feat: don't print relevance score by default

gptme_rag/cli.py

@@ -358,6 +358,11 @@ def index(
     type=click.Choice(["cuda", "cpu"]),
     help="Device to run embeddings on (cuda or cpu)",
 )
+@click.option(
+    "--print-relevance",
+    is_flag=True,
+    help="Print relevance scores",
+)
 def search(
     query: str,
     paths: list[Path],
@@ -371,6 +376,7 @@ def search(
     weights: str | None,
     embedding_function: str | None,
     device: str | None,
+    print_relevance: bool,
 ):
     """Search the index and assemble context."""
     paths = [path.resolve() for path in paths]
@@ -477,7 +483,7 @@ def get_expanded_content(doc: Document, expand: str, indexer: Indexer) -> str:
     if format == "full":
         for i, doc in enumerate(documents):
             # Show relevance info first
-            if distances:
+            if distances and print_relevance:
                 formatter.print_relevance(1 - distances[i])
 
             # Get and format content
@@ -529,7 +535,8 @@ def get_expanded_content(doc: Document, expand: str, indexer: Indexer) -> str:
             console.print(f"\n  {'Total':15} [bold blue]{total:>7.3f}[/bold blue]")
         else:
             # Just show the base relevance score
-            formatter.print_relevance(1 - distances[i])
+            if distances and print_relevance:
+                formatter.print_relevance(1 - distances[i])
 
         # Display preview
         formatter.print_preview(doc)