Update app_multi.py

app_multi.py

@@ -74,7 +74,8 @@ def combine_music(video, audio):
     audio_background = mpe.AudioFileClip(audio)
     final_audio = mpe.CompositeAudioClip([my_clip.audio, audio_background])
     final_clip = my_clip.set_audio(final_audio)
-    return final_clip
+    final_clip.write_videofile("video.mp4")
+    return "video.mp4"
 
 # Reference: https://huggingface.co/spaces/zomehwh/rvc-models/blob/main/app.py#L21  # noqa
 in_hf_space = getenv('SYSTEM') == 'spaces'
@@ -800,6 +801,8 @@ with app:
                                 
                             submit = gr.Button('想把我唱给你听', elem_id="sadtalker_generate", variant='primary')
 
+                            inp_mv_1 = gen_video
+                            inp_mv_2 = as_audio_no_vocals
                             gen_mv = gr.Button('为视频添加伴奏吧', variant='primary')
 
                             with gr.Row():
@@ -816,8 +819,8 @@ with app:
                                 # height = gr.Slider(minimum=64, elem_id="img2img_height", maximum=2048, step=8, label="Manually Crop Height", value=512) # img2img_width
                                 pose_style = gr.Slider(minimum=0, maximum=46, step=1, label="Pose style", value=0, visible=False) # 
                                 size_of_image = gr.Radio([256, 512], value=256, label='face model resolution', info="use 256/512 model?", visible=False) # 
-                                preprocess_type = gr.Radio(['crop', 'full'], value='crop', label='是否聚焦角色面部', info="crop：视频会聚焦角色面部；full：视频会显示图片全貌")
-                                is_still_mode = gr.Checkbox(label="静态模式 (开启静态模式，角色的面部动作会减少；默认开启)", value=True)
+                                preprocess_type = gr.Radio(['crop', 'extfull'], value='crop', label='是否聚焦角色面部', info="crop：视频会聚焦角色面部；extfull：视频会显示图片全貌")
+                                is_still_mode = gr.Checkbox(label="静态模式 (开启静态模式，角色的面部动作会减少；默认开启)", value=True, visible=False)
                                 batch_size = gr.Slider(label="Batch size (数值越大，生成速度越快；若显卡性能好，可增大数值)", step=1, maximum=32, value=4)
                                 enhancer = gr.Checkbox(label="GFPGAN as Face enhancer", visible=False)
                                     
@@ -836,7 +839,7 @@ with app:
                         outputs=[gen_video]
                         )
         
-            gen_mv.click(fn=combine_music, inputs=[gen_video, as_audio_no_vocals], outputs=[music_video])
+            gen_mv.click(fn=combine_music, inputs=[inp_mv_1, inp_mv_2], outputs=[music_video])
         
  
     gr.Markdown("### <center>注意❗：请不要生成会对个人以及组织造成侵害的内容，此程序仅供科研、学习及个人娱乐使用。</center>")