The Marqo team is proud to introduce 0.1.0!
Marqo now includes various timing metrics for the search,
add_documents endpoints when the query parameter
telemetry=True is specified. The metrics will be returned in the response body and provide a breakdown of latencies for various stages of the API call.
By default, Marqo now uses CUDA devices for search and indexing if available. This helps ensure you get the best indexing and search experience without having to explicitly add the device parameter to search and
Model files are validated and removed if corrupted during download. This helps ensure that models are not loaded if they are corrupted.
Marqo will no longer automatically create an index if it does not exist. Attempting to add documents to a non-existent index will now result in an error. This helps provide more certainty about the properties of the index you are adding documents to, and also helps prevent accidental indexing to the wrong index.
Marqo no longer supports server-side parallel indexing. This helps deliver a more stable and efficient indexing experience. Parallelisation can still be implemented by the user.
Improve error messages
Improve API request validation
Add new multimodal search example
Remove autocast for CPU to speed up vectorisation on ARM64 machines
Enhance test stability
Ignore .kibana index
Improve handling of whitespace when indexing documents
Update CUDA version to 11.4.3
Thank you to our 3.1k stargazers!
See full release notes here: Release 0.1.0 · marqo-ai/marqo · GitHub