more

cmd/needle-sidecar/eval/run_eval.py

@@ -63,6 +63,10 @@ def main():
     print(f"\naccuracy {acc:.2%} over {len(results)} rows; "
           f"median latency {sorted(lat)[len(lat)//2]:.0f}ms")
 
+    for src in sorted({r.get("src", "?") for r in results}):
+        sub = [r for r in results if r.get("src", "?") == src]
+        print(f"  {src:<11} {sum(r['ok'] for r in sub)}/{len(sub)}")
+
     wrong = [r for r in results if not r["ok"]]
     if wrong:
         print("\nmisses:")