Weaviate-client

--------------
This patch version includes:
- Fixes a rare bug when the batching algorithm waits for async indexing to complete far beyond the maximum number of allowed retries
- Removes the sync-in-async warning emitted by the sync client when used in an async context due to the warning being erroneously emitted in notebooks

--------------
This patch version includes:
- Adds exponential backoff on the UNAVAILABLE gRPC error code for the BatchObjects method
- This is used in the batching algorithm and in ``data.insert_many``
- Max number of retries is configurable using the ``WEAVIATE_BATCH_MAX_RETRIES`` env var
- Adds support for OIDC when using the ``text2vec-weaviate`` module
- Adds support for updating descriptions of properties in collections (needs compatible server version)
- Adds warning recommending to refactor to use AsyncClient if the SyncClient is used in an async context
- Fixes bugs when importing/exporting collection configurations with respect to property module configs
- Fixes bug when exporting config with multivector enabled between versions

--------------
This patch version incldues:
- Fixes exporting of named vector collections
- Support for upcoming weaviate agents. User ``pip install weaviate-client[agents]`` to try it out
- Raise error if a header with value `None`` has been added
- Introduce ``collections.use`` as an alias for ``collections.get``
- ``collections.use`` is the preferred method to create an in-memory collection object
- ``collections.get`` will be deprecated in the future and should not be used in new code

--------------

This minor version includes:
- Support for new 1.29 features:
- RBAC is ready for production
- Multi-dimensional/COLBert vectors experimental, breaking changes are expected)
- Aggregations are using GRPC now
- Improved exception handling for errors returned by Weaviate
- Add support for NVIDIA modules:
- multi2vec-nvidia
- text2vec-nvidia
- generative-nvidia

--------------
This patch version incldues:
- Fix missing dimensions parameter in text2vec-azure-openai

--------------
This patch version includes:
- Add experimental debug namespace
- Add support for dynamic backup location
- Fix potential deadlock in batching
- Fix backup and restore for embedded mode
- Add extra method for text2vec_google_aistudio