Skip to content

A local Ollama2-based language learning app. Made during AdaHack23 as a team of four.

Notifications You must be signed in to change notification settings

juliammikk/LlamaLearner

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

21 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

LLAMA learner (demo)

An interactive language learner app.

Developed during AdaHack 23 to the following challenge: Design a mini-game that uses an open AI chatbot to teach players basic phrases or vocabulary through interactive dialogue.

Functionality

The app allows you to have a bilingual conversation in a setting of your choice. The goal: ask the right questions to identify five basic vocabulary words.

Input is supported in both English and the second language of choice, and the app can even handle language mixing within a sentence! A locally installed LLAMA2 LLM generates replies. All replies are printed in both languages, as well as played out loud using Google's text-to-speech models. The last reply can also be replayed by pressing the speaker button.

Demo

The demo sets the second language to Italian and the setting to a bakery. Five simple words were generated using the LLM beforehand. image

Requirements

To run the program, Ollama's LLAMA2 model needs to be installed locally, as well as the Python libraries PyQt6, langchain, translate, and gtts).

README author: Jakub Malý (https://github.com/jakub-maly)

About

A local Ollama2-based language learning app. Made during AdaHack23 as a team of four.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages