Skip to content

Latest commit

 

History

History
85 lines (56 loc) · 2.02 KB

README.md

File metadata and controls

85 lines (56 loc) · 2.02 KB

eleventy-plugin-gpt4all

An experimental eleventy plugin for generating static websites using generative ai. Uses GPT4All to generate static assets from prompts stored as .md files.

Prerequisites

Make sure to install GPT4All before use.

Installation

npm i -D @11ty/eleventy eleventy-plugin-gpt4all

Usage

Register the plugin in your .eleventy.js file.

const gpt4AllPlugin = require('eleventy-plugin-gpt4all');

module.exports = function (eleventyConfig) {
    eleventyConfig.addPlugin(gpt4AllPlugin);
};

Then create a .md file for each page that you want on your website that contains a description that is 1600 tokens or fewer (the plugin reserves 448 tokens for prompt engineering).

The index.md file is registered as the homepage.

See the example repository here for an example of how to use the plugin.

Options

modelName

Choose which GPT4All-supported model to use to generate your static website. Defaults to mistral-7b-openorca.Q4_0.gguf.

eleventyConfig.addPlugin(gpt4AllPlugin, { 
    modelName:  'replit-code-v1_5-3b-newbpe-q4_0.gguf'
});

verbose

Enables verbose logging. Good for debugging.

eleventyConfig.addPlugin(gpt4AllPlugin, { 
    verbose: true 
});

prompts

Used to customize prompts.

initialCode

Provide the prompt for the initial code generation.

eleventyConfig.addPlugin(gpt4AllPlugin, { 
    prompts: {
        initialCode: (prompt, lang) => `Please generate code matching the prompt ${prompt} in the language ${lang}`
    }
});

codeReview

Provide the prompt for the code review.

eleventyConfig.addPlugin(gpt4AllPlugin, { 
    prompts: {
        codeReview: (lang, code) => `Please perform a code review of the code ${code} in the language ${lang}`
    }
});

Disclaimer

Please review the code generated by GPT4All before deploying to production.