destinations.impl.lancedb.configuration

LanceDBCredentials Objects

@configspec
class LanceDBCredentials(CredentialsConfiguration)

[view_source]

uri

LanceDB database URI. Defaults to local, on-disk instance.

The available schemas are:

• /path/to/database - local database.
• db://host:port - remote database (LanceDB cloud).

api_key

API key for the remote connections (LanceDB cloud).

embedding_model_provider_api_key

API key for the embedding model provider.

LanceDBClientOptions Objects

@configspec
class LanceDBClientOptions(BaseConfiguration)

[view_source]

max_retries

EmbeddingFunction class wraps the calls for source and query embedding generation inside a rate limit handler that retries the requests with exponential backoff after successive failures.

You can tune it by setting it to a different number, or disable it by setting it to 0.

LanceDBClientConfiguration Objects

@configspec
class LanceDBClientConfiguration(DestinationClientDwhConfiguration)

[view_source]

dataset_separator

Character for the dataset separator.

options

LanceDB client options.

embedding_model_provider

Embedding provider used for generating embeddings. Default is "cohere". You can find the full list of providers at https://github.com/lancedb/lancedb/tree/main/python/python/lancedb/embeddings as well as https://lancedb.github.io/lancedb/embeddings/default_embedding_functions/.

embedding_model_provider_host

Full host URL with protocol and port (e.g. 'http://localhost:11434'). Uses LanceDB's default if not specified, assuming the provider accepts this parameter.

embedding_model

The model used by the embedding provider for generating embeddings. Check with the embedding provider which options are available. Reference https://lancedb.github.io/lancedb/embeddings/default_embedding_functions/.

embedding_model_dimensions

The dimensions of the embeddings generated. In most cases it will be automatically inferred, by LanceDB, but it is configurable in rare cases.

Make sure it corresponds with the associated embedding model's dimensionality.

vector_field_name

Name of the special field to store the vector embeddings.

sentinel_table_name

Name of the sentinel table that encapsulates datasets. Since LanceDB has no concept of schemas, this table serves as a proxy to group related dlt tables together.

fingerprint

def fingerprint() -> str

[view_source]

Returns a fingerprint of a connection string.