https://github.com/rickstaa/onediff_test

Contains a simple test to measure the speedup provided by the OneDiff algorithm.
https://github.com/rickstaa/onediff_test

Last synced: 11 months ago
JSON representation

Contains a simple test to measure the speedup provided by the OneDiff algorithm.

• Host: GitHub
• URL: https://github.com/rickstaa/onediff_test
• Owner: rickstaa
• Created: 2024-05-17T20:58:58.000Z (about 2 years ago)
• Default Branch: main
• Last Pushed: 2024-05-17T20:59:27.000Z (about 2 years ago)
• Last Synced: 2025-03-04T02:44:48.058Z (over 1 year ago)
• Language: Python
• Size: 1.95 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# OneDiff Speedup Test

This repository contains a simple test to measure the speedup provided by the OneDiff algorithm.

## How to Use

Follow these steps to set up and run the test:

1. **Set up a Python 3.10 Conda environment**

   Create a new Conda environment named `onediff-test` with Python 3.10:

   ```bash

   conda env create -n onediff-test python=3.10

   ```

2. **Activate the environment**

   Switch to the new environment:

   ```bash

   conda activate onediff-test

   ```

3. **Install the OneDiff library**

   Run the provided script to install the OneDiff library with CUDA 11.2 support:

   ```bash

   bash install_one_diff_cuda121.bash

   ```

   This script follows the [official OneDiff installation instructions](https://github.com/siliconflow/onediff#1-install-oneflow).

4. **Verify the OneDiff installation**

   Verify that the OneDiff library is successfully installed by importing it in Python:

   ```bash

   python -c "import onediff"

   ```

5. **Run the test**

   Execute the test script:

   ```bash

   python onediff_test.py

   ```

   This script uses the OneDiff library to perform text-to-image inference using a pre-trained model.

# onediff_test