Merge f8b7a48b1a57df51b48eee6a6fea798952c00df2 into c63c25a9a28d32ba74de7c00747dbbd1b17008ec

5355: Support fetching the pooling method from the model configuration r=Kerollmops a=dureuill

# Pull Request

## Related issue
Fixes #5354 

## What does this PR do?
- Fetches the pooling configuration from the model repository
- Use a pooling method that depends on the pooling configuration of that model.
- Allow overriding the pooling method with a new huggingFace embedder parameter `pooling`
  - for backward-compatibility with Meilisearch v1.13
  - for compatibility with embedders that exhibit the same behavior as Meilisearch v1.13
- Handle the default value of that new parameter
   - for compatibility, when importing a db/a dump, it should be set to `forceMean`
   - when (re)set from the settings for an embedder, it should be set to `useModel`


Co-authored-by: Louis Dureuil <louis@meilisearch.com>

.github/workflows/flaky-tests.yml

@@ -9,22 +9,22 @@ jobs:
   flaky:
     runs-on: ubuntu-latest
     container:
-      # Use ubuntu-20.04 to compile with glibc 2.28
+      # Use ubuntu-22.04 to compile with glibc 2.35
-      image: ubuntu:20.04
+      image: ubuntu:22.04
     steps:
-    - uses: actions/checkout@v3
+      - uses: actions/checkout@v3
-    - name: Install needed dependencies
+      - name: Install needed dependencies
-      run: |
+        run: |
-        apt-get update && apt-get install -y curl
+          apt-get update && apt-get install -y curl
-        apt-get install build-essential -y
+          apt-get install build-essential -y
-    - uses: dtolnay/rust-toolchain@1.81
+      - uses: dtolnay/rust-toolchain@1.81
-    - name: Install cargo-flaky
+      - name: Install cargo-flaky
-      run: cargo install cargo-flaky
+        run: cargo install cargo-flaky
-    - name: Run cargo flaky in the dumps
+      - name: Run cargo flaky in the dumps
-      run: cd crates/dump; cargo flaky -i 100 --release
+        run: cd crates/dump; cargo flaky -i 100 --release
-    - name: Run cargo flaky in the index-scheduler
+      - name: Run cargo flaky in the index-scheduler
-      run: cd crates/index-scheduler; cargo flaky -i 100 --release
+        run: cd crates/index-scheduler; cargo flaky -i 100 --release
-    - name: Run cargo flaky in the auth
+      - name: Run cargo flaky in the auth
-      run: cd crates/meilisearch-auth; cargo flaky -i 100 --release
+        run: cd crates/meilisearch-auth; cargo flaky -i 100 --release
-    - name: Run cargo flaky in meilisearch
+      - name: Run cargo flaky in meilisearch
-      run: cd crates/meilisearch; cargo flaky -i 100 --release
+        run: cd crates/meilisearch; cargo flaky -i 100 --release

.github/workflows/publish-apt-brew-pkg.yml

@@ -18,28 +18,28 @@ jobs:
     runs-on: ubuntu-latest
     needs: check-version
     container:
-      # Use ubuntu-20.04 to compile with glibc 2.28
+      # Use ubuntu-22.04 to compile with glibc 2.35
-      image: ubuntu:20.04
+      image: ubuntu:22.04
     steps:
-    - name: Install needed dependencies
+      - name: Install needed dependencies
-      run: |
+        run: |
-        apt-get update && apt-get install -y curl
+          apt-get update && apt-get install -y curl
-        apt-get install build-essential -y
+          apt-get install build-essential -y
-    - uses: dtolnay/rust-toolchain@1.81
+      - uses: dtolnay/rust-toolchain@1.81
-    - name: Install cargo-deb
+      - name: Install cargo-deb
-      run: cargo install cargo-deb
+        run: cargo install cargo-deb
-    - uses: actions/checkout@v3
+      - uses: actions/checkout@v3
-    - name: Build deb package
+      - name: Build deb package
-      run: cargo deb -p meilisearch -o target/debian/meilisearch.deb
+        run: cargo deb -p meilisearch -o target/debian/meilisearch.deb
-    - name: Upload debian pkg to release
+      - name: Upload debian pkg to release
-      uses: svenstaro/upload-release-action@2.7.0
+        uses: svenstaro/upload-release-action@2.7.0
-      with:
+        with:
-        repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
+          repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
-        file: target/debian/meilisearch.deb
+          file: target/debian/meilisearch.deb
-        asset_name: meilisearch.deb
+          asset_name: meilisearch.deb
-        tag: ${{ github.ref }}
+          tag: ${{ github.ref }}
-    - name: Upload debian pkg to apt repository
+      - name: Upload debian pkg to apt repository
-      run: curl -F package=@target/debian/meilisearch.deb https://${{ secrets.GEMFURY_PUSH_TOKEN }}@push.fury.io/meilisearch/
+        run: curl -F package=@target/debian/meilisearch.deb https://${{ secrets.GEMFURY_PUSH_TOKEN }}@push.fury.io/meilisearch/
 
   homebrew:
     name: Bump Homebrew formula

.github/workflows/publish-binaries.yml

@@ -3,7 +3,7 @@ name: Publish binaries to GitHub release
 on:
   workflow_dispatch:
   schedule:
-    - cron: '0 2 * * *' # Every day at 2:00am
+    - cron: "0 2 * * *" # Every day at 2:00am
   release:
     types: [published]
 
@@ -37,26 +37,26 @@ jobs:
     runs-on: ubuntu-latest
     needs: check-version
     container:
-      # Use ubuntu-20.04 to compile with glibc 2.28
+      # Use ubuntu-22.04 to compile with glibc 2.35
-      image: ubuntu:20.04
+      image: ubuntu:22.04
     steps:
-    - uses: actions/checkout@v3
+      - uses: actions/checkout@v3
-    - name: Install needed dependencies
+      - name: Install needed dependencies
-      run: |
+        run: |
-        apt-get update && apt-get install -y curl
+          apt-get update && apt-get install -y curl
-        apt-get install build-essential -y
+          apt-get install build-essential -y
-    - uses: dtolnay/rust-toolchain@1.81
+      - uses: dtolnay/rust-toolchain@1.81
-    - name: Build
+      - name: Build
-      run: cargo build --release --locked
+        run: cargo build --release --locked
-    # No need to upload binaries for dry run (cron)
+      # No need to upload binaries for dry run (cron)
-    - name: Upload binaries to release
+      - name: Upload binaries to release
-      if: github.event_name == 'release'
+        if: github.event_name == 'release'
-      uses: svenstaro/upload-release-action@2.7.0
+        uses: svenstaro/upload-release-action@2.7.0
-      with:
+        with:
-        repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
+          repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
-        file: target/release/meilisearch
+          file: target/release/meilisearch
-        asset_name: meilisearch-linux-amd64
+          asset_name: meilisearch-linux-amd64
-        tag: ${{ github.ref }}
+          tag: ${{ github.ref }}
 
   publish-macos-windows:
     name: Publish binary for ${{ matrix.os }}
@@ -74,19 +74,19 @@ jobs:
             artifact_name: meilisearch.exe
             asset_name: meilisearch-windows-amd64.exe
     steps:
-    - uses: actions/checkout@v3
+      - uses: actions/checkout@v3
-    - uses: dtolnay/rust-toolchain@1.81
+      - uses: dtolnay/rust-toolchain@1.81
-    - name: Build
+      - name: Build
-      run: cargo build --release --locked
+        run: cargo build --release --locked
-    # No need to upload binaries for dry run (cron)
+      # No need to upload binaries for dry run (cron)
-    - name: Upload binaries to release
+      - name: Upload binaries to release
-      if: github.event_name == 'release'
+        if: github.event_name == 'release'
-      uses: svenstaro/upload-release-action@2.7.0
+        uses: svenstaro/upload-release-action@2.7.0
-      with:
+        with:
-        repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
+          repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
-        file: target/release/${{ matrix.artifact_name }}
+          file: target/release/${{ matrix.artifact_name }}
-        asset_name: ${{ matrix.asset_name }}
+          asset_name: ${{ matrix.asset_name }}
-        tag: ${{ github.ref }}
+          tag: ${{ github.ref }}
 
   publish-macos-apple-silicon:
     name: Publish binary for macOS silicon
@@ -127,8 +127,8 @@ jobs:
     env:
       DEBIAN_FRONTEND: noninteractive
     container:
-      # Use ubuntu-20.04 to compile with glibc 2.28
+      # Use ubuntu-22.04 to compile with glibc 2.35
-      image: ubuntu:20.04
+      image: ubuntu:22.04
     strategy:
       matrix:
         include:

.github/workflows/test-suite.yml

@@ -19,11 +19,11 @@ env:
 
 jobs:
   test-linux:
-    name: Tests on ubuntu-20.04
+    name: Tests on ubuntu-22.04
     runs-on: ubuntu-latest
     container:
-      # Use ubuntu-20.04 to compile with glibc 2.28
+      # Use ubuntu-22.04 to compile with glibc 2.35
-      image: ubuntu:20.04
+      image: ubuntu:22.04
     steps:
       - uses: actions/checkout@v3
       - name: Install needed dependencies
@@ -72,8 +72,8 @@ jobs:
     name: Tests almost all features
     runs-on: ubuntu-latest
     container:
-      # Use ubuntu-20.04 to compile with glibc 2.28
+      # Use ubuntu-22.04 to compile with glibc 2.35
-      image: ubuntu:20.04
+      image: ubuntu:22.04
     if: github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'
     steps:
       - uses: actions/checkout@v3
@@ -125,7 +125,7 @@ jobs:
     name: Test disabled tokenization
     runs-on: ubuntu-latest
     container:
-      image: ubuntu:20.04
+      image: ubuntu:22.04
     if: github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'
     steps:
       - uses: actions/checkout@v3
@@ -149,8 +149,8 @@ jobs:
     name: Run tests in debug
     runs-on: ubuntu-latest
     container:
-      # Use ubuntu-20.04 to compile with glibc 2.28
+      # Use ubuntu-22.04 to compile with glibc 2.35
-      image: ubuntu:20.04
+      image: ubuntu:22.04
     steps:
       - uses: actions/checkout@v3
       - name: Install needed dependencies

Cargo.lock

@@ -503,7 +503,7 @@ source = "git+https://github.com/meilisearch/bbqueue#cbb87cc707b5af415ef203bdaf2
 
