black-forest-labs

/

FLUX.1-Krea-dev

like 865

Follow

Black Forest Labs 37.6k

import torch
from diffusers import FluxInpaintPipeline
from diffusers.utils import load_image

pipe = FluxInpaintPipeline.from_pretrained("FLUX.1-Krea-dev", torch_dtype=torch.bfloat16)
pipe.to("cuda")
prompt = "Face of a yellow cat, high resolution, sitting on a park bench"
img_url = "https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo.png"
mask_url = "https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo_mask.png"
source = load_image(img_url)
mask = load_image(mask_url)
image = pipe(prompt=prompt, image=source, mask_image=mask).images[0]
image.save("flux_inpainting.png")


import torch

from diffusers import FluxImg2ImgPipeline
from diffusers.utils import load_image

device = "cuda"
pipe = FluxImg2ImgPipeline.from_pretrained("FLUX.1-Krea-dev", torch_dtype=torch.bfloat16)
pipe = pipe.to(device)

url = "https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo.png"
init_image_og = load_image(url)

init_image = init_image_og.resize((1024, 1024))

prompt = "cat sitting on a bench"

images = pipe(
     prompt=prompt, image=init_image, num_inference_steps=20, strength=0.95, guidance_scale=7.5).images[0]

images = images.resize(init_image_og.size)
# images.save("flux_img2img.png") 



import torch
from diffusers import FluxPipeline

pipe = FluxPipeline.from_pretrained("FLUX.1-Krea-dev", torch_dtype=torch.bfloat16)
pipe.to("cuda")

prompt = """
A cute, high-resolution sticker-style illustration of a happy dog sitting with its tongue out, big round eyes, floppy ears, soft tan and white fur, pastel color palette, outlined with a clean white border, cartoonish yet slightly textured, transparent background, perfect for digital sticker use, Flux render style
"""

# Depending on the variant being used, the pipeline call will slightly vary.
# Refer to the pipeline documentation for more details.
num_inference_steps = 30
image = pipe(prompt, 
                    num_inference_steps=steps,
                    generator=torch.Generator().manual_seed(124),
                    guidance_scale=guidance_scale
                    ).images[0]
image.save(f"flux_text_to_image.png")