About this version

This update focuses on resolving many of the issues present in version V3, while also serving as a more general-purpose base model. The most noticeable change is the change toward a toned-down anime style, leaning closer to Style A, but with some of the visual changes of later updates. Model is also more lora friendly, because of the more base model nature (noob based loras).

Next update: The model displays noticeable noise on some concepts. I'm working to reduce this effect.

Changes in compare to hassaku V3:

- Fixed the issue where prompts with fewer than 75 tokens produced poorer results

- The overall visual style is significantly toned down — less glossy or “shiny”

- Certain tags like "Mature female" now produce more accurate results

- Slight improvements for furry-related prompts

- Better handling of chibi and pixel art styles

- Model has been trained further on more recent data (characters like Varesa from Genshin now work using her tag alone)

- Significantly reduced chances of extreme brightness, discoloration, or overly orange/dark results

- Reduced orange/sepia tint on many concepts (including yellowish bodys)

- Better working artist tags

- Better results for more niche or abstract prompts, not just common ones like “1girl”

Worse for some:

- Composition is less restrictive: As example, other models mostly place multiple characters in a row, sitting or standing on tags 4girls or 5girls. My model creates more complex layouts. While this creates more variety, it also increases the chance of failed images (messy limbs, missplaced Bodyparts)

- The model is less biased toward closer images, often generating wider scenes

Worse:

- Some concepts may look less detailed or simplified, especially those involving characters

- More appearance of multiple views (if this happen, put it into the negatives)

- Colors are less vibrant

- Bodies are sometimes to bright

Note to new characters:

Many of the newer characters still require supporting tags to generate them more accurately

Note2 if some are interested:

Adding new characters and up-to-date knowledge is not as simple as it may seem, especially if you want to avoid bias. For example, every cow-girl might end up resembling Varesa (from Genshin). Avoiding that takes a lot of time. Even now, cow horns having still a bias torwards her horns.

Note3:

Some loras that worked on V1 before, now work on V3 barerly at all or on low strengh, because of further training.

Data-cut early September, adding data in the training progress. Some characters like alice_(genshin_impact) will not work, even if already included.

Base Model Creators position #7

Followers - 14535

14.5k

Likes - 89783

89.8k

Supporter Badge February 2023

Ikena's Avatar

Ikena

Joined Dec 26, 2022

License:

Illustrious License

Hassaku aims to be a anime model with a bright and distinct anime style.

You can run Hassaku XL and use its API on SinkIn: https://sinkin.ai/m/JWknjgr

My Discord for everything related to anime models and art.

If you'd like to support my work, you can do so at: SubscribeStar, every bit helps and is truly appreciated!

____________________________________________________________

Supporters:

Thanks to my supporters pttcot, SETI and Kodokuna

____________________________________________________________

Using the model:

The model is trained using images with minimal disruptive elements such as floating text, logos, speech bubbles, and signatures. If any of these elements are present in an image, please include the prompt "signature" as a negative prompt

Metadata and franchise tags are excluded—please do not use them. Tags such as "highres" and franchise-related tags like "re:zero kara hajimeru isekai seikatsu" are not used in training

Look at the example images, to see how the model should be used

Model is trained to be not complicated in use, tag order:

First are the number of person, then character names, rest.

Example: 1girl, rem \(re:zero\), standing, masterpiece, upper body

NoobAIs Loras working for the most part better in compare to Illustrious loras

Here are some resolution options for SDXL:

1536 x 640

1344 x 768

1216 x 832

1152 x 896

1024 x 1024

896 x 1152

832 x 1216 (most recommended)

768 x 1344

640 x 1536

______________________________________________________

Version and License info:

Below V1 merges ANIMAGINE XL 3.0

Version V1 use Illustrious-XL & WAI-NSFW-illustrious-SDXL with additional training

V2 is trained on its own and don't include any extra merge. Base was V1, it is trained to include newer or missing characters. It was also used for further training tests

V3 is a merge of Illustrious-XL and V2, to fix some issues and was also trained to include newer or missing characters

All Models using the Fair AI Public License 1.0-SD license.

Suggested Resources

Discussion

schneesturmx91988

a day ago

i was waiting for a fix version but after alot testing around its rly hard to get good results for 2girls... because of duplicates or extra faces, same for solo stuff in high resolution dupes appears and ignores solo prompt , solo is not without a reason in every image what makes it kinda hard if u want a duo on a image and not 3 characters or a duplicated one. same for the anatomy for a basic image generation its still fine but as soon u try some more action it breaks sadly : pro no skin shine anymore hand + finger gen is ofc better . thats it .yeah i hope the next version can do this better without duplicates in more characters or extra hands /legs if asked which resolution ? 1536x1024 its the same if i would use a lower reso it can happend to there .what would i love if the next version has the pros that i wrote and the negatives are gone: (stable anatomy,no dupes or extra faces that appears out of a body)<- no franchise tags ofc and just a basic prompt set . dont give up its just my test results after some hours because i love testing stuff

gonfinal12

a day ago

3.1 looks great, it finally works with close up and portrait. also details are better.

bafangsui156

a day ago

Where can I get the "sdxl_vaeFix.safetensors" used in the sample image?

I can't find it anywhere.

The output results are not as good as the sample, probably because the VAE is different.

vernhunt

19 hours ago

does it support artist tag?

See 6 more hidden comments

Civitai | Share your models

About this version

Next update: The model displays noticeable noise on some concepts. I'm working to reduce this effect.

Changes in compare to hassaku V3:

- Fixed the issue where prompts with fewer than 75 tokens produced poorer results

- The overall visual style is significantly toned down — less glossy or “shiny”

- Certain tags like "Mature female" now produce more accurate results

- Slight improvements for furry-related prompts

- Better handling of chibi and pixel art styles

- Model has been trained further on more recent data (characters like Varesa from Genshin now work using her tag alone)

- Significantly reduced chances of extreme brightness, discoloration, or overly orange/dark results

- Reduced orange/sepia tint on many concepts (including yellowish bodys)

- Better working artist tags

- Better results for more niche or abstract prompts, not just common ones like “1girl”

Worse for some:

- The model is less biased toward closer images, often generating wider scenes

Worse:

- Some concepts may look less detailed or simplified, especially those involving characters

- More appearance of multiple views (if this happen, put it into the negatives)

- Colors are less vibrant

- Bodies are sometimes to bright

Note to new characters:

Many of the newer characters still require supporting tags to generate them more accurately

Note2 if some are interested:

Note3:

Some loras that worked on V1 before, now work on V3 barerly at all or on low strengh, because of further training.

Data-cut early September, adding data in the training progress. Some characters like alice_(genshin_impact) will not work, even if already included.