skills.rest
hung-phan
hung-phanCOMMUNITY
Explore all skills by this author →
GitHub

training-workflow

Guide machine learning training workflows with data splitting, validation, and model serialization.

1|Updated Jun 4, 2026
Search Tags:#validation#machine learning#training#cross-validation#serialization
One-click install
npx skills add https://github.com/hung-phan/ml-skills --skill training-workflow
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: training-workflow
Source: https://github.com/hung-phan/ml-skills/tree/main/skills/ml-review/references/ml-training/training-workflow
Command: npx skills add https://github.com/hung-phan/ml-skills --skill training-workflow

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires numpy, scikit-learn, torch, torchvision, torchtext, tensorflow, pytorch-lightning, optuna, ray[tune], mlflow, wandb, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill offers a comprehensive toolkit to address various challenges in model training, from data splitting to final deployment.

Core Features & Use Cases

  • End-to-End Workflow: Guides users through cross-validation, hyperparameter tuning, and serialization for robust training.
  • Splitting Strategies: Covers time series, stratified, group-based splitting techniques.
  • Use Case: If you are starting a new ML project, use this skill to build a strong foundation from data preparation to evaluation.

Quick Start

Utilize the 'training-workflow' skill to understand best practices for your machine learning model training process.

Frequently Asked Questions about training-workflow

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
What is the best way to structure a machine learning training workflow from data splitting to deployment?▼

A structured machine learning training workflow should integrate data splitting, cross-validation, hyperparameter tuning, and model serialization to ensure data integrity and deployment preparedness. This Skill provides comprehensive guidance and tools for this entire process.

How do I implement cross-validation and data splitting strategies for time series or grouped data?▼

Cross-validation and data splitting require tailored strategies for different data types. You can implement time series, stratified, and group-based splitting techniques to maintain data integrity and ensure robust model evaluation during training.

Does this training workflow support hyperparameter tuning and experiment tracking with PyTorch Lightning and MLflow?▼

Yes, the training workflow supports hyperparameter tuning through Optuna and Ray[tune], and experiment tracking via MLflow and Weights & Biases. It integrates seamlessly with PyTorch Lightning and TensorFlow for comprehensive model development.

Can I use this Skill to prepare my models for production through serialization?▼

Yes, model serialization is a core component of this training workflow. It provides tools and best practices to properly serialize your trained machine learning models, ensuring they are fully prepared for downstream deployment.

What are the limitations of using standard validation approaches for complex machine learning projects?▼

Standard validation approaches often fail to account for data leakage and non-independent samples in complex machine learning projects. Utilizing group-based and time series splitting techniques mitigates these limitations and ensures robust model evaluation.

Related Skills

Explore More Software Engineering Skills→
JNZader-VaultJNZader-Vault

scikit-learn

Implement classical machine learning algorithms for tabular data with scikit-learn.

Official
Advanced
jackspacejackspace
16

scikit-learn

Build, evaluate, and deploy classical machine learning models with scikit-learn.

Community
Advanced
DifanaDAPDifanaDAP

weights-and-biases

Automate ML experiment tracking and hyperparameter sweeps via the Python API.

Community
Advanced
anxiety135790anxiety135790

weights-and-biases

Track machine learning experiments, sweeps, artifacts, and model registry state.

Community
Advanced