Update app.py

app.py

@@ -674,6 +674,8 @@ def voice_conversion(source, target, diffusion_steps, length_adjust, inference_c
             ).export(format="mp3", bitrate=bitrate).read()
             yield mp3_bytes, None
 
+
+
 import gradio as gr
 
 gallery_data = {"sikokumetan": {"webp": "default/sikokumetan.webp", "mp3": "default/sikokumetan.mp3"}}
@@ -695,6 +697,7 @@ if __name__ == "__main__":
     inputs = [
         gr.Audio(type="filepath", label="元音声"),
         gr.Audio(type="filepath", label="参考音声"),
+        gallery = gr.Gallery(label="ギャラリー", show_label=True, value=[gallery_data["sikokumetan"]["webp"]]),
         gr.Slider(minimum=1, maximum=200, value=10, step=1, label="拡散ステップ数", info="デフォルトは10、50～100が最適な品質"),
         gr.Slider(minimum=0.5, maximum=2.0, step=0.1, value=1.0, label="長さ調整", info="1.0未満で速度を上げ、1.0以上で速度を遅くします"),
         gr.Slider(minimum=0.0, maximum=1.0, step=0.1, value=0.7, label="推論CFG率", info="わずかな影響があります"),
@@ -717,9 +720,6 @@ if __name__ == "__main__":
             gr.Audio(label="完全出力音声", streaming=False, format='wav')
         ], title="Seed Voice Conversion", examples=examples, cache_examples=False)
 
-        with gr.Row():
-            gallery = gr.Gallery(label="ギャラリー", show_label=True, value=[gallery_data["sikokumetan"]["webp"]])
-
         gallery.select(update_reference, outputs=inputs[1])
 
     interface.launch()