Voluntary Computing
Empower Community AI with Your Idle Compute Power
Voluntary computing enables open-source enthusiasts and community members to donate their spare computer resources (CPU, GPU, and memory) to public-interest projects. By sharing your idle hardware capacity, you directly contribute to training localized AI models (like Telugu LLMs), indexing corpus datasets, and running distributed tasks without relying on expensive, centralized cloud infrastructure.
Spare Resources
The background worker utilizes idle CPU/GPU cores, ensuring minimal impact on your normal daily activities.
Open Source & Secure
All code for the client and backend is fully open source, audited by the community, and focused strictly on transparent data computation.
Decentralized Power
By building a localized, peer-to-peer compute grid, we reduce dependencies on global corporations and retain sovereignty over AI models.
Device Registration & Setup Guide
Follow the terminal setup guide below to register your device and start receiving volunteer computing tasks.
Prerequisite: Install Python & uv
The client uses uv, an extremely fast Python package
installer and resolver. If you don't have it installed, run:
curl -LsSf https://astral.sh/uv/install.sh | sh
Clone the Client CLI Repository
Clone the official Swecha corpus client utility from the self-hosted repository. If you have already cloned it in the past, navigate to the directory and ensure it is updated.
git clone https://code.swecha.org/corpus/corpus-client-cli Navigate & Pull Latest Updates
Switch to the project directory and fetch the latest commits from the master repository.
cd corpus-client-cli
git pull Sync Dependencies
Install all Python project dependencies and configure the
virtual environment seamlessly using uv sync.
uv sync Login to Register Device
Register your device details and link them securely to your volunteer-compute profile by initiating the login sequence.
uv run corpus-client login Start Volunteer Compute Tasks
Initiate the volunteer computing routine. Your console will stay open to accept distributed jobs from the Swecha networks.
uv run corpus-client volunteer-compute Frequently Asked Questions
Can I control how much CPU/RAM is used?
Yes, the configuration files inside the client repository let you customize resource limitations. You can allocate a specific number of cores or hardware thresholds.
Is my data secure?
Absolutely. The client does not scan or access your personal files. It processes isolated computational datasets related to linguistics, text translation, and dataset generation sent from our central orchestrator.