that I made! in connection with this I will prepare something interesting for you :)
Been a big fan of it. Great job! Recently however, maybe there's something to learn for either of us here, any image generated turns out with green/pink static with a general outline of the subject. Isolated to this model. Use to work on any guidance scale value. Has been reinstalled. Checked to see if its on. PC is plugged in. etc. I love this model. I wanna keep using it. Anyone else experiencing this?
Hello,Can you please show me pictures of what you have going on?
@PikaDesigner What's the best way for you to receive images on here? I'm new to this platform and I'd rather not post my problem into your gallery. :)
Edit: I see the link option, but if there's anyway I can avoid making an imgur account specifically for this.
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.