Your open-source personal assistant.
Website :: Documentation :: Roadmap :: Contributing :: Story
I'm taking a lot of time to work on the new core of Leon due to personal reasons. I can only work on it after work and on weekends. Hence, I'm blocking any potential contribution as the whole core of Leon is coming with many breaking changes. Many of you are willing to contribute in Leon (create new skills, help to improve the core, translations and so on...), a big thanks to every one of you!
I wish one day I could work on Leon full time, but it's not possible at the moment as I need to pay my bills. I have some ideas about how to monetize Leon in the future (Leon's core will always remain open source), but before to get there there is still a long way to go.
Until then, any financial support by sponsoring Leon is much appreciated 🙂
Since AI gained in popularity and large language models are getting more and more traction, many of you joined our community. A huge welcome to all of you! 🤗
At the moment, Leon's NLU will remain intents first with his own model without relying on an LLM. It is important that Leon can run 100% offline and I'm confident that with the downsizing techniques such as quantization Leon will sooner or later work with LLMs at his core and still be able to run on edge.
Here is how LLMs may help Leon in the future:
- Intent fallback: when an utterance cannot match an intent, then rely on an LLM to provide results.
- New named entity recognition engine: provide a better solution to extract entities from utterances such as fruits, numbers, cities, durations, persons, etc.
- Skill features: let skills leverage LLMs to provide out-of-the-box NLP features such as summarization, translation, sentiment analysis and so on...
- Skill building: LLMs can help to develop skills such as paraphrasing utterance samples, translate answers, convert code from our Python bridge to the upcoming JavaScript bridge and vice versa, etc.
- More...
Once the new core released, we'll work on the community aspect of Leon. For example, better organize our Discord, planify regular calls, work on skills together, etc. It is very important for Leon to have a real community. At that moment, the skills platform will already be online, so it'll be easier to sync our progress and publish new skills.
- Feel free to check out the Git development branches and our next major milestones.
- And the detailed roadmap.
Many exciting things are coming up, hence no new documentation and test are going to be written until the official release of Leon. Feel free to join us on Discord to know more and to read the "A Much Better NLP and Future" blog post.
Leon is an open-source personal assistant who can live on your server.
He does stuff when you ask him to.
You can talk to him and he can talk to you. You can also text him and he can also text you. If you want to, Leon can communicate with you by being offline to protect your privacy.
- If you are a developer (or not), you may want to build many things that could help in your daily life. Instead of building a dedicated project for each of those ideas, Leon can help you with his Skills structure.
- With this generic structure, everyone can create their own skills and share them with others. Therefore there is only one core (to rule them all).
- Leon uses AI concepts, which is cool.
- Privacy matters, you can configure Leon to talk with him offline. You can already text with him without any third party services.
- Open source is great.
This repository contains the following nodes of Leon:
- The server
- Skills
- The web app
- The hotword node
- The TCP server (for inter-process communication between Leon and third-party nodes such as spaCy)
- The Python bridge (the connector between the core and skills made with Python)
Today, the most interesting part is about his core and the way he can scale up. He is pretty young but can easily scale to have new features (skills). You can find what he is able to do by browsing the skills list.
Please do know that after the official release, we will build many skills along with the community. Feel free to join us on Discord to be part of the journey.
Sounds good to you? Then let's get started!
Gitpod will automatically set up an environment and run an instance for you.
To install these prerequisites, you can follow the How To section of the documentation.
# Install the Leon CLI
npm install --global @leon-ai/cli
# Install Leon (stable branch)
leon create birth
# OR install from the develop branch: leon create birth --develop# Check the setup went well
leon check
# Run
leon start
# Go to http://localhost:1337
# Hooray! Leon is running# Install Leon
leon create birth --docker
# Run
leon start
# Go to http://localhost:1337
# Hooray! Leon is runningFor full documentation, visit docs.getleon.ai.
To know what is going on, follow roadmap.getleon.ai.
If you have an idea for improving Leon, do not hesitate.
Leon needs open source to live, the more skills he has, the more skillful he becomes.
You'll find a write-up on this blog post.
Louis Grenard (@grenlouis)
Appwrite
250 USD / month |
Anonymous
100 USD / month |
herbundkraut
10 USD / month |
You? |
You can also contribute by sponsoring Leon.
Please note that I dedicate most of my free time to Leon.
By sponsoring the project you make the project sustainable and faster to develop features.
The focus is not only limited to the activity you see on GitHub but also a lot of thinking about the direction of the project. Which is naturally related to the overall design, architecture, vision, learning process and so on...
Copyright (c) 2019-present, Louis Grenard louis@getleon.ai
