AutismMix is a Pony V6 based model merge which specializes in anime art. Model's purpose is making Pony more predictable and excels in one style, it doesn't require negative prompts like other Pony models and work with `score` tags. AutismMix has two different version for different purposes. This guide will cover everything about AutismMix.
What This Model Excels At
- Has Pony Diffusion based special tags for better experience.
- Generating human/humanoid, however model still can generate pony/furry images but it's not main purpose of model. (Not recommended to use for furry/pony images.)
- Generating female/male/transgender subjects.
- Crafting SFW/ content imagery.
- Crafting images with wide-range themes. (fantasy, sci-fi, medieval etc.)
- Crafting anime images with certain art styles. (pixel art, abstract, line-art etc.)
- Recognizes many of Danbooru based tags, but model training is not Danbooru based.
- Knowledge of many well-known anime/cartoon/game characters.
- Knowledge of various face/body/characteristic/outfit terms.
- Can create amazing background/landscapes.
Cautions For Optimal Usage
- Trained to work without negative prompt.
- Doesn't recommended to use embeddings made for Pony Models.
- Does not require quality modifiers like `masterpiece, best quality, hd` etc.
- Make sure to use Clip Skip at 2.
- Can't create photorealistic images.
- Use `source_anime` tag to achieve desired anime style.
- `score` tags must be used front of prompt.
- Use negatives just for unwanted stuff or style, e.g `hat, yellow, 3d, child`
- 1024px resolution is recommended, although model can generally work with most of supported SDXL resolutions.
Recommended Settings
- Steps: 20-40 (25-28)
- CFG Scale: 7-8
- Sampler: Euler a, DPM++ 2M Karras, DPM++ SDE Karras
- Clip Skip: 2
- VAE: sdxl_vae
Suggested Resources (AutismMix_pony)
LoRA
- Style LoRAs for Pony Diffusion V6 XL
Prompt Style
- Requires `score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, source_anime` in front of positive prompt.
Note: `score_9` is much weaker than full string because of an training issue. Prefer suggested full string for better results.
- Does not need any negative or `masterpiece, best quality` like terms.
- Data selection tags can be used to make AI focus on certain contents.
- Can work with both of short and long prompts.
- Prefers tag like prompts. Model isn't Danbooru based but can understand Danbooru tags.
Recommended Prompt Order
- `score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, source_anime`
- Data Selection Tags (User Preference)
- Subject
- Describe your suggested composition in general order. Use tag based prompts for better understanding.
Special Data Selection Tags
- Tags can be used in both positive and negative prompts to focus on a specific data of the Model's training database.
Source
Model excels in anime style, be sure to use `source_anime` in your prompt. You can use other source tags too.
- source_cartoon
- source_pony (Not Recommended)
- source_furry (Not Recommended)
Content Safety Rating
- rating_safe
- rating_questionable
- rating_
Version Differences
AutismMix_pony
- Better LoRA usage.
- More similar to Pony V6 XL model.
AutismMix_confetti
- Better at general usage.
- Better at rendering hands.
- More predictable results.
- More stable style.
Confetti is better at general usage but choose Pony one, if you are going to work with LoRA.
Prompt Examples
Graveyard Keeper
Positive Prompt
`score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, source_anime, rating_safe, SFW, 1girl, red eyes, black hair, red streaked hair, white pale skin, , white tank top, (off-shoulder), black shirt, denim shorts, vivid colors, graveyard background, moon`
Negative Prompt
`source_, `
Flower Garden
Positive Prompt
`score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, source_anime, rating_safe, SFW, close-up, front view, 1girl, blue eyes, purple hair, straight long hair, straw hat, formal white dress, walking, flower garden, sunshine, bokeh, line-art`
Negative Prompt
`source_, `
You can join SeaArt Discord Community for more content, also you will have opportunity to choose next model guides with your vote. You can use comments or SeaArt server #guides-feedback to leave feedback or suggestion. We look forward to seeing you on the server.
Happy Generations To Everyone!
