remove

Author: kelindar

dist/MiniLM-L6-v2.Q4_K_M.gguf

@@ -18,7 +18,6 @@ type Result[T any] struct {
 // Index represents a brute-force search index, returning exact results.
 type Index[T any] struct {
 	arr []entry[T]
-	dim int
 }
 
 // NewIndex creates a new exact search index.
@@ -42,7 +41,7 @@ func (b *Index[T]) Search(query Vector, k int) []Result[T] {
 		return nil
 	}
 
-	var h minheap[T]
+	dst := make(minheap[T], 0, k)
 	for _, v := range b.arr {
 		relevance := Cosine(v.Vector, query)
 		result := Result[T]{
@@ -53,29 +52,24 @@ func (b *Index[T]) Search(query Vector, k int) []Result[T] {
 		// If the heap is not full, add the result, otherwise replace
 		// the minimum element
 		switch {
-		case h.Len() < k:
-			h.Push(result)
-		case relevance > h[0].Relevance:
-			h.Pop()
-			h.Push(result)
+		case dst.Len() < k:
+			dst.Push(result)
+		case relevance > dst[0].Relevance:
+			dst.Pop()
+			dst.Push(result)
 		}
 	}
 
 	// Sort the results by relevance
-	sort.Sort(&h)
-	return h
+	sort.Sort(&dst)
+	return dst
 }
 
 // --------------------------------- Heap ---------------------------------
 
 // minheap is a min-heap of top values, ordered by relevance.
 type minheap[T any] []Result[T]
 
-// Reset resets the minheap to an empty state.
-func (h *minheap[T]) Reset() {
-	*h = (*h)[:0]
-}
-
 // Len, Less, Swap implement the sort.Interface.
 func (h *minheap[T]) Len() int           { return len(*h) }
 func (h *minheap[T]) Less(i, j int) bool { return (*h)[i].Relevance > (*h)[j].Relevance }

dist/linux-x64/libggml.so

@@ -10,24 +10,25 @@ import (
 
 /*
 cpu: 13th Gen Intel(R) Core(TM) i7-13700K
-BenchmarkBruteForce-24    	    4123	    301825 ns/op	     744 B/op	       5 allocs/op
+BenchmarkIndex/search-24        	    4401	    265430 ns/op	     264 B/op	       2 allocs/op
 */
-func BenchmarkBruteForce(b *testing.B) {
+func BenchmarkIndex(b *testing.B) {
 	data, err := loadDataset()
 	assert.NoError(b, err)
 
-	bag := NewIndex[string]()
+	index := NewIndex[string]()
 	for _, entry := range data {
-		bag.Add(entry.Vector, entry.Pair[0])
+		index.Add(entry.Vector, entry.Pair[0])
 	}
 
-	b.ReportAllocs()
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		for _, v := range bag.Search(data[i%1000].Vector, 5) {
-			_ = v
+	b.Run("search", func(b *testing.B) {
+		b.ReportAllocs()
+		b.ResetTimer()
+		for i := 0; i < b.N; i++ {
+			_ = index.Search(data[i%1000].Vector, 5)
 		}
-	}
+	})
+
 }
 
 func TestIndex(t *testing.T) {