Installation and configuration steps
InstantCharacter requires a GPU-enabled environment to run, here is a detailed installation guide:
environmental preparation
- Python version 3.8+
- PyTorch (CUDA support)
- Git and Git LFS for large file downloads
- Recommended NVIDIA GPU (minimum 48GB VRAM, optimized to support 24GB)
Installation process
- Install Python and PyTorch:
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
- Install dependent libraries:
pip install transformers accelerate diffusers huggingface_cli
- Cloning Warehouse:
git clone https://github.com/Tencent/InstantCharacter.git
- Download model checkpoints:
huggingface-cli download --resume-download Tencent/InstantCharacter --local-dir checkpoints
caveat
It is necessary to ensure that the GPU driver and CUDA version are compatible with PyTorch, and the model needs to be moved to a CUDA-enabled device. If you encounter download problems, you can use the Hugging Face mirror address.
This answer comes from the articleInstantCharacter: An Open Source Tool for Generating Consistent Characters from a Single ImageThe
May not be reproduced without permission:AI productivity tools " How do I install and configure the InstantCharacter tool? What are the requirements for the runtime environment?
