Skip to main content

Installation

This guide will help you install OpenTryOn on your system.

Prerequisites​

Before installing OpenTryOn, ensure you have:

  • Python 3.10 or higher
  • CUDA-capable GPU (recommended for best performance)
  • Conda or Miniconda (recommended)
  • Git (for cloning the repository)

Installation Methods​

Conda is the recommended installation method as it handles all dependencies including CUDA libraries.

Step 1: Clone the Repository​

git clone https://github.com/tryonlabs/opentryon.git
cd opentryon

Step 2: Create Conda Environment​

conda env create -f environment.yml
conda activate opentryon

This will create a new conda environment named opentryon with all required dependencies.

Step 3: Install Package​

pip install -e .

Method 2: Using pip​

If you prefer using pip, you can install dependencies directly:

Step 1: Clone the Repository​

git clone https://github.com/tryonlabs/opentryon.git
cd opentryon

Step 2: Create Virtual Environment​

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

Step 3: Install Dependencies​

pip install -r requirements.txt
pip install -e .

Method 3: Install from PyPI (Future)​

Once published to PyPI, you'll be able to install directly:

pip install opentryon

Verify Installation​

After installation, verify that OpenTryOn is correctly installed:

python -c "import tryon; print('OpenTryOn installed successfully!')"

Configuration​

Environment Variables​

Create a .env file in the project root with the following variables:

Preprocessing (Required for local preprocessing)​

U2NET_CLOTH_SEG_CHECKPOINT_PATH=path/to/cloth_segm.pth
U2NET_SEGM_CHECKPOINT_PATH=path/to/u2net.pth

API Integrations (Optional, for cloud-based services)​

# Segmind Try-On Diffusion API
SEGMIND_API_KEY=your_segmind_api_key

# Kling AI Virtual Try-On API
KLING_AI_API_KEY=your_kling_api_key
KLING_AI_SECRET_KEY=your_kling_secret_key

# Amazon Nova Canvas (AWS Bedrock)
AWS_ACCESS_KEY_ID=your_aws_access_key
AWS_SECRET_ACCESS_KEY=your_aws_secret_key
AMAZON_NOVA_REGION=us-east-1

# Google Gemini (Nano Banana Image Generation)
GEMINI_API_KEY=your_gemini_api_key

Note: You only need to configure the APIs you plan to use. For preprocessing-only workflows, only the U2Net checkpoints are required.

Download Model Checkpoints​

U2Net Cloth Segmentation​

Download the checkpoint from the huggingface-cloth-segmentation repository:

# Example: Download and place in project root
wget https://huggingface.co/levihsu/OOTDiffusion/resolve/main/cloth_segm.pth

U2Net Human Segmentation​

Download U2Net weights for human segmentation:

# Download from official U2Net repository
# Place in appropriate location and update U2NET_SEGM_CHECKPOINT_PATH

For optimal performance, ensure CUDA is properly configured:

Check CUDA Installation​

import torch
print(f"CUDA available: {torch.cuda.is_available()}")
print(f"CUDA version: {torch.version.cuda}")
print(f"GPU: {torch.cuda.get_device_name(0) if torch.cuda.is_available() else 'None'}")

Install CUDA-Compatible PyTorch​

If CUDA is not detected, install the appropriate PyTorch version:

# For CUDA 11.8
pip install torch torchvision --index-url https://download.pytorch.org/whl/cu118

# For CUDA 12.1
pip install torch torchvision --index-url https://download.pytorch.org/whl/cu121

Troubleshooting​

Common Issues​

Issue: Import Errors​

Solution: Ensure all dependencies are installed:

pip install -r requirements.txt

Issue: CUDA Out of Memory​

Solution: Reduce batch size or image resolution in your code.

Issue: Model Checkpoint Not Found​

Solution: Ensure checkpoint paths in .env are correct and files exist.

Issue: Dependency Conflicts​

Solution: Use conda environment to isolate dependencies:

conda env create -f environment.yml
conda activate opentryon

Getting Help​

If you encounter issues:

  1. Check the Troubleshooting Guide
  2. Search GitHub Issues
  3. Ask on Discord

Next Steps​

Once installed:

  1. Quick Start Guide: Get started with API integrations, datasets, and preprocessing
  2. Configuration Guide: Set up environment variables and API keys
  3. API Reference: Explore cloud-based virtual try-on and image generation APIs
  4. Datasets Module: Learn how to load and work with fashion datasets
  5. Preprocessing: Process garments, models, and images for virtual try-on

Key Features to Explore​

  • 🔌 API Integrations: Use cloud-based services without local model setup

    • Segmind Try-On Diffusion
    • Kling AI Virtual Try-On
    • Amazon Nova Canvas
    • Nano Banana Image Generation

  • 📊 Datasets: Load and work with fashion datasets

    • Fashion-MNIST (60K examples)
    • VITON-HD (11K+ pairs)
    • Subjects200K (200K paired images)

  • 🛠️ Preprocessing: Process images for virtual try-on

    • Garment segmentation and extraction
    • Human segmentation
    • Pose estimation