Rob Meng

Hi folks, I have a similar question to #476. Say that I would be recording thousands if not tens of thousands of processes. Concatenating the files might yield major size...

implement deletion vector handling in index scan

3

closes #916 Implement deletion vector handling in `PQIndex`, where all vector index are stored. TODO: - [ ] add tests - [ ] add TODO comments in places we could...

## Why We don't have an e2e index building benchmark, which is helpful for understanding latency and profiling for bottlenecks in indexing building. ## What changed * extract `create_file` to...

I observed that benchmarks have high variability in latency measurement across runs. (10+%) Increasing the number of samples didn't seem to reduce the variability. I tried to seed testing data...

closes #912 `object_store` has a custom S3 client which does not accept `CredentialProviders` from `aws_config`. This PR implements an adaptor that adapts a `DefaultCredentialsChain` to a `object_store::CredentialProvider`. - [ ]...

#899 enabled `regexp_match`, but if we call with `regexp_match(col, pattern)` where `col` is of type `LargeUtf8` the code panics. a few notes: 1. `regexp_match` on `LargeUtf8` is probably highly inefficient,...

implement sse based argmin

4

## Description WIP we got 5% speed up in sift1m with this single callsite. TODO: convert each argmin callsites to SSE and check benchmark AVX is too wide and isn't...

https://github.com/lancedb/lance/actions/runs/8469231469/job/23204096747

it is possible to get llvm coverage when running python tests. Here is an example https://github.com/cjermain/rust-python-coverage

we have a blanket conversion of `object_store::Error` -> `Error::IO`.