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:

Set up a Python 3.10 Conda environment

Create a new Conda environment named onediff-test with Python 3.10:
```
conda env create -n onediff-test python=3.10
```
Activate the environment

Switch to the new environment:
```
conda activate onediff-test
```
Install the OneDiff library

Run the provided script to install the OneDiff library with CUDA 11.2 support:
```
bash install_one_diff_cuda121.bash
```
This script follows the official OneDiff installation instructions.
Verify the OneDiff installation

Verify that the OneDiff library is successfully installed by importing it in Python:
```
python -c "import onediff"
```
Run the test

Execute the test script:
```
python onediff_test.py
```
This script uses the OneDiff library to perform text-to-image inference using a pre-trained model.