llimo
v1.0.4
Published
> 🚧 This project is a work in progress. For a more complete solution, please use [next-token-prediction](https://github.com/bennyschmidt/next-token-prediction).
Downloads
194
Readme
🚧 This project is a work in progress. For a more complete solution, please use next-token-prediction.
llimo
Large language and image models in pure JavaScript.
Install
npm i llimo
Usage
Put this /training/
directory in the root of your project.
Now you just need to create your app's index.js file and run it. Your model will start training on the .txt files located in /training/documents/
. After training is complete it will run these 3 queries:
const { Conversation: ChatModel } = require('llimo');
const MyChatBot = async () => {
const agent = await ChatModel({
bootstrap: true
});
// Predict the next word
agent.getTokenPrediction('what');
// Predict the next 5 words
agent.getTokenSequencePrediction('what is', 5);
// Chat with LLM
agent.ask('what is Thai food?');
};
MyChatBot();
Demo
Q&A Tests (Paris dataset):
https://github.com/bennyschmidt/llimo/assets/45407493/df8e9627-ff48-41ed-8714-a880beb8680a
A work in progress. Things needed:
A lot more training data - this directly translates to what it accurately "knows"
Better parts-of-speech parsing - to allow for more flexible, even misspelled inputs