 [[package]]
 name = "benchmarks"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "anyhow",
  "bumpalo",
@@ -694,7 +694,7 @@ dependencies = [
 
 [[package]]
 name = "build-info"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "anyhow",
  "time",
@@ -1671,7 +1671,7 @@ dependencies = [
 
 [[package]]
 name = "dump"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "anyhow",
  "big_s",
@@ -1873,7 +1873,7 @@ checksum = "37909eebbb50d72f9059c3b6d82c0463f2ff062c9e95845c43a6c9c0355411be"
 
 [[package]]
 name = "file-store"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "tempfile",
  "thiserror 2.0.9",
@@ -1895,7 +1895,7 @@ dependencies = [
 
 [[package]]
 name = "filter-parser"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "insta",
  "nom",
@@ -1915,7 +1915,7 @@ dependencies = [
 
 [[package]]
 name = "flatten-serde-json"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "criterion",
  "serde_json",
@@ -2054,7 +2054,7 @@ dependencies = [
 
 [[package]]
 name = "fuzzers"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "arbitrary",
  "bumpalo",
@@ -2743,7 +2743,7 @@ checksum = "206ca75c9c03ba3d4ace2460e57b189f39f43de612c2f85836e65c929701bb2d"
 
 [[package]]
 name = "index-scheduler"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "anyhow",
  "arroy 0.5.0 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -2950,7 +2950,7 @@ dependencies = [
 
 [[package]]
 name = "json-depth-checker"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "criterion",
  "serde_json",
@@ -3569,7 +3569,7 @@ checksum = "490cc448043f947bae3cbee9c203358d62dbee0db12107a74be5c30ccfd09771"
 
 [[package]]
 name = "meili-snap"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "insta",
  "md5",
@@ -3578,7 +3578,7 @@ dependencies = [
 
 [[package]]
 name = "meilisearch"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "actix-cors",
  "actix-http",
@@ -3670,7 +3670,7 @@ dependencies = [
 
 [[package]]
 name = "meilisearch-auth"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "base64 0.22.1",
  "enum-iterator",
@@ -3689,7 +3689,7 @@ dependencies = [
 
 [[package]]
 name = "meilisearch-types"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "actix-web",
  "anyhow",
@@ -3723,7 +3723,7 @@ dependencies = [
 
 [[package]]
 name = "meilitool"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "anyhow",
  "arroy 0.5.0 (git+https://github.com/meilisearch/arroy/?tag=DO-NOT-DELETE-upgrade-v04-to-v05)",
@@ -3758,7 +3758,7 @@ dependencies = [
 
 [[package]]
 name = "milli"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "allocator-api2",
  "arroy 0.5.0 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -4270,7 +4270,7 @@ checksum = "e3148f5046208a5d56bcfc03053e3ca6334e51da8dfb19b6cdc8b306fae3283e"
 
 [[package]]
 name = "permissive-json-pointer"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "big_s",
  "serde_json",
@@ -6847,7 +6847,7 @@ dependencies = [
 
 [[package]]
 name = "xtask"
-version = "1.13.0"
+version = "1.13.1"
 dependencies = [
  "anyhow",
  "build-info",

Cargo.toml

@@ -22,7 +22,7 @@ members = [
 ]
 
 [workspace.package]
-version = "1.13.0"
+version = "1.13.1"
 authors = [
     "Quentin de Quelen <quentin@dequelen.me>",
     "Clément Renault <clement@meilisearch.com>",
@@ -35,6 +35,7 @@ license = "MIT"
 
 [profile.release]
 codegen-units = 1
+lto = true
 
 [profile.dev.package.flate2]
 opt-level = 3

bors.toml

@@ -1,5 +1,5 @@
 status = [
-    'Tests on ubuntu-20.04',
+    'Tests on ubuntu-22.04',
     'Tests on macos-13',
     'Tests on windows-2022',
     'Run Clippy',

crates/dump/src/reader/snapshots/dump__reader__test__import_dump_v6_with_vectors-5.snap

@@ -1,5 +1,5 @@
 ---
-source: dump/src/reader/mod.rs
+source: crates/dump/src/reader/mod.rs
 expression: vector_index.settings().unwrap()
 ---
 {
@@ -49,6 +49,7 @@ expression: vector_index.settings().unwrap()
       "source": "huggingFace",
       "model": "BAAI/bge-base-en-v1.5",
       "revision": "617ca489d9e86b49b8167676d8220688b99db36e",
+      "pooling": "forceMean",
       "documentTemplate": "{% for field in fields %} {{ field.name }}: {{ field.value }}\n{% endfor %}"
     }
   },

crates/dump/src/reader/v6/mod.rs

@@ -3,6 +3,7 @@ use std::io::{BufRead, BufReader, ErrorKind};
 use std::path::Path;
 
 pub use meilisearch_types::milli;
+use meilisearch_types::milli::vector::hf::OverridePooling;
 use tempfile::TempDir;
 use time::OffsetDateTime;
 use tracing::debug;
@@ -252,7 +253,29 @@ impl V6IndexReader {
     }
 
     pub fn settings(&mut self) -> Result<Settings<Checked>> {
-        let settings: Settings<Unchecked> = serde_json::from_reader(&mut self.settings)?;
+        let mut settings: Settings<Unchecked> = serde_json::from_reader(&mut self.settings)?;
+        patch_embedders(&mut settings);
         Ok(settings.check())
     }
 }
+
+fn patch_embedders(settings: &mut Settings<Unchecked>) {
+    if let Setting::Set(embedders) = &mut settings.embedders {
+        for settings in embedders.values_mut() {
+            let Setting::Set(settings) = &mut settings.inner else {
+                continue;
+            };
+            if settings.source != Setting::Set(milli::vector::settings::EmbedderSource::HuggingFace)
+            {
+                continue;
+            }
+            settings.pooling = match settings.pooling {
+                Setting::Set(pooling) => Setting::Set(pooling),
+                // if the pooling for a hugging face embedder is not set, force it to `forceMean`
+                // for backward compatibility with v1.13
+                // dumps created in v1.14 and up will have the setting set for hugging face embedders
+                Setting::Reset | Setting::NotSet => Setting::Set(OverridePooling::ForceMean),
+            };
+        }
+    }
+}

crates/index-scheduler/src/index_mapper/mod.rs

@@ -6,6 +6,7 @@ use std::{fs, thread};
 use meilisearch_types::heed::types::{SerdeJson, Str};
 use meilisearch_types::heed::{Database, Env, RoTxn, RwTxn};
 use meilisearch_types::milli;
+use meilisearch_types::milli::database_stats::DatabaseStats;
 use meilisearch_types::milli::update::IndexerConfig;
 use meilisearch_types::milli::{FieldDistribution, Index};
 use serde::{Deserialize, Serialize};
@@ -98,8 +99,9 @@ pub enum IndexStatus {
 /// The statistics that can be computed from an `Index` object.
 #[derive(Serialize, Deserialize, Debug)]
 pub struct IndexStats {
-    /// Number of documents in the index.
+    /// Stats of the documents database.
-    pub number_of_documents: u64,
+    #[serde(default)]
+    pub documents_database_stats: DatabaseStats,
     /// Size taken up by the index' DB, in bytes.
     ///
     /// This includes the size taken by both the used and free pages of the DB, and as the free pages
@@ -138,9 +140,9 @@ impl IndexStats {
     pub fn new(index: &Index, rtxn: &RoTxn) -> milli::Result<Self> {
         let arroy_stats = index.arroy_stats(rtxn)?;
         Ok(IndexStats {
-            number_of_documents: index.number_of_documents(rtxn)?,
             number_of_embeddings: Some(arroy_stats.number_of_embeddings),
             number_of_embedded_documents: Some(arroy_stats.documents.len()),
+            documents_database_stats: index.documents_stats(rtxn)?.unwrap_or_default(),
             database_size: index.on_disk_size()?,
             used_database_size: index.used_size()?,
             primary_key: index.primary_key(rtxn)?.map(|s| s.to_string()),

crates/index-scheduler/src/insta_snapshot.rs

@@ -370,7 +370,8 @@ pub fn snapshot_index_mapper(rtxn: &RoTxn, mapper: &IndexMapper) -> String {
         let stats = mapper.stats_of(rtxn, &name).unwrap();
         s.push_str(&format!(
             "{name}: {{ number_of_documents: {}, field_distribution: {:?} }}\n",
-            stats.number_of_documents, stats.field_distribution
+            stats.documents_database_stats.number_of_entries(),
+            stats.field_distribution
         ));
     }
 

crates/index-scheduler/src/scheduler/snapshots/index_scheduler__scheduler__test_embedders__import_vectors-8.snap

@@ -1,12 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_embedders.rs
 expression: simple_hf_config.embedder_options
-snapshot_kind: text
 ---
 {
   "HuggingFace": {
     "model": "sentence-transformers/all-MiniLM-L6-v2",
     "revision": "e4ce9877abf3edfe10b0d82785e83bdcb973e22e",
-    "distribution": null
+    "distribution": null,
+    "pooling": "useModel"
   }
 }

crates/index-scheduler/src/scheduler/snapshots/test.rs/test_settings_update/after_registering_settings_task.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, status: enqueued, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:
 enqueued [0,]

crates/index-scheduler/src/scheduler/snapshots/test.rs/test_settings_update/settings_update_processed.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:
 enqueued []

crates/index-scheduler/src/scheduler/snapshots/test_embedders.rs/import_vectors/Intel to kefir succeeds.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_embedders.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { received_documents: 1, indexed_documents: Some(1) }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: UpdateDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}
 2 {uid: 2, batch_uid: 2, status: succeeded, details: { received_documents: 1, indexed_documents: Some(1) }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: None, method: UpdateDocuments, content_file: 00000000-0000-0000-0000-000000000001, documents_count: 1, allow_index_creation: true }}
 ----------------------------------------------------------------------

crates/index-scheduler/src/scheduler/snapshots/test_embedders.rs/import_vectors/Intel to kefir.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_embedders.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { received_documents: 1, indexed_documents: Some(1) }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: UpdateDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}
 2 {uid: 2, status: enqueued, details: { received_documents: 1, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: None, method: UpdateDocuments, content_file: 00000000-0000-0000-0000-000000000001, documents_count: 1, allow_index_creation: true }}
 ----------------------------------------------------------------------

crates/index-scheduler/src/scheduler/snapshots/test_embedders.rs/import_vectors/adding Intel succeeds.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_embedders.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { received_documents: 1, indexed_documents: Some(1) }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: UpdateDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:

crates/index-scheduler/src/scheduler/snapshots/test_embedders.rs/import_vectors/after adding Intel.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_embedders.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 1 {uid: 1, status: enqueued, details: { received_documents: 1, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: UpdateDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:

crates/index-scheduler/src/scheduler/snapshots/test_embedders.rs/import_vectors/after_registering_settings_task_vectors.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_embedders.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, status: enqueued, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:
 enqueued [0,]

crates/index-scheduler/src/scheduler/snapshots/test_embedders.rs/import_vectors/settings_update_processed_vectors.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_embedders.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:
 enqueued []

