Create lancedb index after materializing dataset (#1519)

jdye64 · web-flow · commit 8e2d80f80c23 · 2026-03-09T16:44:54.000-04:00
diff --git a/nemo_retriever/src/nemo_retriever/examples/batch_pipeline.py b/nemo_retriever/src/nemo_retriever/examples/batch_pipeline.py
@@ -732,6 +732,9 @@ def main(
             .materialize()
         )
 
+        if hasattr(ingestor, "_create_lancedb_index"):
+            ingestor._create_lancedb_index()
+
         ingest_elapsed_s = time.perf_counter() - ingest_start
         rows_processed = _count_materialized_rows(ingest_results)
         logger.info(

Original file line number	Diff line number	Diff line change
`@@ -732,6 +732,9 @@ def main(`
`732`	`732`	`.materialize()`
`733`	`733`	`)`
`734`	`734`
	`735`	`+ if hasattr(ingestor, "_create_lancedb_index"):`
	`736`	`+ ingestor._create_lancedb_index()`
	`737`	`+`
`735`	`738`	`ingest_elapsed_s = time.perf_counter() - ingest_start`
`736`	`739`	`rows_processed = _count_materialized_rows(ingest_results)`
`737`	`740`	`logger.info(`