Skip to main content
Need help with Syft Space? We’re here to support you through multiple channels.

​
Community support

GitHub Discussions

Ask questions, share ideas, and connect with the community

GitHub Issues

Report bugs, request features, and track ongoing work

OpenMined Slack

Chat with the OpenMined community in real-time

Documentation

Comprehensive guides and API references

​
Getting started

​
Common issues

If Syft Space cannot connect to Docker:
  1. Ensure Docker is running: docker ps
  2. Check Docker socket permissions: 
    # Linux
sudo usermod -aG docker $USER
newgrp docker
  3. For Docker Desktop on Mac/Windows, ensure it’s running and the socket is exposed
  4. Verify the volume mount in your docker run command includes: 
    -v /var/run/docker.sock:/var/run/docker.sock
If local dataset provisioning fails:
  1. Check available ports (8081, 50051 for ChromaDB)
  2. Verify Docker has enough resources (4GB RAM minimum, 8GB recommended)
  3. Check Docker logs: 
    docker logs syft-chromadb
  4. Restart the provisioner from the Datasets page
If you’re getting 401 Unauthorized errors:
  1. For admin endpoints, ensure you’ve set SYFT_ADMIN_API_KEY
  2. For query endpoints, verify your SyftHub token is valid
  3. Include the token in the Authorization header: 
    -H "Authorization: Bearer your-token"
  4. In development mode, leave SYFT_ADMIN_API_KEY empty to disable auth
If files aren’t being indexed:
  1. Check the ingestion path exists and is accessible
  2. Verify the dataset is in RUNNING state
  3. Check ingestion status in the Datasets page
  4. Review logs for file processing errors
  5. Supported formats: PDF, TXT, MD, DOCX, HTML
If you can’t connect to your AI model:
  1. Verify the API key is correct
  2. For custom endpoints (Ollama, vLLM), ensure the base_url is reachable
  3. Check model health in the Models page
  4. Test the endpoint directly: 
    curl http://localhost:11434/api/generate # Ollama
curl http://localhost:8000/v1/models     # vLLM
If Syft Space is using too much memory:
  1. Reduce the number of concurrent ingestion workers
  2. For large PDFs, the system uses docling which requires significant memory
  3. Allocate more memory to Docker (8GB+ for heavy workloads)
  4. Monitor memory in your system’s task manager

​
Report a bug

Found a bug? Help us improve Syft Space by reporting it.
1

Check existing issues

Search GitHub Issues to see if the bug has already been reported.
2

Gather information

Collect details about the bug:
  • Syft Space version (check Settings → About)
  • Operating system and version
  • Docker version: docker --version
  • Python version: python --version
  • Steps to reproduce the issue
  • Error messages or logs
3

Create an issue

Open a new issue with:
  • Clear, descriptive title
  • Detailed description of the problem
  • Steps to reproduce
  • Expected vs actual behavior
  • Screenshots or error logs (if applicable)

​
Request a feature

Have an idea for improving Syft Space?
1

Check existing requests

Search GitHub Discussions to see if the feature has been suggested.
2

Start a discussion

Create a new discussion with:
  • Clear description of the feature
  • Use case or problem it solves
  • How it would benefit users
  • Any implementation ideas
3

Get community feedback

Engage with the community to refine the idea and gather support.

​
Contributing

Want to help improve Syft Space? We welcome contributions!

Contributing guide

Learn how to contribute code, documentation, and more

​
System requirements

​
Minimum requirements

  • RAM: 4GB minimum, 8GB recommended
  • Storage: 2GB for application, additional space for datasets
  • Docker: Required for local vector databases
  • Python: 3.12 or higher (for development)
  • Node.js: 20 or higher (for frontend development)

​
Supported platforms

  • macOS 11+ (Intel and Apple Silicon)
  • Ubuntu 20.04+
  • Debian 11+
  • Fedora 36+
  • Windows 10+ (coming soon)

​
Diagnostic information

When seeking support, it’s helpful to provide diagnostic information:

​
Check application logs

Logs are written to:
  • macOS: ~/Library/Logs/syft-space/
  • Linux: ~/.local/share/syft-space/logs/

​
Check system status

# Check if Docker is running
docker ps

# Check Syft Space containers
docker ps -a | grep syft

# Check database
ls -lh ~/.syft-space/app.db

# Test API health
curl http://localhost:8080/health

​
Check provisioner status

Visit http://localhost:8080/api/v1/datasets/provisioners/ to see the status of all dataset provisioners.

​
Security issues

If you discover a security vulnerability, please DO NOT open a public issue.
Instead, report security issues privately to the OpenMined security team:
  • Email: [email protected]
  • Include detailed steps to reproduce
  • Allow time for the team to address the issue before public disclosure

​
Community resources

​
OpenMined ecosystem

Syft Space is part of the larger OpenMined ecosystem for privacy-preserving AI:

PySyft

Privacy-preserving machine learning framework

SyftHub

Decentralized marketplace for knowledge endpoints

OpenMined

Learn about the OpenMined mission and projects

OpenMined Blog

News, tutorials, and updates from the community

​
Stay updated

Can’t find what you’re looking for? Ask on GitHub Discussions or Slack.

Build docs developers (and LLMs) love

Get started for freeTalk to us