GPU Detection Scripts

This repository contains Python scripts to check GPU availability and status using PyTorch and TensorFlow.

Files

• gpu_test.py: Checks GPU availability using PyTorch and TensorFlow.
• gpu_test_pytorch.py: Checks GPU availability using PyTorch only.
• gpu_test_tf.py: Checks GPU availability using TensorFlow only.

Usage

To check for GPU availability, run the appropriate script:

• For both PyTorch and TensorFlow:

  python gpu_test.py

• For PyTorch only:

  python gpu_test_pytorch.py

• For TensorFlow only:

  python gpu_test_tf.py

Output

Each script will display the following information:

• PyTorch:

  • If CUDA is available.
  • Number of CUDA devices.
  • Name of the first CUDA device (if available).

• TensorFlow:

  • Number of GPUs available (prefixed with Num GPUs Available:).

Requirements

Ensure you have the necessary libraries installed. The scripts will notify you if either library is missing. Install them individually with:

pip install torch tensorflow pytest

Requirements

Ensure you have the necessary libraries installed. The scripts will notify you if either library is missing. Install them individually with:

pip install torch tensorflow pytest

Alternatively, install everything from the provided requirements file:

pip install -r requirements.txt

Testing

This project uses pytest for its test suite. After installing the required packages, run:

pytest

This command executes the tests under the tests directory and verifies that each GPU detection script runs without errors.

Acknowledgments

Thanks to the open-source community for the tools and libraries used.