FaceFusion
Current
CurrentNext
Current
CurrentNext
  • Introduction
    • Changelog
    • Disclaimer
    • Licenses
  • Installation
    • Platform
      • Linux
      • macOS
      • Windows
    • Accelerator
      • Linux
      • Windows
  • Usage
    • CLI Commands
      • General
      • Job Manager
      • Job Runner
    • CLI Arguments
      • Paths
      • Patterns
      • Face Detector
      • Face Landmarker
      • Face Selector
      • Face Masker
      • Frame Extraction
      • Output Creation
      • Processors
        • Age Modifier
        • Deep Swapper
        • Expression Restorer
        • Face Debugger
        • Face Editor
        • Face Enhancer
        • Face Swapper
        • Frame Colorizer
        • Frame Enhancer
        • Lip Syncer
      • UIs
      • Execution
      • Download
      • Memory
      • Misc
    • Deepfake Webcam
      • Linux
    • Platform Benchmark
    • Run With Docker
  • Troubleshooting
    • macOS
    • Windows
  • Workshop
    • Processor
    • UI Layout
    • UI Component
  • Knowledgebase
    • Face Vocabulary
    • Technical Terms
    • Error Codes
    • GPU Capabilities
  • Donation
  • FAQ
  • Project Links
    • GitHub
    • Discord
    • Shop
  • Partner Links
    • RunDiffusion
    • Pinokio
Powered by GitBook
On this page
  • CUDA
  • TensorRT
  • OpenVINO
  1. Installation
  2. Accelerator

Windows

CUDA

Compatible with NVIDIA graphic cards:

conda install conda-forge::cuda-runtime=12.8.0 conda-forge::cudnn=9.7.1.26

TensorRT

Suitable for high performance NVIDIA graphic cards:

pip install tensorrt==10.8.0.43 --extra-index-url https://pypi.nvidia.com

OpenVINO

Suitable for Intel Arc graphic cards:

conda install conda-forge::openvino=2025.0.0

Last updated 21 days ago