Proposed change

This PR adds the ability for users to select the embedding model version (Jina CLIP v1 or v2) for Semantic Search.

• The V2 model adds multilingual support (89 languages).
• The V2 model has significantly higher RAM usage and inference time than the V1 model with only a 3% improvement on embedding retrieval, so the general recommendation is that users (especially native English speakers) continue to use the V1 model.
• Users who want to run V2 should have a discrete, powerful GPU. CPU inference with the small model is not recommended.

Type of change

• Dependency upgrade
• Bugfix (non-breaking change which fixes an issue)
• New feature
• Breaking change (fix/feature causing existing functionality to break)
• Code quality improvements to existing code
• Documentation Update

Additional information

• This PR fixes or closes issue: fixes #
• This PR is related to issue:

Checklist

• The code change is tested and works locally.
• Local tests pass. Your PR cannot be merged unless tests pass
• There is no commented out code in this PR.
• The code has been formatted using Ruff (ruff format frigate)