Add sparse float vector support

[zhengbuqian]

Milvus 2.4 has added beta support to sparse float vector to allow efficient storing, indexing and searching of such vectors. The main targeted use case is to support learned sparse models such as SPLADE and statistics based methods such as BM25.

We need to add sparse float vector to C# SDK as well. See: milvus-io/milvus#29419 (comment) for more details.

Thanks a lot for the efforts!

[roji]

Thanks, absolutely! I've created a 2.4.0 milestone to start tracking that work.

Are you interested in working on adding this support?

[zhengbuqian]

Hi Roji, I don't have any experiences working with c#, so I guess I won't be able to work on this. But this is not urgent(at least in the coming week) as we plan to provide sparse as a GA feature in the next major release, not in 2.4. PyMilvus is sufficient for 2.4.

[roji]

OK. I've just submitted #72 to bump the SDK version to 2.4 and start running tests against Milvus 2.4. I'll be looking at implementing 2.4 features at some point, if nobody else is available to work on them.

@zhengbuqian please feel free to open additional issues for 2.4 features, so that we track support for them!