pipcook

Machine learning platform for Web developers

What is pipcook?

Machine learning platform for Web developers

Overview of pipcook

What it is

Pipcook is a JavaScript application framework for machine learning and its engineering. It is designed to enable JavaScript engineers to utilize machine learning without prerequisites. The project provides subprojects including a machine learning pipeline framework, management tools, and a JavaScript runtime for machine learning. These can be used as building blocks in conjunction with other projects. Pipcook is an open-source project guided by principles of being modular and swappable. It is intended for Web engineers looking to learn machine learning, train and serve models, or optimize models for better evaluation results. The project includes Pipcook Pipeline, which represents ML pipelines consisting of scripts, and Pipcook Bridge to Python, which provides access to Python packages via N-API using the Boa module.

Key features

  • Modular architecture with well-defined functions and APIs that work together.
  • Swappable modules that allow most components to be replaced by different implementations.
  • Pipcook Pipeline for representing ML pipelines, using a plug-in mechanism to support dataset, training, validations, and deployment.
  • Pipcook Bridge to Python (Boa) that provides access to Python packages such as numpy, scikit-learn, jieba, tensorflow in the Node.js runtime through JavaScript.
  • Outputs trained models as NPM packages containing the model and JavaScript functions.
  • Command-line tool for managing projects, training, predicting, and serving models.

Use cases

  • Web engineers learning machine learning concepts and workflows.
  • Training image classification models, such as classifying avatar and blurBackground images.
  • Training text classification tasks using bayes.
  • Training object detection models for component recognition.
  • Serving trained models as a local HTTP server for inference.

Getting started

Prepare Node.js >= 12.17 or >= 14.0.0 and npm >= 6.14.4. Install the command-line tool:shell$ npm install -g @pipcook/cliTrain an image classification pipeline:shell$ pipcook train https://cdn.jsdelivr.net/gh/alibaba/pipcook@main/example/pipelines/image-classification-mobilenet.json -o ./outputPredict using the trained model:shell$ pipcook predict ./output/image-classification-mobilenet.json -s ./output/data/validation/blurBackground/71197_223__30.7_36.jpgDeploy the model as a server:shell$ pipcook serve ./output

License and pricing

Pipcook is licensed under Apache 2.0. It is open source with no stated pricing.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/alibabapipcook.svg)](https://yunodb.com/tools/alibabapipcook?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/alibabapipcook?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/alibabapipcook.svg" alt="Featured on YunoDB" height="20"></a>