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/after_processing_everything.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 13, 1) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, batch_uid: 2, status: succeeded, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
 3 {uid: 3, batch_uid: 3, status: failed, error: ResponseError { code: 200, message: "Index `doggo` already exists.", error_code: "index_already_exists", error_type: "invalid_request", error_link: "https://docs.meilisearch.com/errors#index_already_exists" }, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
@@ -58,7 +57,7 @@ girafo: { number_of_documents: 0, field_distribution: {} }
 [timestamp] [4,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.0"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.1"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
 1 {uid: 1, details: {"primaryKey":"mouse"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"indexCreation":1},"indexUids":{"catto":1}}, }
 2 {uid: 2, details: {"primaryKey":"bone"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"indexCreation":1},"indexUids":{"doggo":1}}, }
 3 {uid: 3, details: {"primaryKey":"bone"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"indexCreation":1},"indexUids":{"doggo":1}}, }

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/register_automatic_upgrade_task.snap

@@ -7,7 +7,7 @@ snapshot_kind: text
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 13, 1) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 ----------------------------------------------------------------------
 ### Status:
 enqueued [0,]

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/registered_a_task_while_the_upgrade_task_is_enqueued.snap

@@ -7,7 +7,7 @@ snapshot_kind: text
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 13, 1) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 ----------------------------------------------------------------------
 ### Status:

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/upgrade_task_failed.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 13, 1) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 ----------------------------------------------------------------------
 ### Status:
@@ -38,7 +37,7 @@ catto [1,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.0"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.1"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/upgrade_task_failed_again.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 13, 1) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, status: enqueued, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
 ----------------------------------------------------------------------
@@ -41,7 +40,7 @@ doggo [2,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.0"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.1"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/upgrade_task_succeeded.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 13, 1) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, status: enqueued, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
 3 {uid: 3, status: enqueued, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
@@ -44,7 +43,7 @@ doggo [2,3,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.0"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.1"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

crates/index-scheduler/src/scheduler/test.rs

@@ -910,7 +910,11 @@ fn create_and_list_index() {
         [
           "kefir",
           {
-            "number_of_documents": 0,
+            "documents_database_stats": {
+              "numberOfEntries": 0,
+              "totalKeySize": 0,
+              "totalValueSize": 0
+            },
             "database_size": "[bytes]",
             "number_of_embeddings": 0,
             "number_of_embedded_documents": 0,

crates/index-scheduler/src/scheduler/test_embedders.rs

@@ -404,31 +404,32 @@ fn import_vectors_first_and_embedder_later() {
     // even though we specified the vector for the ID 3, it shouldn't be marked
     // as user provided since we explicitely marked it as NOT user provided.
     snapshot!(format!("{conf:#?}"), @r###"
-        [
+    [
-            IndexEmbeddingConfig {
+        IndexEmbeddingConfig {
-                name: "my_doggo_embedder",
+            name: "my_doggo_embedder",
-                config: EmbeddingConfig {
+            config: EmbeddingConfig {
-                    embedder_options: HuggingFace(
+                embedder_options: HuggingFace(
-                        EmbedderOptions {
+                    EmbedderOptions {
-                            model: "sentence-transformers/all-MiniLM-L6-v2",
+                        model: "sentence-transformers/all-MiniLM-L6-v2",
-                            revision: Some(
+                        revision: Some(
-                                "e4ce9877abf3edfe10b0d82785e83bdcb973e22e",
+                            "e4ce9877abf3edfe10b0d82785e83bdcb973e22e",
-                            ),
-                            distribution: None,
-                        },
-                    ),
-                    prompt: PromptData {
-                        template: "{{doc.doggo}}",
-                        max_bytes: Some(
-                            400,
                         ),
+                        distribution: None,
+                        pooling: UseModel,
                     },
-                    quantized: None,
+                ),
+                prompt: PromptData {
+                    template: "{{doc.doggo}}",
+                    max_bytes: Some(
+                        400,
+                    ),
                 },
-                user_provided: RoaringBitmap<[1, 2]>,
+                quantized: None,
             },
-        ]
+            user_provided: RoaringBitmap<[1, 2]>,
-        "###);
+        },
+    ]
+    "###);
     let docid = index.external_documents_ids.get(&rtxn, "0").unwrap().unwrap();
     let embeddings = index.embeddings(&rtxn, docid).unwrap();
     let embedding = &embeddings["my_doggo_embedder"];

crates/index-scheduler/src/upgrade/mod.rs

@@ -46,20 +46,19 @@ pub fn upgrade_index_scheduler(
         }
     };
 
-    let mut current_version = from;
-
     info!("Upgrading the task queue");
+    let mut local_from = from;
     for upgrade in upgrade_functions[start..].iter() {
         let target = upgrade.target_version();
         info!(
             "Upgrading from v{}.{}.{} to v{}.{}.{}",
-            from.0, from.1, from.2, current_version.0, current_version.1, current_version.2
+            local_from.0, local_from.1, local_from.2, target.0, target.1, target.2
         );
         let mut wtxn = env.write_txn()?;
-        upgrade.upgrade(env, &mut wtxn, from)?;
+        upgrade.upgrade(env, &mut wtxn, local_from)?;
         versioning.set_version(&mut wtxn, target)?;
         wtxn.commit()?;
-        current_version = target;
+        local_from = target;
     }
 
     let mut wtxn = env.write_txn()?;

crates/meilisearch/Cargo.toml

@@ -170,5 +170,5 @@ german = ["meilisearch-types/german"]
 turkish = ["meilisearch-types/turkish"]
 
 [package.metadata.mini-dashboard]
-assets-url = "https://github.com/meilisearch/mini-dashboard/releases/download/v0.2.16/build.zip"
+assets-url = "https://github.com/meilisearch/mini-dashboard/releases/download/v0.2.17/build.zip"
-sha1 = "68f83438a114aabbe76bc9fe480071e741996662"
+sha1 = "29e92ce25f306208a9c86f013279c736bdc1e034"

crates/meilisearch/src/routes/indexes/mod.rs

@@ -494,6 +494,10 @@ pub async fn delete_index(
 pub struct IndexStats {
     /// Number of documents in the index
     pub number_of_documents: u64,
+    /// Size of the documents database, in bytes.
+    pub raw_document_db_size: u64,
+    /// Average size of a document in the documents database.
+    pub avg_document_size: u64,
     /// Whether or not the index is currently ingesting document
     pub is_indexing: bool,
     /// Number of embeddings in the index
@@ -510,7 +514,9 @@ pub struct IndexStats {
 impl From<index_scheduler::IndexStats> for IndexStats {
     fn from(stats: index_scheduler::IndexStats) -> Self {
         IndexStats {
-            number_of_documents: stats.inner_stats.number_of_documents,
+            number_of_documents: stats.inner_stats.documents_database_stats.number_of_entries(),
+            raw_document_db_size: stats.inner_stats.documents_database_stats.total_value_size(),
+            avg_document_size: stats.inner_stats.documents_database_stats.average_value_size(),
             is_indexing: stats.is_indexing,
             number_of_embeddings: stats.inner_stats.number_of_embeddings,
             number_of_embedded_documents: stats.inner_stats.number_of_embedded_documents,
@@ -532,6 +538,8 @@ impl From<index_scheduler::IndexStats> for IndexStats {
         (status = OK, description = "The stats of the index", body = IndexStats, content_type = "application/json", example = json!(
             {
                 "numberOfDocuments": 10,
+                "rawDocumentDbSize": 10,
+                "avgDocumentSize": 10,
                 "numberOfEmbeddings": 10,
                 "numberOfEmbeddedDocuments": 10,
                 "isIndexing": true,

crates/meilisearch/src/routes/mod.rs

@@ -392,6 +392,9 @@ pub struct Stats {
                 "indexes": {
                     "movies": {
                         "numberOfDocuments": 10,
+                        "rawDocumentDbSize": 100,
+                        "maxDocumentSize": 16,
+                        "avgDocumentSize": 10,
                         "isIndexing": true,
                         "fieldDistribution": {
                             "genre": 10,

crates/meilisearch/tests/documents/delete_documents.rs

@@ -160,6 +160,8 @@ async fn delete_document_by_filter() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 4,
+      "rawDocumentDbSize": 42,
+      "avgDocumentSize": 10,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -209,6 +211,8 @@ async fn delete_document_by_filter() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 2,
+      "rawDocumentDbSize": 16,
+      "avgDocumentSize": 8,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -277,6 +281,8 @@ async fn delete_document_by_filter() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 1,
+      "rawDocumentDbSize": 12,
+      "avgDocumentSize": 12,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,

