Update HNSW Brute Force Threshold While Searching with Bitset (#829)

Signed-off-by: Patrick Weizhi Xu <weizhi.xu@zilliz.com>
milvus-io · Apr 21, 2023 · 5cd495a · 5cd495a
1 parent b2cab04
commit 5cd495a
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 3 deletions.
diff --git a/thirdparty/DiskANN/src/pq_flash_index.cpp b/thirdparty/DiskANN/src/pq_flash_index.cpp
@@ -1001,7 +1001,7 @@ namespace diskann {
     float query_norm = query_norm_opt.value();
     auto ctx = this->reader->get_ctx();
 
-    if (!bitset_view.empty() && bitset_view.count() > bitset_view.size() * kDiskAnnBruteForceFilterRate) {
+    if (!bitset_view.empty() && bitset_view.count() >= bitset_view.size() * kDiskAnnBruteForceFilterRate) {
         brute_force_beam_search(data, query_norm, k_search, indices, distances,
                                 beam_width, ctx, stats, feder, bitset_view);
         this->thread_data.push(data);

diff --git a/thirdparty/hnswlib/hnswlib/hnswalg.h b/thirdparty/hnswlib/hnswlib/hnswalg.h
@@ -35,7 +35,7 @@
 namespace hnswlib {
 typedef unsigned int tableint;
 typedef unsigned int linklistsizeint;
-constexpr float kHnswBruteForceFilterRate = 0.8f;
+constexpr float kHnswBruteForceFilterRate = 0.93f;
 
 template <typename dist_t>
 class HierarchicalNSW : public AlgorithmInterface<dist_t> {
@@ -1030,7 +1030,7 @@ class HierarchicalNSW : public AlgorithmInterface<dist_t> {
         if (cur_element_count == 0)
             return {};
 
-        if (!bitset.empty() && bitset.count() > (cur_element_count * kHnswBruteForceFilterRate)) {
+        if (!bitset.empty() && bitset.count() >= (cur_element_count * kHnswBruteForceFilterRate)) {
             assert(cur_element_count == bitset.size());
             knowhere::ResultMaxHeap<dist_t, labeltype> max_heap(k);
             for (labeltype id = 0; id < cur_element_count; ++id) {