简体中文 | English
GPUSnatcher is a tool for GPU resource monitoring and snatching, designed to help users temporarily monitor and grab idle GPU resources.
- Real-time GPU usage monitoring
- Command-line interface, easy to integrate into workflows
- Email notifications
- Scheduled automatic resource release
# Without pytorch
pip install gpusnatcher
# with pytorch
pip install gpusnatcher[cuda129]
Local installation：
git clone https://github.com/wilmerwang/GPUSnatcher.git
cd GPUSnatcher
# Without pytorch
pip install .
# with pytorch
pip install .[cuda129]
# Run for the first time to configure parameters
gpusk
Parameter description:
class ConfigData:
"""Configuration data for GPU Snatcher."""
gpu_nums: int
gpu_times_min: int
email_host: str
email_user: str
email_pwd: str
email_sender: str
email_receivers: list[str]
- gpu_nums: Number of GPUs to snatch
- gpu_times_min: Time to release resources after completing the specified GPU snatching plan
- email_host: Email server, e.g., smtp.qq.com
- email_user: Email address
- email_pwd: SMTP authorization code
- email_sender: Sender
- email_receivers: Recipients
Issues and pull requests are welcome. Please follow the project's code style guidelines.
This project is licensed under the MIT License.