Originally posted to HuggingFace by yehiaserag
A dreambooth finetune of stable diffusion 1.5 model that will output stuff in anime pencil concept drawing style.
Usage
Keep the words anime pencil concept style towards the beginning of your prompt to invoke the finetuned style.
Description
FAQ
Comments (7)
any chance of adding a safetensor version maybe? :D
There is in the huggingface repo
any settings suggestions such as no of steps , promts, sampling method , etc.. will be really helpfull
Will update the readme in huggingface since stuff are different for v4
A small note: Use DDIM sampler, anything starting 20 steps is good, 50 works great too, try with one or multiple tokens from the trigger words especially for v4
can we get a pencil drawing with no colors , just the drawing?
If you mean in just black pencil then you are in luck, v5 will be trained on a dataset of 5k images that includes 10% images with pencil only, soooon :D
Now you can with V5, just include the words "black and white"