Skip to content

Latest commit

 

History

7 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

LangAdvisor

A desktop GUI or web application providing a fun toolkit for polyglots out there, including grammar assessment, practice and translation. Leverages local LLMs via Ollama for enhanced privacy, or cloud LLMs via Ollama Cloud service. Free-tier cloud LLMs such as minimax-m3 and gemma4-31b work well for most uses, though any tool-enabled LLM should work.

Features

Self-conscious about mistakes you might be making in messages or emails? You can quickly submit your message for an assessment of its correctness and intelligibility by a native speaker:

assessment

Want some interactive practice, suited to your level and a topic of interest? You can generate phrases and have your translations graded:

alt text

A lesson mode can be used to queue up a series of phrases that are graded at the end, with help provided along the way as needed. lesson

A translation mode enables you to get a variety of different potential translations for phrases: translation

Virtually any major language is supported, including ancient languages, and you can decide which Ollama model you wish to use depending on your desire for local execution, speed or accuracy. Latinization is supported for non-Western languages, where such a standard exists. More obscure languages may require a more powerful model to get decent results:

alt text

Prerequisites

  • Python 3.10+
  • uv
  • Ollama running locally with a model available (e.g. ollama pull llama3.2)

Running

You can download pre-built desktop apps for Windows and Linux in the Releases pane if you don't want to clone the source.

It's also quite easy to run from source if you have uv installed. Simply clone the source code and run:

   uv run main.py

uv will automatically create a local virtual environment and pull down needed dependencies.

As a webserver

If you have a server where you'd like to host the app for multiple users to access over the web, it's as simple as:

FLET_FORCE_WEB_SERVER=true FLET_SERVER_PORT=8008 uv run main.py

Then the app will be available on http://myserver:8008/ . Ensure you have ollama running wherever you are running this from. A raspberry pi or equivalent works well if you are routing requests to Ollama Cloud and want to make this available to devices in your local network.

About

Language instruction desktop and web app

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages