Nuevo Foundation Workshops

Navigation:

Nuvi wants quick, private translations—so you’ll host the translation “brain” right on your machine.

Nuvi says: Keeping it local means no internet delay and better privacy for practice projects.

1. Add it to requirements.txt

Open your project’s requirements.txt and add (or confirm these lines are present):

libretranslate
requests

Save the file.

To open the terminal you can:

Use the menu: View → Terminal
OR press the keyboard shortcut: Ctrl+`` (backtick key, usually above Tab). On macOS it’s Cmd+``.
OR click the Terminal tab / plus (+) icon at the top of the panel area.

Make sure your virtual environment is selected (look for .venv in the bottom status bar). If not, select it.
Open a new terminal in VS Code (Ctrlxd+` ).
Run:

pip install -r requirements.txt

pip show libretranslate

Start the server in the terminal (keep this window open while you test):

libretranslate --load-only en,es

This will start the server with support for translating English and Spanish—the two languages Nuvi wants to support first.

Nuvi tip: Need more languages later? Restart without `--load-only` or add codes (e.g. `--load-only en,es,fr`).

The first time running it may download language data. That can take a few minutes.

LibreTranslate starting

Once it’s finished, open a browser to:

http://127.0.0.1:5000

You should see a simple page. To see languages the server supports, visit:

http://127.0.0.1:5000/languages

Problem	Fix
`No module named libretranslate`	Ensure your terminal shows the (.venv) text or ask an adult for help.
Server starts then quits	Read the error text; reinstall or try again.
Can’t open 127.0.0.1:5000	Make sure the server is still running; check firewall.
First translation slow	Models still loading. Wait a bit.
Output is empty / weird	Make sure you used POST with keys: `q`, `source`, `target`, `format`.
Port already in use	Use a different one: `python -m libretranslate --port 5001` and update the URL.