crates/meilisearch/tests/dumps/mod.rs

@@ -32,6 +32,8 @@ async fn import_dump_v1_movie_raw() {
       @r###"
     {
       "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -187,6 +189,8 @@ async fn import_dump_v1_movie_with_settings() {
         @r###"
     {
       "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -355,6 +359,8 @@ async fn import_dump_v1_rubygems_with_settings() {
       @r###"
     {
       "numberOfDocuments": 53,
+      "rawDocumentDbSize": 8606,
+      "avgDocumentSize": 162,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -520,6 +526,8 @@ async fn import_dump_v2_movie_raw() {
       @r###"
     {
       "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -675,6 +683,8 @@ async fn import_dump_v2_movie_with_settings() {
       @r###"
     {
       "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -840,6 +850,8 @@ async fn import_dump_v2_rubygems_with_settings() {
       @r###"
     {
       "numberOfDocuments": 53,
+      "rawDocumentDbSize": 8606,
+      "avgDocumentSize": 162,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -1002,6 +1014,8 @@ async fn import_dump_v3_movie_raw() {
       @r###"
     {
       "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -1157,6 +1171,8 @@ async fn import_dump_v3_movie_with_settings() {
       @r###"
     {
       "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -1322,6 +1338,8 @@ async fn import_dump_v3_rubygems_with_settings() {
       @r###"
     {
       "numberOfDocuments": 53,
+      "rawDocumentDbSize": 8606,
+      "avgDocumentSize": 162,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -1484,6 +1502,8 @@ async fn import_dump_v4_movie_raw() {
       @r###"
     {
       "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -1639,6 +1659,8 @@ async fn import_dump_v4_movie_with_settings() {
       @r###"
     {
       "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -1804,6 +1826,8 @@ async fn import_dump_v4_rubygems_with_settings() {
       @r###"
     {
       "numberOfDocuments": 53,
+      "rawDocumentDbSize": 8606,
+      "avgDocumentSize": 162,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -1973,6 +1997,8 @@ async fn import_dump_v5() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 10,
+      "rawDocumentDbSize": 6782,
+      "avgDocumentSize": 678,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -2009,6 +2035,8 @@ async fn import_dump_v5() {
       @r###"
     {
       "numberOfDocuments": 10,
+      "rawDocumentDbSize": 6782,
+      "avgDocumentSize": 678,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -2386,6 +2414,7 @@ async fn generate_and_import_dump_containing_vectors() {
           "source": "huggingFace",
           "model": "sentence-transformers/all-MiniLM-L6-v2",
           "revision": "e4ce9877abf3edfe10b0d82785e83bdcb973e22e",
+          "pooling": "useModel",
           "documentTemplate": "{{doc.doggo}}",
           "documentTemplateMaxBytes": 400
         }

crates/meilisearch/tests/search/mod.rs

@@ -128,6 +128,40 @@ async fn search_with_stop_word() {
           .await;
 }
 
+#[actix_rt::test]
+async fn search_with_typo_settings() {
+    // related to https://github.com/meilisearch/meilisearch/issues/5240
+    let server = Server::new().await;
+    let index = server.index("test");
+
+    let (_, code) = index
+        .update_settings(json!({"typoTolerance": { "disableOnAttributes": ["title", "id"]}}))
+        .await;
+    meili_snap::snapshot!(code, @"202 Accepted");
+
+    let documents = DOCUMENTS.clone();
+    let (task, _status_code) = index.add_documents(documents, None).await;
+    index.wait_task(task.uid()).await.succeeded();
+
+    index
+        .search(json!({"q": "287947" }), |response, code| {
+            assert_eq!(code, 200, "{}", response);
+            snapshot!(json_string!(response["hits"]), @r###"
+            [
+              {
+                "title": "Shazam!",
+                "id": "287947",
+                "color": [
+                  "green",
+                  "blue"
+                ]
+              }
+            ]
+            "###);
+        })
+        .await;
+}
+
 #[actix_rt::test]
 async fn phrase_search_with_stop_word() {
     // related to https://github.com/meilisearch/meilisearch/issues/3521

crates/meilisearch/tests/stats/mod.rs

@@ -113,6 +113,8 @@ async fn add_remove_embeddings() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 2,
+      "rawDocumentDbSize": 27,
+      "avgDocumentSize": 13,
       "isIndexing": false,
       "numberOfEmbeddings": 5,
       "numberOfEmbeddedDocuments": 2,
@@ -136,6 +138,8 @@ async fn add_remove_embeddings() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 2,
+      "rawDocumentDbSize": 27,
+      "avgDocumentSize": 13,
       "isIndexing": false,
       "numberOfEmbeddings": 3,
       "numberOfEmbeddedDocuments": 2,
@@ -159,6 +163,8 @@ async fn add_remove_embeddings() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 2,
+      "rawDocumentDbSize": 27,
+      "avgDocumentSize": 13,
       "isIndexing": false,
       "numberOfEmbeddings": 2,
       "numberOfEmbeddedDocuments": 2,
@@ -183,6 +189,8 @@ async fn add_remove_embeddings() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 2,
+      "rawDocumentDbSize": 27,
+      "avgDocumentSize": 13,
       "isIndexing": false,
       "numberOfEmbeddings": 2,
       "numberOfEmbeddedDocuments": 1,
@@ -231,6 +239,8 @@ async fn add_remove_embedded_documents() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 2,
+      "rawDocumentDbSize": 27,
+      "avgDocumentSize": 13,
       "isIndexing": false,
       "numberOfEmbeddings": 5,
       "numberOfEmbeddedDocuments": 2,
@@ -250,6 +260,8 @@ async fn add_remove_embedded_documents() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 1,
+      "rawDocumentDbSize": 13,
+      "avgDocumentSize": 13,
       "isIndexing": false,
       "numberOfEmbeddings": 3,
       "numberOfEmbeddedDocuments": 1,
@@ -281,6 +293,8 @@ async fn update_embedder_settings() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 2,
+      "rawDocumentDbSize": 108,
+      "avgDocumentSize": 54,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,
@@ -315,6 +329,8 @@ async fn update_embedder_settings() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 2,
+      "rawDocumentDbSize": 108,
+      "avgDocumentSize": 54,
       "isIndexing": false,
       "numberOfEmbeddings": 3,
       "numberOfEmbeddedDocuments": 2,

crates/meilisearch/tests/upgrade/mod.rs

@@ -43,7 +43,7 @@ async fn version_too_old() {
     std::fs::write(db_path.join("VERSION"), "1.11.9999").unwrap();
     let options = Opt { experimental_dumpless_upgrade: true, ..default_settings };
     let err = Server::new_with_options(options).await.map(|_| ()).unwrap_err();
-    snapshot!(err, @"Database version 1.11.9999 is too old for the experimental dumpless upgrade feature. Please generate a dump using the v1.11.9999 and import it in the v1.13.0");
+    snapshot!(err, @"Database version 1.11.9999 is too old for the experimental dumpless upgrade feature. Please generate a dump using the v1.11.9999 and import it in the v1.13.1");
 }
 
 #[actix_rt::test]
@@ -58,7 +58,7 @@ async fn version_requires_downgrade() {
     std::fs::write(db_path.join("VERSION"), format!("{major}.{minor}.{patch}")).unwrap();
     let options = Opt { experimental_dumpless_upgrade: true, ..default_settings };
     let err = Server::new_with_options(options).await.map(|_| ()).unwrap_err();
-    snapshot!(err, @"Database version 1.13.1 is higher than the Meilisearch version 1.13.0. Downgrade is not supported");
+    snapshot!(err, @"Database version 1.13.2 is higher than the Meilisearch version 1.13.1. Downgrade is not supported");
 }
 
 #[actix_rt::test]

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/batches_filter_afterEnqueuedAt_equal_2025-01-16T16_47_41.snap

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "progress": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.13.0"
+        "upgradeTo": "v1.13.1"
       },
       "stats": {
         "totalNbTasks": 1,

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/batches_filter_afterFinishedAt_equal_2025-01-16T16_47_41.snap

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "progress": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.13.0"
+        "upgradeTo": "v1.13.1"
       },
       "stats": {
         "totalNbTasks": 1,

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/batches_filter_afterStartedAt_equal_2025-01-16T16_47_41.snap

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "progress": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.13.0"
+        "upgradeTo": "v1.13.1"
       },
       "stats": {
         "totalNbTasks": 1,

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/tasks_filter_afterEnqueuedAt_equal_2025-01-16T16_47_41.snap

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "canceledBy": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.13.0"
+        "upgradeTo": "v1.13.1"
       },
       "error": null,
       "duration": "[duration]",

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/tasks_filter_afterFinishedAt_equal_2025-01-16T16_47_41.snap

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "canceledBy": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.13.0"
+        "upgradeTo": "v1.13.1"
       },
       "error": null,
       "duration": "[duration]",

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/tasks_filter_afterStartedAt_equal_2025-01-16T16_47_41.snap

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "canceledBy": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.13.0"
+        "upgradeTo": "v1.13.1"
       },
       "error": null,
       "duration": "[duration]",

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/the_whole_batch_queue_once_everything_has_been_processed.snap

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "progress": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.13.0"
+        "upgradeTo": "v1.13.1"
       },
       "stats": {
         "totalNbTasks": 1,

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/the_whole_task_queue_once_everything_has_been_processed.snap

@@ -1,6 +1,5 @@
 ---
 source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
-snapshot_kind: text
 ---
 {
   "results": [
@@ -13,7 +12,7 @@ snapshot_kind: text
       "canceledBy": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.13.0"
+        "upgradeTo": "v1.13.1"
       },
       "error": null,
       "duration": "[duration]",

crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs

@@ -139,6 +139,8 @@ async fn check_the_index_scheduler(server: &Server) {
       "indexes": {
         "kefir": {
           "numberOfDocuments": 1,
+          "rawDocumentDbSize": 109,
+          "avgDocumentSize": 109,
           "isIndexing": false,
           "numberOfEmbeddings": 0,
           "numberOfEmbeddedDocuments": 0,
@@ -225,6 +227,8 @@ async fn check_the_index_scheduler(server: &Server) {
       "indexes": {
         "kefir": {
           "numberOfDocuments": 1,
+          "rawDocumentDbSize": 109,
+          "avgDocumentSize": 109,
           "isIndexing": false,
           "numberOfEmbeddings": 0,
           "numberOfEmbeddedDocuments": 0,
@@ -244,6 +248,8 @@ async fn check_the_index_scheduler(server: &Server) {
     snapshot!(stats, @r###"
     {
       "numberOfDocuments": 1,
+      "rawDocumentDbSize": 109,
+      "avgDocumentSize": 109,
       "isIndexing": false,
       "numberOfEmbeddings": 0,
       "numberOfEmbeddedDocuments": 0,

