@zerda.js/runtime

Zerda.js Runtime Environment | Cross-platform | Modular | Wide GPU support

What is the Zerda.js Runtime Environment?

The Zerda Runtime Environment (Zerda RE) is a JavaScript runtime environment (JSRE) that sits on top of an existing JSRE to supply it with the Brain.js library, with GPU-acceleration supporting a wide variety of GPUs to ensure optimal performance across all devices. This makes it easy to train and deploy AI models not only on individual devices, but also across a large network of internet-connected devices.

Furthermore, Zerda also provides several classes and functions that make peer-to-peer (P2P) communication between nodes in the network effortless. Simply create a new session, share your passphrase with other users (or programmatically share it with other nodes), and you'll have a distributed AI network put together in no time.

Installation

Install the command-line interface

To install globally to use from the command line, please run the following command:

npm i -g @zerda.js/runtime

Install as a dependency

To install as a dependency to use as a library, please run the following command:

To install as a dependency

npm i @zerda.js/runtime

Usage

How to use the command line interface

Process a string

echo "[ 0, 1 ]" | zerda gh:voidvoxel/zerda-example-plugin

Run a zerda project

zerda -p ./samples

or

zerda -p ./samples/zerda.json

Process an input file (npm)

cat samples/numbers.txt | zerda zerda-example-plugin zerda-example-plugin > samples/numbers.json

Process an input file (GitHub)

cat samples/numbers.txt | zerda gh:voidvoxel/zerda-example-plugin > samples/numbers.json

How to use as a library

import ZerdaRuntime from "@zerda.js/runtime";


// Create a new runtime instance.
const zerdaRuntime = new ZerdaRuntime();

// Get the plugin.
const plugin = zerdaRuntime.evalGitHub("gh:voidvoxel/zerda-example-plugin");

// Set the input values.
const input = [ 1, 2 ];

// Get the output values.
const output = plugin(input);

// Log the input values.
console.log(
    "input:",
    JSON.stringify(input)
);

// Log the output values.
console.log(
    "output:",
    JSON.stringify(output)
);