add indexing docs/sec column to knnPerfTest tabular output

mikemccand · Oct 30, 2024 · 0cdcbb3 · 0cdcbb3
1 parent d9d5c64
commit 0cdcbb3
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 7 deletions.
diff --git a/src/main/knn/KnnGraphTester.java b/src/main/knn/KnnGraphTester.java
@@ -690,9 +690,10 @@ private void testSearch(Path indexPath, Path queryPath, Path outputPath, int[][]
       } else {
         quantizeDesc = "no";
       }
+      double reindexSec = reindexTimeMsec / 1000.0;
       System.out.printf(
           Locale.ROOT,
-          "SUMMARY: %5.3f\t%5.3f\t%d\t%d\t%d\t%d\t%d\t%s\t%d\t%.2f\t%.2f\t%d\t%.2f\t%.2f\t%s\n",
+          "SUMMARY: %5.3f\t%5.3f\t%d\t%d\t%d\t%d\t%d\t%s\t%d\t%.2f\t%.2f\t%.2f\t%d\t%.2f\t%.2f\t%s\n",
           recall,
           totalCpuTimeMS / (float) numIters,
           numDocs,
@@ -702,7 +703,8 @@ private void testSearch(Path indexPath, Path queryPath, Path outputPath, int[][]
           beamWidth,
           quantizeDesc,
           totalVisited,
-          reindexTimeMsec / 1000.0,
+          reindexSec,
+          numDocs / reindexSec,
           forceMergeTimeSec,
           indexNumSegments,
           indexSizeOnDiskMB,

diff --git a/src/python/knnPerfTest.py b/src/python/knnPerfTest.py
@@ -56,16 +56,16 @@
     #'fanout': (20, 100, 250)
     'fanout': (6,),
     #'quantize': None,
-    'quantizeBits': (32,),
+    #'quantizeBits': (32,),
     'numMergeWorker': (12,),
     'numMergeThread': (4,),
     'encoding': ('float32',),
     # 'metric': ('angular',),  # default is angular (dot_product)
     #'quantize': (True,),
-    'quantizeBits': (4,),
+    'quantizeBits': (4, 7, 32),
     #'fanout': (0,),
     'topK': (10,),
-    'quantizeCompress': (True, False),
+    'quantizeCompress': (False,),
     #'niter': (10,),
 }
 
@@ -143,7 +143,7 @@ def run_knn_benchmark(checkout, values):
             '-docs', doc_vectors,
             '-reindex',
             '-search-and-stats', query_vectors,
-            #'-metric', 'euclidean',
+            '-metric', 'mip',
             # '-parentJoin', parentJoin_meta_file,
             # '-numMergeThread', '8', '-numMergeWorker', '8',
             '-forceMerge',
@@ -170,7 +170,7 @@ def run_knn_benchmark(checkout, values):
         all_results.append(summary)
     print('\nResults:')
 
-    header = 'recall\tlatency (ms)\tnDoc\ttopK\tfanout\tmaxConn\tbeamWidth\tquantized\tvisited\tindex s\tforce merge s\tnum segments\tindex size (MB)\tselectivity\tfilterType'
+    header = 'recall\tlatency (ms)\tnDoc\ttopK\tfanout\tmaxConn\tbeamWidth\tquantized\tvisited\tindex s\tindex docs/s\tforce merge s\tnum segments\tindex size (MB)\tselectivity\tfilterType'
 
     # crazy logic to make everything fixed width so rendering in fixed width font "aligns":
     headers = header.split('\t')