sdwebuiapi
sdwebuiapi copied to clipboard
mix1009
Python API client for AUTOMATIC1111/stable-diffusion-webui
If we use script, then the txt2img only generate 1 image, even set batch_size=8 For instance result1 = api.txt2img(prompt="a man on tropical beach", negative_prompt="ugly, out of frame", seed=1003, styles=["anime"], cfg_scale=7,...
 Following the code shared in the repo.  I keep getting this timeout. Also, i am defaulting to port 7860 instead of 7861. Do i need to permit that...
Hi this API is amazing! I'm looking to build a docker image for deforum specifically -- and I wanted to know how I can use this API to make deforum...
I'm trying to use controlnet in the img2img-inpainting task: ``` unit1 = webuiapi.ControlNetUnit(input_image=img, module='depth', model='control_sd15_depth [fef5e48e]') result2 = api.img2img(images=[img], prompt="PHOTOREALISTIC, DREAMLIKEART, Masterpiece, best quality, 1 man, closeup of face, ,...
I use txt2img with depth ControlNet, but the result's picture is more greyish compared to the result produced by web-ui. I check the parameter one by one, but I do...
Any chance we can have [Multidiffusion](https://github.com/pkuliyi2015/multidiffusion-upscaler-for-automatic1111) extension implemented? Ty!
I used the following code to generate an image using `v1-5-pruned-emaonly` - `enable_hr=False` does not seem to work: ``` result = api.txt2img(prompt="dog", negative_prompt="ugly", seed=-1, cfg_scale=7, sampler_name='Euler a', steps=5, width=512, height=512,...
The ability to control the parameters of controlnet is only set by the GUI (the reference image, and the model).
in `webuiapi.py`: ``` class Upscaler(str, Enum): none = 'None' Lanczos = 'Lanczos' Nearest = 'Nearest' LDSR = 'LDSR' BSRGAN = 'BSRGAN' ESRGAN_4x = 'ESRGAN_4x' R_ESRGAN_General_4xV3 = 'R-ESRGAN General 4xV3' ScuNET_GAN...
