Update app.py

app.py

@@ -143,7 +143,7 @@ with gr.Blocks() as demo:
             | :--- | :--- | :--- | :--- | :--- |
             | **Zero-Shot Text Prompting** | Direct Generation | Full Dataset (535 files) | 67.3% | High variance across vocal pitches |
             | **Linear Embedding Head** | Stratified 80/20 Split | Unseen 20% Subset | **97.2%** | Rare confusion on acoustic twins |
-            | **Linear Embedding Head** | Cross-Speaker (Leave-8-Out) | 6 Unseen Speakers (Blind) | **92.2%** | Robust across unique vocal anatomy |
+            | **Linear Embedding Head** | Cross-Speaker (Leave-6-Out) | 6 Unseen Speakers (Blind) | **92.2%** | Robust across unique vocal anatomy |
 
             ---