model-registry

What problem does it solve?

This Skill streamlines the management of AI model versions, enabling efficient promotion through different deployment stages and facilitating rollbacks when necessary.

Core Features & Use Cases

• Model Listing: View all registered models and their current deployment stages.
• Version Management: List specific versions of a model and compare different versions.
• Stage Promotion: Promote models to staging or production environments after validation.
• Rollback: Revert to a previous stable model version in case of issues.
• Archiving: Archive older model versions.
• Export: Export models in formats like ONNX for inference optimization.
• Use Case: A data science team can use this skill to safely deploy a new version of a trading strategy model to production after it has passed staging tests, or quickly roll back to a previous version if the new one underperforms.

Quick Start

Use the model-registry skill to promote the tsmom_v2 model to the staging environment.