XLabs-AI/flux-controlnet-depth-diffusers

1---2license: other3language:4- en5base_model:6- black-forest-labs/FLUX.1-dev7pipeline_tag: text-to-image8tags:9- diffusers10- controlnet11- Flux12- image-generation13---14 15# Description16This repository provides a Diffusers version of FLUX.1-dev Depth ControlNet checkpoint by Xlabs AI, [original repo](https://huggingface.co/XLabs-AI/flux-controlnet-depth-v3). 17 18![Example Picture 1](depth_result.png?raw=true)19 20# How to use21This model can be used directly with the diffusers library22 23```24import torch25from diffusers.utils import load_image26from diffusers import FluxControlNetModel27from diffusers.pipelines import FluxControlNetPipeline28from PIL import Image29import numpy as np30 31generator = torch.Generator(device="cuda").manual_seed(87544357)32 33controlnet = FluxControlNetModel.from_pretrained(34  "Xlabs-AI/flux-controlnet-depth-diffusers",35  torch_dtype=torch.bfloat16,36  use_safetensors=True,37)38pipe = FluxControlNetPipeline.from_pretrained(39  "black-forest-labs/FLUX.1-dev",40  controlnet=controlnet,41  torch_dtype=torch.bfloat1642)43pipe.to("cuda")44 45control_image = load_image("https://huggingface.co/Xlabs-AI/flux-controlnet-depth-diffusers/resolve/main/depth_example.png")46prompt = "photo of fashion woman in the street"47 48image = pipe(49    prompt,50    control_image=control_image,51    controlnet_conditioning_scale=0.7,52    num_inference_steps=25,53    guidance_scale=3.5,54    height=768,55    width=1024,56    generator=generator,57    num_images_per_prompt=1,58).images[0]59 60image.save("output_test_controlnet.png")61```62 63## License64 65Our weights fall under the [FLUX.1 [dev]](https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/LICENSE.md) Non-Commercial License<br/>