Introduction to Pipcook

Pipcook is a comprehensive JavaScript-based machine learning and artificial intelligence development platform. It simplifies the process of creating, training, and deploying ML models with easy-to-use APIs. This guide introduces Pipcook’s powerful APIs with practical code snippets and a sample application.

API Examples

Pipcook.createPipeline()

This function allows you to create a pipeline for executing machine learning processes.

  const pipcook = require('@pipcook/core');

  pipcook.createPipeline({
    plugins: [
      {
        package: '@pipcook/plugins-csv-data-collect',
        params: {
          url: 'https://example.com/dataset.csv'
        }
      },
      {
        package: '@pipcook/plugins-bayesian-classify-model-define'
      },
      {
        package: '@pipcook/plugins-model-train'
      },
      {
        package: '@pipcook/plugins-model-evaluate'
      }
    ]
  });

Pipcook.runPipeline()

This function runs the created pipeline and trains the model based on the dataset.

  pipcook.runPipeline('path/to/your/pipeline/config').then(result => {
    console.log(result);
  });

Pipcook.loadDataset()

This function loads a dataset for training the model.

  pipcook.loadDataset('path/to/your/dataset').then(dataset => {
    console.log(dataset);
  });

Pipcook.trainModel()

This function trains the model using the specified dataset.

  pipcook.trainModel(model, dataset).then(trainedModel => {
    console.log(trainedModel);
  });

Sample Application

The following is an example application utilizing Pipcook APIs to create and train a machine learning model for image classification:

  const pipcook = require('@pipcook/core');

  async function createImageClassificationModel() {
    // Create a pipeline
    await pipcook.createPipeline({
      plugins: [
        {
          package: '@pipcook/plugins-image-classification-data-collect',
          params: {
            url: 'https://example.com/image-dataset.zip'
          }
        },
        {
          package: '@pipcook/plugins-image-classification-data-process'
        },
        {
          package: '@pipcook/plugins-image-classification-model-define'
        },
        {
          package: '@pipcook/plugins-model-train'
        },
        {
          package: '@pipcook/plugins-model-evaluate'
        }
      ]
    });

    // Run the pipeline, train the model, and get the result
    const result = await pipcook.runPipeline('path/to/your/pipeline/config');
    console.log(result);
  }

  // Execute the function
  createImageClassificationModel();

---

Hash: 549be0179e18739dfd5562b717112136caa9e09b8742cecda8af4e84f8a2a212