Download vosk model in package

[ftnext]

No description provided.

[Copilot]

Pull Request Overview

This PR updates the Vosk model download and loading workflow to store the model in the package's "models/vosk" directory and improve error handling.

• Recognizers now use pathlib for model path resolution and raise a SetupError if the model is not found.
• The CLI is updated to download the model directly into the package’s designated "models/vosk" directory.
• The models directory is now fully ignored via .gitignore, preserving only the .gitignore file itself.

Reviewed Changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated no comments.

File	Description
speech_recognition/recognizers/vosk.py	Updated model path resolution using pathlib and raised SetupError on missing model
speech_recognition/models/.gitignore	Configured to ignore all files except .gitignore
speech_recognition/cli.py	Adjusted CLI to download the Vosk model to the package’s models/vosk directory