crates/milli/src/database_stats.rs

@@ -0,0 +1,96 @@
+use heed::types::Bytes;
+use heed::Database;
+use heed::RoTxn;
+use serde::{Deserialize, Serialize};
+
+#[derive(Debug, Clone, Copy, Serialize, Deserialize, PartialEq, Eq, Default)]
+#[serde(rename_all = "camelCase")]
+/// The stats of a database.
+pub struct DatabaseStats {
+    /// The number of entries in the database.
+    number_of_entries: u64,
+    /// The total size of the keys in the database.
+    total_key_size: u64,
+    /// The total size of the values in the database.
+    total_value_size: u64,
+}
+
+impl DatabaseStats {
+    /// Returns the stats of the database.
+    ///
+    /// This function iterates over the whole database and computes the stats.
+    /// It is not efficient and should be cached somewhere.
+    pub(crate) fn new(database: Database<Bytes, Bytes>, rtxn: &RoTxn<'_>) -> heed::Result<Self> {
+        let mut database_stats =
+            Self { number_of_entries: 0, total_key_size: 0, total_value_size: 0 };
+
+        let mut iter = database.iter(rtxn)?;
+        while let Some((key, value)) = iter.next().transpose()? {
+            let key_size = key.len() as u64;
+            let value_size = value.len() as u64;
+            database_stats.total_key_size += key_size;
+            database_stats.total_value_size += value_size;
+        }
+
+        database_stats.number_of_entries = database.len(rtxn)?;
+
+        Ok(database_stats)
+    }
+
+    /// Recomputes the stats of the database and returns the new stats.
+    ///
+    /// This function is used to update the stats of the database when some keys are modified.
+    /// It is more efficient than the `new` function because it does not iterate over the whole database but only the modified keys comparing the before and after states.
+    pub(crate) fn recompute<I, K>(
+        mut stats: Self,
+        database: Database<Bytes, Bytes>,
+        before_rtxn: &RoTxn<'_>,
+        after_rtxn: &RoTxn<'_>,
+        modified_keys: I,
+    ) -> heed::Result<Self>
+    where
+        I: IntoIterator<Item = K>,
+        K: AsRef<[u8]>,
+    {
+        for key in modified_keys {
+            let key = key.as_ref();
+            if let Some(value) = database.get(after_rtxn, key)? {
+                let key_size = key.len() as u64;
+                let value_size = value.len() as u64;
+                stats.total_key_size = stats.total_key_size.saturating_add(key_size);
+                stats.total_value_size = stats.total_value_size.saturating_add(value_size);
+            }
+
+            if let Some(value) = database.get(before_rtxn, key)? {
+                let key_size = key.len() as u64;
+                let value_size = value.len() as u64;
+                stats.total_key_size = stats.total_key_size.saturating_sub(key_size);
+                stats.total_value_size = stats.total_value_size.saturating_sub(value_size);
+            }
+        }
+
+        stats.number_of_entries = database.len(after_rtxn)?;
+
+        Ok(stats)
+    }
+
+    pub fn average_key_size(&self) -> u64 {
+        self.total_key_size.checked_div(self.number_of_entries).unwrap_or(0)
+    }
+
+    pub fn average_value_size(&self) -> u64 {
+        self.total_value_size.checked_div(self.number_of_entries).unwrap_or(0)
+    }
+
+    pub fn number_of_entries(&self) -> u64 {
+        self.number_of_entries
+    }
+
+    pub fn total_key_size(&self) -> u64 {
+        self.total_key_size
+    }
+
+    pub fn total_value_size(&self) -> u64 {
+        self.total_value_size
+    }
+}

crates/milli/src/index.rs

@@ -11,6 +11,7 @@ use rstar::RTree;
 use serde::{Deserialize, Serialize};
 
 use crate::constants::{self, RESERVED_VECTORS_FIELD_NAME};
+use crate::database_stats::DatabaseStats;
 use crate::documents::PrimaryKey;
 use crate::error::{InternalError, UserError};
 use crate::fields_ids_map::FieldsIdsMap;
@@ -74,6 +75,7 @@ pub mod main_key {
     pub const LOCALIZED_ATTRIBUTES_RULES: &str = "localized_attributes_rules";
     pub const FACET_SEARCH: &str = "facet_search";
     pub const PREFIX_SEARCH: &str = "prefix_search";
+    pub const DOCUMENTS_STATS: &str = "documents_stats";
 }
 
 pub mod db_name {
@@ -403,6 +405,58 @@ impl Index {
         Ok(count.unwrap_or_default())
     }
 
+    /// Updates the stats of the documents database based on the previous stats and the modified docids.
+    pub fn update_documents_stats(
+        &self,
+        wtxn: &mut RwTxn<'_>,
+        modified_docids: roaring::RoaringBitmap,
+    ) -> Result<()> {
+        let before_rtxn = self.read_txn()?;
+        let document_stats = match self.documents_stats(&before_rtxn)? {
+            Some(before_stats) => DatabaseStats::recompute(
+                before_stats,
+                self.documents.remap_types(),
+                &before_rtxn,
+                wtxn,
+                modified_docids.iter().map(|docid| docid.to_be_bytes()),
+            )?,
+            None => {
+                // This should never happen when there are already documents in the index, the documents stats should be present.
+                // If it happens, it means that the index was not properly initialized/upgraded.
+                debug_assert_eq!(
+                    self.documents.len(&before_rtxn)?,
+                    0,
+                    "The documents stats should be present when there are documents in the index"
+                );
+                tracing::warn!("No documents stats found, creating new ones");
+                DatabaseStats::new(self.documents.remap_types(), &*wtxn)?
+            }
+        };
+
+        self.put_documents_stats(wtxn, document_stats)?;
+        Ok(())
+    }
+
+    /// Writes the stats of the documents database.
+    pub fn put_documents_stats(
+        &self,
+        wtxn: &mut RwTxn<'_>,
+        stats: DatabaseStats,
+    ) -> heed::Result<()> {
+        self.main.remap_types::<Str, SerdeJson<DatabaseStats>>().put(
+            wtxn,
+            main_key::DOCUMENTS_STATS,
+            &stats,
+        )
+    }
+
+    /// Returns the stats of the documents database.
+    pub fn documents_stats(&self, rtxn: &RoTxn<'_>) -> heed::Result<Option<DatabaseStats>> {
+        self.main
+            .remap_types::<Str, SerdeJson<DatabaseStats>>()
+            .get(rtxn, main_key::DOCUMENTS_STATS)
+    }
+
     /* primary key */
 
     /// Writes the documents primary key, this is the field name that is used to store the id.

crates/milli/src/lib.rs

@@ -10,6 +10,7 @@ pub mod documents;
 
 mod asc_desc;
 mod criterion;
+pub mod database_stats;
 mod error;
 mod external_documents_ids;
 pub mod facet;

crates/milli/src/search/new/query_term/compute_derivations.rs

@@ -215,7 +215,7 @@ pub fn partially_initialized_term_from_word(
     let mut zero_typo = None;
     let mut prefix_of = BTreeSet::new();
 
-    if fst.contains(word) {
+    if fst.contains(word) || ctx.index.exact_word_docids.get(ctx.txn, word)?.is_some() {
         zero_typo = Some(word_interned);
     }
 

crates/milli/src/update/index_documents/mod.rs

@@ -307,6 +307,7 @@ where
         let current_span = tracing::Span::current();
 
         // Run extraction pipeline in parallel.
+        let mut modified_docids = RoaringBitmap::new();
         pool.install(|| {
                 let settings_diff_cloned = settings_diff.clone();
                 rayon::spawn(move || {
@@ -367,7 +368,7 @@ where
                         Err(status) => {
                             if let Some(typed_chunks) = chunk_accumulator.pop_longest() {
                                 let (docids, is_merged_database) =
-                                    write_typed_chunk_into_index(self.wtxn, self.index, &settings_diff, typed_chunks)?;
+                                    write_typed_chunk_into_index(self.wtxn, self.index, &settings_diff, typed_chunks, &mut modified_docids)?;
                                 if !docids.is_empty() {
                                     final_documents_ids |= docids;
                                     let documents_seen_count = final_documents_ids.len();
@@ -467,6 +468,10 @@ where
                 Ok(())
             }).map_err(InternalError::from)??;
 
+        if !settings_diff.settings_update_only {
+            // Update the stats of the documents database when there is a document update.
+            self.index.update_documents_stats(self.wtxn, modified_docids)?;
+        }
         // We write the field distribution into the main database
         self.index.put_field_distribution(self.wtxn, &field_distribution)?;
 
@@ -2763,6 +2768,7 @@ mod tests {
                         source: Setting::Set(crate::vector::settings::EmbedderSource::UserProvided),
                         model: Setting::NotSet,
                         revision: Setting::NotSet,
+                        pooling: Setting::NotSet,
                         api_key: Setting::NotSet,
                         dimensions: Setting::Set(3),
                         document_template: Setting::NotSet,

crates/milli/src/update/index_documents/typed_chunk.rs

@@ -129,6 +129,7 @@ pub(crate) fn write_typed_chunk_into_index(
     index: &Index,
     settings_diff: &InnerIndexSettingsDiff,
     typed_chunks: Vec<TypedChunk>,
+    modified_docids: &mut RoaringBitmap,
 ) -> Result<(RoaringBitmap, bool)> {
     let mut is_merged_database = false;
     match typed_chunks[0] {
@@ -214,6 +215,7 @@ pub(crate) fn write_typed_chunk_into_index(
                         kind: DocumentOperationKind::Create,
                     });
                     docids.insert(docid);
+                    modified_docids.insert(docid);
                 } else {
                     db.delete(wtxn, &docid)?;
                     operations.push(DocumentOperation {
@@ -222,6 +224,7 @@ pub(crate) fn write_typed_chunk_into_index(
                         kind: DocumentOperationKind::Delete,
                     });
                     docids.remove(docid);
+                    modified_docids.insert(docid);
                 }
             }
             let external_documents_docids = index.external_documents_ids();

crates/milli/src/update/new/extract/cache.rs

@@ -711,15 +711,17 @@ impl DelAddRoaringBitmap {
         DelAddRoaringBitmap { del, add }
     }
 
