So after multiple questions on Deviantart about how I make my artwork, I decided to release this for everyone to use. It's a mix of 3 different mixes, and YES it is of course NSFW compatible!
In order to get the same visual art style you MUST use this in your positive prompt box
(((by Atey Ghailan and Emmanuel Shiu,SamDoesArt1)))
Description
1.5 just the base version :)
FAQ
Comments (23)
Would it be possible to share some prompts? I'm using the trigger words and I'm having no success getting anything close to these previews.
Of course!
(unity 8k wallpaper:1.5), (best quality:1.3),(illustration:1.3),(100 layer:1.3),(ultra detailed:1.3),(sharp:1.3), ((extreme detail)), delicate and beautiful, Character Enhance, Perfect Anatomy, Beautiful and Aesthetic, Brilliant Colors, {Intricate Detail}, highres, light particles, flares, light rays, ((masterpiece)), extremely detailed, solo, fancy boobs, fancy pussy, fancy hair, masterpiece, best quality, highly detailed,masterpiece, best quality, highly detailed,pussy focus, realistic hair, realistic body, realistic pussy, dynamic pose, 1girl, realistic full body white girl, (((dark brown hair))) , body tattoos, by Atey Ghailan and Emmanuel Shiu, SamDoesArt1, slim waist, anime, bright, colorful, high contrast, saturated, leg tattoos, thigh tattoos, chest tattoos, ((((naked))), ((((slim))))
Thank you :)
Do you use any specific negative prompts?
I tried again, both the copied prompt but also putting the trigger words first and some other stuff. I don't get anything close to this. Strange.
@barefooter1234 Have you installed the SamDoesart1 textual inversion/embed? It's available on this website and there are several (so try messing around with a few! I personally like 2). You have to have it installed or else the trigger word has nothing to actually call up and it won't do anything.
Link for SamDoesArt1:
https://civitai.com/models/1247/samdoesarts-embedding-1
If you're using AUTOMATIC111's GUI, just drop it into the embeddings folder.
Stable Diffusion 1.5\stable-diffusion-webui\embeddings
The samdoesart embeds are really potent, so you will see a difference if its working.
@anxious_vivi I'm using MNKD's GUI and if it's possible to use embeddings/txt inversions with it, I have to admit that I don't know how.
@barefooter1234 I'm not...quite sure. Never used that one or helped anyone with that one before. There should be a folder somewhere for it, or a place to make it. Its late for me or else I'd install it and see where it is myself, but I have to sleep.
I imagine it's pretty similar to automatic111's directory layout, so try looking inside the main folder.
There is an official discord which I found on the itch.io page, so if you have Discord, ask and someone hopefully should be able to sort you out.
@anxious_vivi I use it because I prefer not to use a web browser for my UI, and since I've found it well made and easy to use in general. I thank you for your help anyway. I'll keep on trying.
Do you use any VAE with this?
At the visual level, VAE only affect color. Anything v3 is a good all around VAE that most of these #anime checkpoints are run on, but feel free to mess around with any you find (or none at all—it gives a subdued, pastel look sometimes). Overall it will not affect the prompt, just the color intensity or lack thereof.
I'm not sure what the author is running on, but it's probably this.
(direct download)
https://civitai.com/api/download/1/model/Anything-V3.0.vae.pt
Yes, I use Ema-Pruned
Do you use any VAE to generate theese images?
At the visual level, VAE only affect color. Anything v3 is a good all around VAE that most of these #anime checkpoints are run on, but feel free to mess around with any you find (or none at all—it gives a subdued, pastel look sometimes). Overall it will not affect the prompt, just the color intensity or lack thereof.
I'm not sure what the author is running on, but it's probably this.
(direct download)
https://civitai.com/api/download/1/model/Anything-V3.0.vae.pt
Yes, I use Ema-Pruned
If you could upload a safetensor file would be awesome
Absolutely
so I've been trying to replicate your cowboy prompt without any success. Besides Samdoesart1, are there any embedings/lora that I am missing?
I am using a VAE which I have posted on another comment, besides that I am not using anything else. Try putting the image into the info tab of SD and try copying the prompt, see if that works!
@woundded Thank you! I noticed it was on clip skip 1 and from what I gather from the resolution - I assume you have quite the machine and are not using --xformers. Could be exactly that, but honestly it was just out of personal curiocity. But yeah the only thing I noticed is that I get 2 girls from time to time.
The Safetensor version is giving me the following error:
safetensors_rust.SafetensorError: Error while deserializing header: HeaderTooLarge
I was able to fix this by changing the extension from .safetensors to .ckpt
I keep getting an AssertionError in my client that says its not a safetensors file. Is this supposed to be in my Stable-Diffusion models or another folder?
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.