Open-source alternatives to Nbdev

RISE is an interactive notebook presentation tool and slideshow manager designed to transform Jupyter and IPython notebooks into structured slide decks. It functions as an extension that allows a user to toggle between a standard editable notebook view and a full-screen slideshow format during live demonstrations. The system utilizes the Reveal.js framework to render notebook cells as formatted slides, mapping Jupyter metadata to determine slide breaks and fragment sequences. This integration enables the creation of computational slide decks where running Python code and interactive visualiza

Determined is an open-source machine learning platform that simplifies distributed training, hyperparameter tuning, experiment tracking, and resource management. Works with PyTorch and TensorFlow.

Jupyter Book is a computational book publisher and static site generator that converts Jupyter notebooks and markdown files into interactive web books and publication-quality PDF documents. It serves as a markdown-based documentation tool that executes embedded code at build time and caches the resulting outputs for static display. The system distinguishes itself by supporting interactive data publications, allowing readers to engage with live computational widgets and launch notebooks in remote execution environments. It extends standard markdown with a system of roles and directives to supp

Flyte is a Kubernetes-based machine learning orchestrator and containerized pipeline manager designed for coordinating AI workflows and data pipelines. It functions as an engine for defining and executing resilient pipelines, utilizing a data lineage tracker to maintain immutable execution states and ensure reproducible outputs. The platform distinguishes itself by packaging individual tasks into separate containers to ensure dependency isolation and environment consistency. It provides specialized capabilities for machine learning, including the transformation of trained models into scalable

CML is a pipeline automation tool for training and evaluating machine learning models, functioning as a CI/CD system for machine learning. It serves as a cloud compute orchestrator and Git-based workflow manager that automates model training cycles through branch management, automated commits, and integrated reporting. The project distinguishes itself by provisioning ephemeral cloud instances or Kubernetes nodes to provide specialized hardware for compute-heavy tasks. It also manages remote compute runners, allowing the connection of self-hosted GPU clusters or on-premise machines to execute

Mercury is a framework for transforming Jupyter notebooks into interactive web applications, a notebook execution API, and a static site generator. It functions as a self-hosted application server that allows users to deploy password-protected notebooks as functional user interfaces without writing frontend code. The system distinguishes itself by mapping notebook widgets to a reactive web interface and synchronizing live application sessions across multiple users in real time. It enables remote execution of notebooks via an API to retrieve computation results as structured data and supports

Jupytext is a synchronization tool and text converter for Jupyter Notebooks. It transforms notebook files into plain text formats, such as Markdown or Python scripts, to enable line-by-line diffs and peer reviews within version control systems. The tool pairs notebook files with corresponding text files to maintain a dual representation of the same content. It uses bidirectional synchronization to update linked files based on the most recent modifications, allowing notebook content to be edited inside standard text editors. Beyond file conversion and synchronization, the project provides cap

Metaflow is a Python machine learning framework and MLOps workflow orchestrator designed to manage the lifecycle of data pipelines from local prototyping to production. It serves as a distributed compute manager and an experiment tracking system, enabling the creation of reproducible pipelines that transition between development and high-availability production environments. The framework distinguishes itself through an integrated checkpointing system that automatically persists intermediate data artifacts to remote storage, allowing failed runs to be resumed from the last successful step. It

Papermill is a Jupyter notebook execution engine and parameterization framework designed to run notebooks programmatically. It allows users to inject custom input values into notebooks to execute the same logic across different datasets, transforming interactive notebooks into reproducible data science pipelines. The project functions as a language-agnostic notebook runner and orchestrator, supporting kernels for Python, R, Julia, and Scala. It is distinguished by its cloud-integrated runner capabilities, featuring built-in handlers to read and write notebooks directly to storage providers su

The fastest ⚡️ way to build data pipelines. Develop iteratively, deploy anywhere. ☁️

Kedro is a data science pipeline framework and production toolbox designed to build reproducible, modular workflows using software engineering best practices. It functions as a data engineering orchestrator and catalog manager, bridging the gap between interactive analysis and maintainable production pipelines. The framework distinguishes itself by using a data catalog to decouple data access from processing logic and providing tools to transition analysis from interactive notebooks into structured workflows. It includes a workflow visualization tool that generates visual maps of data pipelin