-    pub fn apply_to(&self, documents_ids: &mut RoaringBitmap) {
+    pub fn apply_to(&self, documents_ids: &mut RoaringBitmap, modified_docids: &mut RoaringBitmap) {
         let DelAddRoaringBitmap { del, add } = self;
 
         if let Some(del) = del {
             *documents_ids -= del;
+            *modified_docids |= del;
         }
 
         if let Some(add) = add {
             *documents_ids |= add;
+            *modified_docids |= add;
         }
     }
 }

crates/milli/src/update/new/indexer/extract.rs

@@ -32,6 +32,7 @@ pub(super) fn extract_all<'pl, 'extractor, DC, MSP>(
     field_distribution: &mut BTreeMap<String, u64>,
     mut index_embeddings: Vec<IndexEmbeddingConfig>,
     document_ids: &mut RoaringBitmap,
+    modified_docids: &mut RoaringBitmap,
 ) -> Result<(FacetFieldIdsDelta, Vec<IndexEmbeddingConfig>)>
 where
     DC: DocumentChanges<'pl>,
@@ -70,7 +71,7 @@ where
                 // adding the delta should never cause a negative result, as we are removing fields that previously existed.
                 *current = current.saturating_add_signed(delta);
             }
-            document_extractor_data.docids_delta.apply_to(document_ids);
+            document_extractor_data.docids_delta.apply_to(document_ids, modified_docids);
         }
 
         field_distribution.retain(|_, v| *v != 0);
@@ -256,7 +257,7 @@ where
                     let Some(deladd) = data.remove(&config.name) else {
                         continue 'data;
                     };
-                    deladd.apply_to(&mut config.user_provided);
+                    deladd.apply_to(&mut config.user_provided, modified_docids);
                 }
             }
         }

crates/milli/src/update/new/indexer/mod.rs

@@ -130,6 +130,7 @@ where
     let index_embeddings = index.embedding_configs(wtxn)?;
     let mut field_distribution = index.field_distribution(wtxn)?;
     let mut document_ids = index.documents_ids(wtxn)?;
+    let mut modified_docids = roaring::RoaringBitmap::new();
 
     let congestion = thread::scope(|s| -> Result<ChannelCongestion> {
         let indexer_span = tracing::Span::current();
@@ -138,6 +139,7 @@ where
         // prevent moving the field_distribution and document_ids in the inner closure...
         let field_distribution = &mut field_distribution;
         let document_ids = &mut document_ids;
+        let modified_docids = &mut modified_docids;
         let extractor_handle =
             Builder::new().name(S("indexer-extractors")).spawn_scoped(s, move || {
                 pool.install(move || {
@@ -152,6 +154,7 @@ where
                         field_distribution,
                         index_embeddings,
                         document_ids,
+                        modified_docids,
                     )
                 })
                 .unwrap()
@@ -227,6 +230,7 @@ where
         embedders,
         field_distribution,
         document_ids,
+        modified_docids,
     )?;
 
     Ok(congestion)

crates/milli/src/update/new/indexer/write.rs

@@ -121,7 +121,8 @@ where
     Ok(())
 }
 
-pub fn update_index(
+#[allow(clippy::too_many_arguments)]
+pub(super) fn update_index(
     index: &Index,
     wtxn: &mut RwTxn<'_>,
     new_fields_ids_map: FieldIdMapWithMetadata,
@@ -129,6 +130,7 @@ pub fn update_index(
     embedders: EmbeddingConfigs,
     field_distribution: std::collections::BTreeMap<String, u64>,
     document_ids: roaring::RoaringBitmap,
+    modified_docids: roaring::RoaringBitmap,
 ) -> Result<()> {
     index.put_fields_ids_map(wtxn, new_fields_ids_map.as_fields_ids_map())?;
     if let Some(new_primary_key) = new_primary_key {
@@ -140,6 +142,7 @@ pub fn update_index(
     index.put_field_distribution(wtxn, &field_distribution)?;
     index.put_documents_ids(wtxn, &document_ids)?;
     index.set_updated_at(wtxn, &OffsetDateTime::now_utc())?;
+    index.update_documents_stats(wtxn, modified_docids)?;
     Ok(())
 }
 

crates/milli/src/update/settings.rs

@@ -1676,6 +1676,7 @@ fn validate_prompt(
             source,
             model,
             revision,
+            pooling,
             api_key,
             dimensions,
             document_template: Setting::Set(template),
@@ -1709,6 +1710,7 @@ fn validate_prompt(
                 source,
                 model,
                 revision,
+                pooling,
                 api_key,
                 dimensions,
                 document_template: Setting::Set(template),
@@ -1735,6 +1737,7 @@ pub fn validate_embedding_settings(
         source,
         model,
         revision,
+        pooling,
         api_key,
         dimensions,
         document_template,
@@ -1776,6 +1779,7 @@ pub fn validate_embedding_settings(
             source,
             model,
             revision,
+            pooling,
             api_key,
             dimensions,
             document_template,
@@ -1791,6 +1795,7 @@ pub fn validate_embedding_settings(
     match inferred_source {
         EmbedderSource::OpenAi => {
             check_unset(&revision, EmbeddingSettings::REVISION, inferred_source, name)?;
+            check_unset(&pooling, EmbeddingSettings::POOLING, inferred_source, name)?;
 
             check_unset(&request, EmbeddingSettings::REQUEST, inferred_source, name)?;
             check_unset(&response, EmbeddingSettings::RESPONSE, inferred_source, name)?;
@@ -1829,6 +1834,7 @@ pub fn validate_embedding_settings(
         EmbedderSource::Ollama => {
             check_set(&model, EmbeddingSettings::MODEL, inferred_source, name)?;
             check_unset(&revision, EmbeddingSettings::REVISION, inferred_source, name)?;
+            check_unset(&pooling, EmbeddingSettings::POOLING, inferred_source, name)?;
 
             check_unset(&request, EmbeddingSettings::REQUEST, inferred_source, name)?;
             check_unset(&response, EmbeddingSettings::RESPONSE, inferred_source, name)?;
@@ -1846,6 +1852,7 @@ pub fn validate_embedding_settings(
         EmbedderSource::UserProvided => {
             check_unset(&model, EmbeddingSettings::MODEL, inferred_source, name)?;
             check_unset(&revision, EmbeddingSettings::REVISION, inferred_source, name)?;
+            check_unset(&pooling, EmbeddingSettings::POOLING, inferred_source, name)?;
             check_unset(&api_key, EmbeddingSettings::API_KEY, inferred_source, name)?;
             check_unset(
                 &document_template,
@@ -1869,6 +1876,7 @@ pub fn validate_embedding_settings(
         EmbedderSource::Rest => {
             check_unset(&model, EmbeddingSettings::MODEL, inferred_source, name)?;
             check_unset(&revision, EmbeddingSettings::REVISION, inferred_source, name)?;
+            check_unset(&pooling, EmbeddingSettings::POOLING, inferred_source, name)?;
             check_set(&url, EmbeddingSettings::URL, inferred_source, name)?;
             check_set(&request, EmbeddingSettings::REQUEST, inferred_source, name)?;
             check_set(&response, EmbeddingSettings::RESPONSE, inferred_source, name)?;
@@ -1878,6 +1886,7 @@ pub fn validate_embedding_settings(
         source,
         model,
         revision,
+        pooling,
         api_key,
         dimensions,
         document_template,

crates/milli/src/update/upgrade/mod.rs

@@ -3,7 +3,7 @@ mod v1_13;
 
 use heed::RwTxn;
 use v1_12::{V1_12_3_To_V1_13_0, V1_12_To_V1_12_3};
-use v1_13::V1_13_0_To_Current;
+use v1_13::{V1_13_0_To_V1_13_1, V1_13_1_To_Current};
 
 use crate::progress::{Progress, VariableNameStep};
 use crate::{Index, InternalError, Result};
@@ -28,13 +28,18 @@ pub fn upgrade(
     progress: Progress,
 ) -> Result<bool> {
     let from = index.get_version(wtxn)?.unwrap_or(db_version);
-    let upgrade_functions: &[&dyn UpgradeIndex] =
+    let upgrade_functions: &[&dyn UpgradeIndex] = &[
-        &[&V1_12_To_V1_12_3 {}, &V1_12_3_To_V1_13_0 {}, &V1_13_0_To_Current()];
+        &V1_12_To_V1_12_3 {},
+        &V1_12_3_To_V1_13_0 {},
+        &V1_13_0_To_V1_13_1 {},
+        &V1_13_1_To_Current {},
+    ];
 
     let start = match from {
         (1, 12, 0..=2) => 0,
         (1, 12, 3..) => 1,
         (1, 13, 0) => 2,
+        (1, 13, 1) => 3,
         // We must handle the current version in the match because in case of a failure some index may have been upgraded but not other.
         (1, 13, _) => return Ok(false),
         (major, minor, patch) => {

crates/milli/src/update/upgrade/v1_13.rs

@@ -2,13 +2,44 @@ use heed::RwTxn;
 
 use super::UpgradeIndex;
 use crate::constants::{VERSION_MAJOR, VERSION_MINOR, VERSION_PATCH};
+use crate::database_stats::DatabaseStats;
 use crate::progress::Progress;
-use crate::{Index, Result};
+use crate::{make_enum_progress, Index, Result};
 
 #[allow(non_camel_case_types)]
-pub(super) struct V1_13_0_To_Current();
+pub(super) struct V1_13_0_To_V1_13_1();
 
-impl UpgradeIndex for V1_13_0_To_Current {
+impl UpgradeIndex for V1_13_0_To_V1_13_1 {
+    fn upgrade(
+        &self,
+        wtxn: &mut RwTxn,
+        index: &Index,
+        _original: (u32, u32, u32),
+        progress: Progress,
+    ) -> Result<bool> {
+        make_enum_progress! {
+            enum DocumentsStats {
+                CreatingDocumentsStats,
+            }
+        };
+
+        // Create the new documents stats.
+        progress.update_progress(DocumentsStats::CreatingDocumentsStats);
+        let stats = DatabaseStats::new(index.documents.remap_types(), wtxn)?;
+        index.put_documents_stats(wtxn, stats)?;
+
+        Ok(true)
+    }
+
+    fn target_version(&self) -> (u32, u32, u32) {
+        (1, 13, 1)
+    }
+}
+
+#[allow(non_camel_case_types)]
+pub(super) struct V1_13_1_To_Current();
+
+impl UpgradeIndex for V1_13_1_To_Current {
     fn upgrade(
         &self,
         _wtxn: &mut RwTxn,

crates/milli/src/vector/error.rs

@@ -262,6 +262,31 @@ impl NewEmbedderError {
         }
     }
 
+    pub fn open_pooling_config(
+        pooling_config_filename: PathBuf,
+        inner: std::io::Error,
+    ) -> NewEmbedderError {
+        let open_config = OpenPoolingConfig { filename: pooling_config_filename, inner };
+
+        Self {
+            kind: NewEmbedderErrorKind::OpenPoolingConfig(open_config),
+            fault: FaultSource::Runtime,
+        }
+    }
+
+    pub fn deserialize_pooling_config(
+        model_name: String,
+        pooling_config_filename: PathBuf,
+        inner: serde_json::Error,
+    ) -> NewEmbedderError {
+        let deserialize_pooling_config =
+            DeserializePoolingConfig { model_name, filename: pooling_config_filename, inner };
+        Self {
+            kind: NewEmbedderErrorKind::DeserializePoolingConfig(deserialize_pooling_config),
+            fault: FaultSource::Runtime,
+        }
+    }
+
     pub fn open_tokenizer(
         tokenizer_filename: PathBuf,
         inner: Box<dyn std::error::Error + Send + Sync>,
@@ -319,6 +344,13 @@ pub struct OpenConfig {
     pub inner: std::io::Error,
 }
 
