Local AI Development in VS Code

Using Continue Plugin for VS Code

  • Download from https://ollama.com/download (~ 1GB) and install (That actually is starting the setup file, one click to „Next“ and after a few minutes – done! VS Code is waiting for a connection, and a second after the installation success message VS Code is detecting that correctly:
  • Next is copying the suggested command into the terminal and execute it

This fails; in Windows you have to perform a restart, i guess, ollama is written to the path variable.

After restarting, the command works perfectly, it downloads ~5 GB of data

After successing, there is immediate response in VS Code

  • Let’s continue with downloading the autocomplete model:

Another 1 GB of data.

Aaaaand check!

  • And last step according to the guide: Downloading the embeddings model:

Another 0,3 GB are downloaded and a bunch of small ones…

Now all prerequisits are installed and we can connect:

Done.