Use Cases

• Train SCVI model with custom dataloaders
• Perform downstream analysis tasks of SCVI models
• Optimize SCVI model with hyperparameter tuning
  • RAY
  • MLFLOW
• Using LLM Engines with scvi-tools
  • Claude (Anthropic)
  • ChatGPT (OpenAI)
  • OpenClaw
  • Gemini (Google)
  • BioMNI (Stanford)
  • Summary
• Train SCVI model with multi-GPU support
  • Some of the benefits of using MultiGPU training
  • Using MultiGPU training in SCVI-Tools
• Saving and loading SCVI models
• SCVI Criticism
  • There are a few metrics we calculate to achieve that:
  • Example of use:
  • Creating Report:
• Training configuration
  • When to use which config
  • Example (SCVI)
  • Example (SCANVI)
  • Choosing the right plan config
• Train SCVI model with callbacks