+#[derive(Debug, thiserror::Error)]
+#[error("could not open pooling config at {filename}: {inner}")]
+pub struct OpenPoolingConfig {
+    pub filename: PathBuf,
+    pub inner: std::io::Error,
+}
+
 #[derive(Debug, thiserror::Error)]
 #[error("for model '{model_name}', could not deserialize config at {filename} as JSON: {inner}")]
 pub struct DeserializeConfig {
@@ -327,6 +359,14 @@ pub struct DeserializeConfig {
     pub inner: serde_json::Error,
 }
 
+#[derive(Debug, thiserror::Error)]
+#[error("for model '{model_name}', could not deserialize file at `{filename}` as a pooling config: {inner}")]
+pub struct DeserializePoolingConfig {
+    pub model_name: String,
+    pub filename: PathBuf,
+    pub inner: serde_json::Error,
+}
+
 #[derive(Debug, thiserror::Error)]
 #[error("model `{model_name}` appears to be unsupported{}\n  - inner error: {inner}",
 if architectures.is_empty() {
@@ -354,8 +394,12 @@ pub enum NewEmbedderErrorKind {
     #[error(transparent)]
     OpenConfig(OpenConfig),
     #[error(transparent)]
+    OpenPoolingConfig(OpenPoolingConfig),
+    #[error(transparent)]
     DeserializeConfig(DeserializeConfig),
     #[error(transparent)]
+    DeserializePoolingConfig(DeserializePoolingConfig),
+    #[error(transparent)]
     UnsupportedModel(UnsupportedModel),
     #[error(transparent)]
     OpenTokenizer(OpenTokenizer),

crates/milli/src/vector/hf.rs

@@ -34,6 +34,30 @@ pub struct EmbedderOptions {
     pub model: String,
     pub revision: Option<String>,
     pub distribution: Option<DistributionShift>,
+    #[serde(default)]
+    pub pooling: OverridePooling,
+}
+
+#[derive(
+    Debug,
+    Clone,
+    Copy,
+    Default,
+    Hash,
+    PartialEq,
+    Eq,
+    serde::Deserialize,
+    serde::Serialize,
+    utoipa::ToSchema,
+    deserr::Deserr,
+)]
+#[deserr(rename_all = camelCase, deny_unknown_fields)]
+#[serde(rename_all = "camelCase")]
+pub enum OverridePooling {
+    UseModel,
+    ForceCls,
+    #[default]
+    ForceMean,
 }
 
 impl EmbedderOptions {
@@ -42,6 +66,7 @@ impl EmbedderOptions {
             model: "BAAI/bge-base-en-v1.5".to_string(),
             revision: Some("617ca489d9e86b49b8167676d8220688b99db36e".into()),
             distribution: None,
+            pooling: OverridePooling::UseModel,
         }
     }
 }
@@ -58,6 +83,7 @@ pub struct Embedder {
     tokenizer: Tokenizer,
     options: EmbedderOptions,
     dimensions: usize,
+    pooling: Pooling,
 }
 
 impl std::fmt::Debug for Embedder {
@@ -66,10 +92,62 @@ impl std::fmt::Debug for Embedder {
             .field("model", &self.options.model)
             .field("tokenizer", &self.tokenizer)
             .field("options", &self.options)
+            .field("pooling", &self.pooling)
             .finish()
     }
 }
 
+#[derive(Clone, Copy, serde::Deserialize)]
+struct PoolingConfig {
+    #[serde(default)]
+    pub pooling_mode_cls_token: bool,
+    #[serde(default)]
+    pub pooling_mode_mean_tokens: bool,
+    #[serde(default)]
+    pub pooling_mode_max_tokens: bool,
+    #[serde(default)]
+    pub pooling_mode_mean_sqrt_len_tokens: bool,
+    #[serde(default)]
+    pub pooling_mode_lasttoken: bool,
+}
+
+#[derive(Debug, Clone, Copy, Default)]
+pub enum Pooling {
+    #[default]
+    Mean,
+    Cls,
+    Max,
+    MeanSqrtLen,
+    LastToken,
+}
+impl Pooling {
+    fn override_with(&mut self, pooling: OverridePooling) {
+        match pooling {
+            OverridePooling::UseModel => {}
+            OverridePooling::ForceCls => *self = Pooling::Cls,
+            OverridePooling::ForceMean => *self = Pooling::Mean,
+        }
+    }
+}
+
+impl From<PoolingConfig> for Pooling {
+    fn from(value: PoolingConfig) -> Self {
+        if value.pooling_mode_cls_token {
+            Self::Cls
+        } else if value.pooling_mode_mean_tokens {
+            Self::Mean
+        } else if value.pooling_mode_lasttoken {
+            Self::LastToken
+        } else if value.pooling_mode_mean_sqrt_len_tokens {
+            Self::MeanSqrtLen
+        } else if value.pooling_mode_max_tokens {
+            Self::Max
+        } else {
+            Self::default()
+        }
+    }
+}
+
 impl Embedder {
     pub fn new(options: EmbedderOptions) -> std::result::Result<Self, NewEmbedderError> {
         let device = match candle_core::Device::cuda_if_available(0) {
@@ -83,7 +161,7 @@ impl Embedder {
             Some(revision) => Repo::with_revision(options.model.clone(), RepoType::Model, revision),
             None => Repo::model(options.model.clone()),
         };
-        let (config_filename, tokenizer_filename, weights_filename, weight_source) = {
+        let (config_filename, tokenizer_filename, weights_filename, weight_source, pooling) = {
             let api = Api::new().map_err(NewEmbedderError::new_api_fail)?;
             let api = api.repo(repo);
             let config = api.get("config.json").map_err(NewEmbedderError::api_get)?;
@@ -97,7 +175,38 @@ impl Embedder {
                     })
                     .map_err(NewEmbedderError::api_get)?
             };
-            (config, tokenizer, weights, source)
+            let pooling = match api.get("1_Pooling/config.json") {
+                Ok(pooling) => Some(pooling),
+                Err(hf_hub::api::sync::ApiError::RequestError(error))
+                    if matches!(*error, ureq::Error::Status(404, _,)) =>
+                {
+                    // ignore the error if the file simply doesn't exist
+                    None
+                }
+                Err(error) => return Err(NewEmbedderError::api_get(error)),
+            };
+            let mut pooling: Pooling = match pooling {
+                Some(pooling_filename) => {
+                    let pooling = std::fs::read_to_string(&pooling_filename).map_err(|inner| {
+                        NewEmbedderError::open_pooling_config(pooling_filename.clone(), inner)
+                    })?;
+
+                    let pooling: PoolingConfig =
+                        serde_json::from_str(&pooling).map_err(|inner| {
+                            NewEmbedderError::deserialize_pooling_config(
+                                options.model.clone(),
+                                pooling_filename,
+                                inner,
+                            )
+                        })?;
+                    pooling.into()
+                }
+                None => Pooling::default(),
+            };
+
+            pooling.override_with(options.pooling);
+
+            (config, tokenizer, weights, source, pooling)
         };
 
         let config = std::fs::read_to_string(&config_filename)
@@ -122,6 +231,8 @@ impl Embedder {
             },
         };
 
+        tracing::debug!(model = options.model, weight=?weight_source, pooling=?pooling, "model config");
+
         let model = BertModel::load(vb, &config).map_err(NewEmbedderError::load_model)?;
 
         if let Some(pp) = tokenizer.get_padding_mut() {
@@ -134,7 +245,7 @@ impl Embedder {
             tokenizer.with_padding(Some(pp));
         }
 
-        let mut this = Self { model, tokenizer, options, dimensions: 0 };
+        let mut this = Self { model, tokenizer, options, dimensions: 0, pooling };
 
         let embeddings = this
             .embed(vec!["test".into()])
@@ -168,17 +279,53 @@ impl Embedder {
             .forward(&token_ids, &token_type_ids, None)
             .map_err(EmbedError::model_forward)?;
 
-        // Apply some avg-pooling by taking the mean embedding value for all tokens (including padding)
+        let embeddings = Self::pooling(embeddings, self.pooling)?;
-        let (_n_sentence, n_tokens, _hidden_size) =
-            embeddings.dims3().map_err(EmbedError::tensor_shape)?;
-
-        let embeddings = (embeddings.sum(1).map_err(EmbedError::tensor_value)? / (n_tokens as f64))
-            .map_err(EmbedError::tensor_shape)?;
 
         let embeddings: Vec<Embedding> = embeddings.to_vec2().map_err(EmbedError::tensor_shape)?;
         Ok(embeddings)
     }
 
+    fn pooling(embeddings: Tensor, pooling: Pooling) -> Result<Tensor, EmbedError> {
+        match pooling {
+            Pooling::Mean => Self::mean_pooling(embeddings),
+            Pooling::Cls => Self::cls_pooling(embeddings),
+            Pooling::Max => Self::max_pooling(embeddings),
+            Pooling::MeanSqrtLen => Self::mean_sqrt_pooling(embeddings),
+            Pooling::LastToken => Self::last_token_pooling(embeddings),
+        }
+    }
+
+    fn cls_pooling(embeddings: Tensor) -> Result<Tensor, EmbedError> {
+        embeddings.get_on_dim(1, 0).map_err(EmbedError::tensor_value)
+    }
+
+    fn mean_sqrt_pooling(embeddings: Tensor) -> Result<Tensor, EmbedError> {
+        let (_n_sentence, n_tokens, _hidden_size) =
+            embeddings.dims3().map_err(EmbedError::tensor_shape)?;
+
+        (embeddings.sum(1).map_err(EmbedError::tensor_value)? / (n_tokens as f64).sqrt())
+            .map_err(EmbedError::tensor_shape)
+    }
+
+    fn mean_pooling(embeddings: Tensor) -> Result<Tensor, EmbedError> {
+        let (_n_sentence, n_tokens, _hidden_size) =
+            embeddings.dims3().map_err(EmbedError::tensor_shape)?;
+
+        (embeddings.sum(1).map_err(EmbedError::tensor_value)? / (n_tokens as f64))
+            .map_err(EmbedError::tensor_shape)
+    }
+
+    fn max_pooling(embeddings: Tensor) -> Result<Tensor, EmbedError> {
+        embeddings.max(1).map_err(EmbedError::tensor_shape)
+    }
+
+    fn last_token_pooling(embeddings: Tensor) -> Result<Tensor, EmbedError> {
+        let (_n_sentence, n_tokens, _hidden_size) =
+            embeddings.dims3().map_err(EmbedError::tensor_shape)?;
+
+        embeddings.get_on_dim(1, n_tokens - 1).map_err(EmbedError::tensor_value)
+    }
+
     pub fn embed_one(&self, text: &str) -> std::result::Result<Embedding, EmbedError> {
         let tokens = self.tokenizer.encode(text, true).map_err(EmbedError::tokenize)?;
         let token_ids = tokens.get_ids();
@@ -192,11 +339,8 @@ impl Embedder {
             .forward(&token_ids, &token_type_ids, None)
             .map_err(EmbedError::model_forward)?;
 
