databricks-ml-bundle 0.1.0

CLI tool to generate Databricks ML platform project structure with governance and best practices

Databricks ML Bundle CLI

A CLI tool to generate Databricks ML platform project structures with governance and best practices.

Installation

pip install databricks-ml-bundle

Usage

Generate a new ML project:

databricks-ml-init --name my-ml-project --workspace-host https://your-workspace.cloud.databricks.com --model-type segmentation --use-gpu

Or use the short command:

dml-init -n my-ml-project -w https://your-workspace.cloud.databricks.com -m classification

Options

• --name, -n: Project name (required)
• --output-dir, -o: Output directory (default: current directory)
• --workspace-host, -w: Databricks workspace URL (required)
• --model-type, -m: Model type - classification, regression, segmentation, nlp, custom (default: custom)
• --use-gpu: Enable GPU configuration for training

Generated Structure

The CLI generates a complete ML platform project with:

• Multi-environment support (dev/stg/prod)
• Unity Catalog integration
• MLflow experiment tracking
• Quality gates and approvals
• CI/CD pipeline with GitHub Actions
• Cluster policies and security
• Modular Python package structure

Example

# Generate a computer vision project
databricks-ml-init \
  --name vista-segmentation \
  --workspace-host https://my-workspace.cloud.databricks.com \
  --model-type segmentation \
  --use-gpu

# Navigate to project
cd vista-segmentation

# Install dependencies
pip install -r requirements.txt

# Deploy to Databricks
databricks bundle validate --target dev
databricks bundle deploy --target dev

Features

• ✅ Governance-first: Built-in security, permissions, and audit trails
• ✅ Multi-environment: Separate dev/staging/production environments
• ✅ Model-specific: Templates optimized for different ML use cases
• ✅ Production-ready: Includes serving endpoints, monitoring, and CI/CD
• ✅ Unity Catalog: Full integration with Databricks governance platform

Development

# Clone repository
git clone https://github.com/yourusername/databricks-ml-bundle-cli
cd databricks-ml-bundle-cli

# Install with Poetry
poetry install

# Run locally
poetry run databricks-ml-init --help

Publishing to PyPI

# Build package
poetry build

# Publish to PyPI
poetry publish

License

MIT License