In this example, we'll build a simple LanceDB table containing embeddings for different languages that can be used for universal semantic search.
- The Dataset used will be wikipedia dataset in English and French
- The Model used will be cohere's multi-lingual model
In this example, we'll explore LanceDB's Embeddings API that allows you to create tables that automatically vectorize data once you define the config at the time of table creation. Let's dive right in!
To learn more about LanceDB, visit our docs
Run python script:
COHERE_API_KEY=... python main.py