snowykami/meilisearch
1
0
Fork 0
mirror of https://github.com/meilisearch/meilisearch.git synced 2025-09-06 12:46:31 +00:00
Code Issues Packages Projects Releases Wiki Activity

Use Hannoy instead of arroy

Browse Source
This commit is contained in:
Kerollmops
2025-07-21 11:42:46 +02:00
parent 6f8c414a75
commit 94c1e458bd
25 changed files with 380 additions and 356 deletions
Download Patch File Download Diff File Expand all files Collapse all files

28
crates/milli/src/index.rs
View File

@ -31,7 +31,7 @@ use crate::prompt::PromptData;
use crate::proximity::ProximityPrecision;
use crate::update::new::StdResult;
use crate::vector::db::IndexEmbeddingConfigs;
use crate::vector::{ArroyStats, ArroyWrapper, Embedding};
use crate::vector::{Embedding, HannoyStats, HannoyWrapper};
use crate::{
default_criteria, CboRoaringBitmapCodec, Criterion, DocumentId, ExternalDocumentsIds,
FacetDistribution, FieldDistribution, FieldId, FieldIdMapMissingEntry, FieldIdWordCountCodec,
@ -113,7 +113,7 @@ pub mod db_name {
pub const FIELD_ID_DOCID_FACET_F64S: &str = "field-id-docid-facet-f64s";
pub const FIELD_ID_DOCID_FACET_STRINGS: &str = "field-id-docid-facet-strings";
pub const VECTOR_EMBEDDER_CATEGORY_ID: &str = "vector-embedder-category-id";
pub const VECTOR_ARROY: &str = "vector-arroy";
pub const VECTOR_HANNOY: &str = "vector-hannoy";
pub const DOCUMENTS: &str = "documents";
}
const NUMBER_OF_DBS: u32 = 25;
@ -177,10 +177,10 @@ pub struct Index {
/// Maps the document id, the facet field id and the strings.
pub field_id_docid_facet_strings: Database<FieldDocIdFacetStringCodec, Str>,
/// Maps an embedder name to its id in the arroy store.
/// Maps an embedder name to its id in the hannoy store.
pub(crate) embedder_category_id: Database<Unspecified, Unspecified>,
/// Vector store based on arroy™.
pub vector_arroy: arroy::Database<Unspecified>,
/// Vector store based on hannoy™.
pub vector_hannoy: hannoy::Database<Unspecified>,
/// Maps the document id to the document as an obkv store.
pub(crate) documents: Database<BEU32, ObkvCodec>,
@ -237,7 +237,7 @@ impl Index {
// vector stuff
let embedder_category_id =
env.create_database(&mut wtxn, Some(VECTOR_EMBEDDER_CATEGORY_ID))?;
let vector_arroy = env.create_database(&mut wtxn, Some(VECTOR_ARROY))?;
let vector_hannoy = env.create_database(&mut wtxn, Some(VECTOR_HANNOY))?;
let documents = env.create_database(&mut wtxn, Some(DOCUMENTS))?;
@ -264,7 +264,7 @@ impl Index {
facet_id_is_empty_docids,
field_id_docid_facet_f64s,
field_id_docid_facet_strings,
vector_arroy,
vector_hannoy,
embedder_category_id,
documents,
};
@ -1772,8 +1772,8 @@ impl Index {
for config in embedders.embedding_configs(rtxn)? {
let embedder_info = embedders.embedder_info(rtxn, &config.name)?.unwrap();
let has_fragments = config.config.embedder_options.has_fragments();
let reader = ArroyWrapper::new(
self.vector_arroy,
let reader = HannoyWrapper::new(
self.vector_hannoy,
embedder_info.embedder_id,
config.config.quantized(),
);
@ -1792,13 +1792,13 @@ impl Index {
Ok(PrefixSettings { compute_prefixes, max_prefix_length: 4, prefix_count_threshold: 100 })
}
pub fn arroy_stats(&self, rtxn: &RoTxn<'_>) -> Result<ArroyStats> {
let mut stats = ArroyStats::default();
pub fn hannoy_stats(&self, rtxn: &RoTxn<'_>) -> Result<HannoyStats> {
let mut stats = HannoyStats::default();
let embedding_configs = self.embedding_configs();
for config in embedding_configs.embedding_configs(rtxn)? {
let embedder_id = embedding_configs.embedder_id(rtxn, &config.name)?.unwrap();
let reader =
ArroyWrapper::new(self.vector_arroy, embedder_id, config.config.quantized());
HannoyWrapper::new(self.vector_hannoy, embedder_id, config.config.quantized());
reader.aggregate_stats(rtxn, &mut stats)?;
}
Ok(stats)
@ -1842,7 +1842,7 @@ impl Index {
facet_id_is_empty_docids,
field_id_docid_facet_f64s,
field_id_docid_facet_strings,
vector_arroy,
vector_hannoy,
embedder_category_id,
documents,
} = self;
@ -1913,7 +1913,7 @@ impl Index {
"field_id_docid_facet_strings",
field_id_docid_facet_strings.stat(rtxn).map(compute_size)?,
);
sizes.insert("vector_arroy", vector_arroy.stat(rtxn).map(compute_size)?);
sizes.insert("vector_hannoy", vector_hannoy.stat(rtxn).map(compute_size)?);
sizes.insert("embedder_category_id", embedder_category_id.stat(rtxn).map(compute_size)?);
sizes.insert("documents", documents.stat(rtxn).map(compute_size)?);