Train ML models with feature views and register them in the Hopsworks model registry.
npx skills add https://github.com/logicalclocks/hopsworks-api --skill hops-train
Please help me install this Agent Skill. Skill: hops-train Source: https://github.com/logicalclocks/hopsworks-api/tree/main/skills/hops/hops-train Command: npx skills add https://github.com/logicalclocks/hopsworks-api --skill hops-train
This Skill facilitates the training of ML models using feature views within the Hopsworks platform, automating the process of loading data, training models, and registering them in the Hopsworks model registry.
To train a model using a feature view named 'my_fv', use the hops-train skill with the command: hops-train --feature-view my_fv.
High-intent search queries and answers about installing and using this skill.
You can train an ML model using a Hopsworks feature view by executing the training command with your feature view name. This automates loading data from the feature view, training the model, and registering it in the Hopsworks model registry.
Model registration in the Hopsworks model registry works by automatically saving the trained models along with their associated metrics and plots. This ensures the models are versioned and ready for downstream deployment for real-time serving or batch inference.
Yes, you need access to a Hopsworks cluster and the Python SDK to train and register ML models. The Skill requires an active cluster environment to utilize feature views for data input and to access the Hopsworks model registry for deployment.
The best way to automate ML training and deployment in Hopsworks is to use a Skill that integrates feature views for data loading and the model registry for deployment. This streamlines the pipeline from data ingestion to real-time serving or batch inference.
Yes, you can use this approach for both batch inference and real-time serving. Once the model is trained using a feature view and registered in the Hopsworks model registry, it can be deployed directly to support either inference pattern.
logicalclocksMonitor Hopsworks feature stores for drift and data quality issues.
belokonmDeploy ML models to production with infrastructure-as-code templates and monitoring.
logicalclocksManage Hopsworks Feature Store resources via the hops CLI with JSON output.
scawfulPlan, train, evaluate, and deploy Zelda/ASM models with a unified lifecycle tool.