-        // Apply some avg-pooling by taking the mean embedding value for all tokens (including padding)
+        let embedding = Self::pooling(embeddings, self.pooling)?;
-        let (_n_sentence, n_tokens, _hidden_size) =
+
-            embeddings.dims3().map_err(EmbedError::tensor_shape)?;
-        let embedding = (embeddings.sum(1).map_err(EmbedError::tensor_value)? / (n_tokens as f64))
-            .map_err(EmbedError::tensor_shape)?;
         let embedding = embedding.squeeze(0).map_err(EmbedError::tensor_shape)?;
         let embedding: Embedding = embedding.to_vec1().map_err(EmbedError::tensor_shape)?;
         Ok(embedding)

crates/milli/src/vector/settings.rs

@@ -6,6 +6,7 @@ use roaring::RoaringBitmap;
 use serde::{Deserialize, Serialize};
 use utoipa::ToSchema;
 
+use super::hf::OverridePooling;
 use super::{ollama, openai, DistributionShift};
 use crate::prompt::{default_max_bytes, PromptData};
 use crate::update::Setting;
@@ -30,6 +31,10 @@ pub struct EmbeddingSettings {
     pub revision: Setting<String>,
     #[serde(default, skip_serializing_if = "Setting::is_not_set")]
     #[deserr(default)]
+    #[schema(value_type = Option<OverridePooling>)]
+    pub pooling: Setting<OverridePooling>,
+    #[serde(default, skip_serializing_if = "Setting::is_not_set")]
+    #[deserr(default)]
     #[schema(value_type = Option<String>)]
     pub api_key: Setting<String>,
     #[serde(default, skip_serializing_if = "Setting::is_not_set")]
@@ -164,6 +169,7 @@ impl SettingsDiff {
                     mut source,
                     mut model,
                     mut revision,
+                    mut pooling,
                     mut api_key,
                     mut dimensions,
                     mut document_template,
@@ -180,6 +186,7 @@ impl SettingsDiff {
                     source: new_source,
                     model: new_model,
                     revision: new_revision,
+                    pooling: new_pooling,
                     api_key: new_api_key,
                     dimensions: new_dimensions,
                     document_template: new_document_template,
@@ -210,6 +217,7 @@ impl SettingsDiff {
                         &source,
                         &mut model,
                         &mut revision,
+                        &mut pooling,
                         &mut dimensions,
                         &mut url,
                         &mut request,
@@ -225,6 +233,9 @@ impl SettingsDiff {
                 if revision.apply(new_revision) {
                     ReindexAction::push_action(&mut reindex_action, ReindexAction::FullReindex);
                 }
+                if pooling.apply(new_pooling) {
+                    ReindexAction::push_action(&mut reindex_action, ReindexAction::FullReindex);
+                }
                 if dimensions.apply(new_dimensions) {
                     match source {
                         // regenerate on dimensions change in OpenAI since truncation is supported
@@ -290,6 +301,7 @@ impl SettingsDiff {
                     source,
                     model,
                     revision,
+                    pooling,
                     api_key,
                     dimensions,
                     document_template,
@@ -338,6 +350,7 @@ fn apply_default_for_source(
     source: &Setting<EmbedderSource>,
     model: &mut Setting<String>,
     revision: &mut Setting<String>,
+    pooling: &mut Setting<OverridePooling>,
     dimensions: &mut Setting<usize>,
     url: &mut Setting<String>,
     request: &mut Setting<serde_json::Value>,
@@ -350,6 +363,7 @@ fn apply_default_for_source(
         Setting::Set(EmbedderSource::HuggingFace) => {
             *model = Setting::Reset;
             *revision = Setting::Reset;
+            *pooling = Setting::Reset;
             *dimensions = Setting::NotSet;
             *url = Setting::NotSet;
             *request = Setting::NotSet;
@@ -359,6 +373,7 @@ fn apply_default_for_source(
         Setting::Set(EmbedderSource::Ollama) => {
             *model = Setting::Reset;
             *revision = Setting::NotSet;
+            *pooling = Setting::NotSet;
             *dimensions = Setting::Reset;
             *url = Setting::NotSet;
             *request = Setting::NotSet;
@@ -368,6 +383,7 @@ fn apply_default_for_source(
         Setting::Set(EmbedderSource::OpenAi) | Setting::Reset => {
             *model = Setting::Reset;
             *revision = Setting::NotSet;
+            *pooling = Setting::NotSet;
             *dimensions = Setting::NotSet;
             *url = Setting::Reset;
             *request = Setting::NotSet;
@@ -377,6 +393,7 @@ fn apply_default_for_source(
         Setting::Set(EmbedderSource::Rest) => {
             *model = Setting::NotSet;
             *revision = Setting::NotSet;
+            *pooling = Setting::NotSet;
             *dimensions = Setting::Reset;
             *url = Setting::Reset;
             *request = Setting::Reset;
@@ -386,6 +403,7 @@ fn apply_default_for_source(
         Setting::Set(EmbedderSource::UserProvided) => {
             *model = Setting::NotSet;
             *revision = Setting::NotSet;
+            *pooling = Setting::NotSet;
             *dimensions = Setting::Reset;
             *url = Setting::NotSet;
             *request = Setting::NotSet;
@@ -419,6 +437,7 @@ impl EmbeddingSettings {
     pub const SOURCE: &'static str = "source";
     pub const MODEL: &'static str = "model";
     pub const REVISION: &'static str = "revision";
+    pub const POOLING: &'static str = "pooling";
     pub const API_KEY: &'static str = "apiKey";
     pub const DIMENSIONS: &'static str = "dimensions";
     pub const DOCUMENT_TEMPLATE: &'static str = "documentTemplate";
@@ -446,6 +465,7 @@ impl EmbeddingSettings {
                 &[EmbedderSource::HuggingFace, EmbedderSource::OpenAi, EmbedderSource::Ollama]
             }
             Self::REVISION => &[EmbedderSource::HuggingFace],
+            Self::POOLING => &[EmbedderSource::HuggingFace],
             Self::API_KEY => {
                 &[EmbedderSource::OpenAi, EmbedderSource::Ollama, EmbedderSource::Rest]
             }
@@ -500,6 +520,7 @@ impl EmbeddingSettings {
                 Self::SOURCE,
                 Self::MODEL,
                 Self::REVISION,
+                Self::POOLING,
                 Self::DOCUMENT_TEMPLATE,
                 Self::DOCUMENT_TEMPLATE_MAX_BYTES,
                 Self::DISTRIBUTION,
@@ -592,10 +613,12 @@ impl From<EmbeddingConfig> for EmbeddingSettings {
                 model,
                 revision,
                 distribution,
+                pooling,
             }) => Self {
                 source: Setting::Set(EmbedderSource::HuggingFace),
                 model: Setting::Set(model),
                 revision: Setting::some_or_not_set(revision),
+                pooling: Setting::Set(pooling),
                 api_key: Setting::NotSet,
                 dimensions: Setting::NotSet,
                 document_template: Setting::Set(prompt.template),
@@ -617,6 +640,7 @@ impl From<EmbeddingConfig> for EmbeddingSettings {
                 source: Setting::Set(EmbedderSource::OpenAi),
                 model: Setting::Set(embedding_model.name().to_owned()),
                 revision: Setting::NotSet,
+                pooling: Setting::NotSet,
                 api_key: Setting::some_or_not_set(api_key),
                 dimensions: Setting::some_or_not_set(dimensions),
                 document_template: Setting::Set(prompt.template),
@@ -638,6 +662,7 @@ impl From<EmbeddingConfig> for EmbeddingSettings {
                 source: Setting::Set(EmbedderSource::Ollama),
                 model: Setting::Set(embedding_model),
                 revision: Setting::NotSet,
+                pooling: Setting::NotSet,
                 api_key: Setting::some_or_not_set(api_key),
                 dimensions: Setting::some_or_not_set(dimensions),
                 document_template: Setting::Set(prompt.template),
@@ -656,6 +681,7 @@ impl From<EmbeddingConfig> for EmbeddingSettings {
                 source: Setting::Set(EmbedderSource::UserProvided),
                 model: Setting::NotSet,
                 revision: Setting::NotSet,
+                pooling: Setting::NotSet,
                 api_key: Setting::NotSet,
                 dimensions: Setting::Set(dimensions),
                 document_template: Setting::NotSet,
@@ -679,6 +705,7 @@ impl From<EmbeddingConfig> for EmbeddingSettings {
                 source: Setting::Set(EmbedderSource::Rest),
                 model: Setting::NotSet,
                 revision: Setting::NotSet,
+                pooling: Setting::NotSet,
                 api_key: Setting::some_or_not_set(api_key),
                 dimensions: Setting::some_or_not_set(dimensions),
                 document_template: Setting::Set(prompt.template),
@@ -701,6 +728,7 @@ impl From<EmbeddingSettings> for EmbeddingConfig {
             source,
             model,
             revision,
+            pooling,
             api_key,
             dimensions,
             document_template,
@@ -764,6 +792,9 @@ impl From<EmbeddingSettings> for EmbeddingConfig {
                     if let Some(revision) = revision.set() {
                         options.revision = Some(revision);
                     }
+                    if let Some(pooling) = pooling.set() {
+                        options.pooling = pooling;
+                    }
                     options.distribution = distribution.set();
                     this.embedder_options = super::EmbedderOptions::HuggingFace(options);
                 }