Julie Tibshirani
Julie Tibshirani
Would it be possible to report vector search recall in addition to QPS? When assessing vector search, users are interested in both computational performance and accuracy. Tracking recall through nightly...
In addition to word embedding models like GloVe, there are now text embedding models like BERT and Universal Sentence Encoder that work at the level of sentences. These embedding models...
After C++ training completes, we serialize the contents of the forest to a byte stream, and pass this up to R. This serialized forest is then passed back to subsequent...
We tell users not to force merge unless their index is read-only. This PR proposes to soften the warning and make it more precise. This way, more users can consider...
This 'how to' guide explains performance considerations specific to kNN search. It takes inspiration from the 'tune for search speed' guide.
### Description HNSW search is most efficient when all vector data fits in page cache. So good to keep the size of vector files as small as possible. We currently...
After searching with `IndexHNSWFlat`, the reported number of distance computations `hnsw_stats.ndis` can remain 0. Looking through the code, it doesn't seem like `ndis` is being updated. It wonder if it's...
This PR adds a new heuristic for C++ that checks for the preprocessor macros __has_cpp_attribute and __cplusplus. These are commonly-used macros that check what version of C++ is supported, and...
On dot com, we see occasional panics in the frontend search jobs code from deeply nested regular expressions. We should be catching this sooner and returning a normal error to...
We've heard from a few customers that they'd to be able to promote usage of certain repositories which are known to be high quality and promote best practices. There may...