AI/diffusers
1
0
Fork 0
mirror of https://github.com/huggingface/diffusers.git synced 2026-01-27 17:22:53 +03:00
Code Activity

Updates Image to Image Inpainting community pipeline README (#1370)

Browse Source 
* updates img2img_inpainting README

* Adds example image to community pipeline README
This commit is contained in:
Alex McKinney
2022-11-29 12:17:22 +00:00
committed by GitHub
parent 4d1e4e24e5
commit bcb6cc16df
1 changed files with 6 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
examples/community/README.md
View File

@@ -602,7 +602,7 @@ For example, this could be used to place a logo on a shirt and make it blend sea
import PIL
import torch
from diffusers import StableDiffusionInpaintPipeline
from diffusers import DiffusionPipeline
image_path = "./path-to-image.png"
inner_image_path = "./path-to-inner-image.png"
@@ -612,10 +612,11 @@ init_image = PIL.Image.open(image_path).convert("RGB").resize((512, 512))
inner_image = PIL.Image.open(inner_image_path).convert("RGBA").resize((512, 512))
mask_image = PIL.Image.open(mask_path).convert("RGB").resize((512, 512))
pipe = StableDiffusionInpaintPipeline.from_pretrained(
pipe = DiffusionPipeline.from_pretrained(
"runwayml/stable-diffusion-inpainting",
custom_pipeline="img2img_inpainting",
revision="fp16",
torch_dtype=torch.float16,
torch_dtype=torch.float16
)
pipe = pipe.to("cuda")
@@ -623,6 +624,8 @@ prompt = "Your prompt here!"
image = pipe(prompt=prompt, image=init_image, inner_image=inner_image, mask_image=mask_image).images[0]
```
![2 by 2 grid demonstrating image to image inpainting.](https://user-images.githubusercontent.com/44398246/203506577-ec303be4-887e-4ebd-a773-c83fcb3dd01a.png)
### Text Based Inpainting Stable Diffusion
Use a text prompt to generate the mask for the area to be inpainted.