安装PhotoDoodle需要遵循严格的步骤，以下是详细的环境准备指南：
- Basic Environment Configuration
- 安装Git版本控制系统
- 准备Python 3.11.10环境
- 安装Conda包管理器
- Get Project Code
- fulfillment
git clone git@github.com:showlab/PhotoDoodle.git - Go to the project directory
cd PhotoDoodle
- fulfillment
- Creating a Virtual Environment
- (of a computer) run
conda create -n doodle python=3.11.10 - activation environment
conda activate doodle
- (of a computer) run
- Installation of dependent libraries
- 优先安装GPU版本的PyTorch：
pip install torch==2.5.1 torchvision==0.20.1 torchaudio==2.5.1 - Install other dependencies:
pip install --upgrade -r requirements.txt
- 优先安装GPU版本的PyTorch：
- Download model files
- 从GitHub Releases或Hugging Face下载预训练模型
- 将模型放置到
checkpoints/directory
强烈建议使用NVIDIA GPU加速处理速度，安装过程中需注意Python和PyTorch版本的严格匹配。如果遇到依赖冲突，可以尝试重建虚拟环境。
This answer comes from the articlePhotoDoodle: AI tool for adding artistic doodles to photos with text commandsThe
May not be reproduced without permission:AI productivity tools " 如何安装和配置PhotoDoodle运行环境？
English 
简体中文 
日本語 
Deutsch 
Português do Brasil