ovos-padatious-pipeline-plugin

An efficient neural network intent parser for the OpenVoiceOS (OVOS) ecosystem, implemented in pure numpy with a FANN-compatible model format.

This repository bundles a fork of the original padatious from Mycroft AI and exposes it as an OVOS pipeline plugin.

Documentation

Document	Description
Installation	How to install dependencies and the package
Intent Format	Syntax for writing .intent and .entity files
API Reference	Python API for IntentContainer and DomainIntentContainer
OVOS Pipeline	Using the plugin inside OVOS / configuration options
Architecture	Internal architecture and data flow
Theory	Algorithm design, decisions, and limitations
Prerelease Quirks	User-visible changes since the last stable release

Quick Example

from ovos_padatious import IntentContainer

container = IntentContainer('intent_cache')
container.add_intent('hello', ['Hi there!', 'Hello.'])
container.add_intent('search', ['Search for {query} (using|on) {engine}.'])
container.train()

result = container.calc_intent('Search for cats on CatTube.')
print(result.name)    # 'search'
print(result.matches) # {'query': 'cats', 'engine': 'CatTube'}

License

Licensed under the Apache 2.0 license.