How does it ensure data reproducibility?

The skill guides you in implementing data validation (like Great Expectations) and versioning (like DVC) to maintain a clear lineage of datasets and models.

Can I use this for experiment tracking?

Yes, it includes integration patterns for popular platforms like MLflow and Weights & Biases to track training metrics and manage model registries.

Does it handle model deployment?

It includes comprehensive guidance for various serving architectures and deployment strategies, including canary rollouts, shadow deployments, and A/B testing.

What orchestration tools does this skill support?

It provides patterns and templates for major industry-standard tools including Apache Airflow, Dagster, Kubeflow Pipelines, and Prefect.

ML Pipeline Workflow

Name: ML Pipeline Workflow
Author: HermeticOrmus

byHermeticOrmus

数据科学与机器学习

Orchestrates end-to-end MLOps pipelines from data preparation and model training to production deployment and monitoring.

关于

This skill provides a comprehensive framework for building and automating production-grade machine learning pipelines. It guides users through the entire ML lifecycle, including data ingestion, feature engineering, experimental tracking, and robust deployment strategies like canary or blue-green releases. By implementing DAG-based orchestration patterns and best practices for reproducibility, this skill helps developers and data scientists transform experimental models into reliable, observable, and scalable production systems.

主要功能

End-to-end DAG orchestration for complex ML workflows
Production-ready deployment patterns including Canary and Blue-Green
Experiment tracking integration with MLflow and Weights & Biases
0 GitHub stars
Automated data validation and versioning for reproducible pipelines
Integrated model monitoring and performance regression detection

使用场景

Building automated retraining pipelines triggered by data drift
Standardizing model validation and deployment across production environments
Designing reproducible training workflows for team collaboration

关于

主要功能

End-to-end DAG orchestration for complex ML workflows
Production-ready deployment patterns including Canary and Blue-Green
Experiment tracking integration with MLflow and Weights & Biases
0 GitHub stars
Automated data validation and versioning for reproducible pipelines
Integrated model monitoring and performance regression detection

使用场景

Building automated retraining pipelines triggered by data drift
Standardizing model validation and deployment across production environments
Designing reproducible training workflows for team collaboration