snowykami/meilisearch
1
0
Fork 0
mirror of https://github.com/meilisearch/meilisearch.git synced 2025-09-26 14:56:29 +00:00
Code Issues Packages Projects Releases Wiki Activity

Update arroy-hannoy conversion internals

Browse Source
This commit is contained in:
nnethercott
2025-09-15 16:10:56 +02:00
parent 06b3ca9eb5
commit 5bef2f4d86
1 changed files with 10 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
crates/milli/src/vector/store.rs
View File

@ -1046,6 +1046,10 @@ impl VectorStore {
where where
R: rand::Rng + rand::SeedableRng, R: rand::Rng + rand::SeedableRng,
{ {
// No work if distances are the same
if AD::name() == HD::name() {
return Ok(());
}
for index in vector_store_range_for_embedder(self.embedder_index) { for index in vector_store_range_for_embedder(self.embedder_index) {
let arroy_reader: arroy::Reader<AD> = let arroy_reader: arroy::Reader<AD> =
match arroy::Reader::open(arroy_rtxn, index, self.database.remap_types()) { match arroy::Reader::open(arroy_rtxn, index, self.database.remap_types()) {
@ -1084,6 +1088,10 @@ impl VectorStore {
where where
R: rand::Rng + rand::SeedableRng, R: rand::Rng + rand::SeedableRng,
{ {
// No work if distances are the same
if AD::name() == HD::name() {
return Ok(());
}
for index in vector_store_range_for_embedder(self.embedder_index) { for index in vector_store_range_for_embedder(self.embedder_index) {
let hannoy_reader: hannoy::Reader<HD> = let hannoy_reader: hannoy::Reader<HD> =
match hannoy::Reader::open(hannoy_rtxn, index, self.database.remap_types()) { match hannoy::Reader::open(hannoy_rtxn, index, self.database.remap_types()) {
@ -1098,12 +1106,8 @@ impl VectorStore {
arroy_writer.clear(arroy_wtxn)?; arroy_writer.clear(arroy_wtxn)?;
for entry in hannoy_reader.iter(hannoy_rtxn)? { for entry in hannoy_reader.iter(hannoy_rtxn)? {
let (item, mut vector) = entry?; let (item, mut vector) = entry?;
// hannoy bug? the `vector` here can be longer than `dimensions`. debug_assert!(vector.len() == dimensions);
// workaround: truncating. // arroy and hannoy disagreement over the 0 value if distance is Hamming
if vector.len() > dimensions {
vector.truncate(dimensions);
}
// arroy and hannoy disagreement over the 0 value
// - arroy does: // - arroy does:
// - if x >= 0 => 1 // - if x >= 0 => 1
// - if x < 0 => -1 // - if x < 0 => -1