test

app.py

@@ -178,6 +178,7 @@ def run_edit(audio_file, caption, num_steps, guidance_scale, guidance_rescale, s
     from omegaconf import OmegaConf
     from safetensors.torch import load_file
     import diffusers.schedulers as noise_schedulers
+    logger.info("🚀 Starting ..")
     torch.backends.cuda.matmul.allow_tf32 = False
     torch.backends.cudnn.allow_tf32 = False
 
@@ -310,7 +311,7 @@ def build_demo():
                         # 示例 1 (原本的)
                         ["./Ym8O802VvJes.wav", "Mix in dog barking around the middle."],
                         ["./YDKM2KjNkX18.wav", "Incorporate Telephone bell ringing into the background."],
-                        ["./drop_audiocaps_1.wav", "Remove the sound of several beeps."],
+                        ["./drop_audiocaps_1.wav", "Erase the rain falling sound  from the background."],
                         ["./reorder_audiocaps_1.wav", "Switch the positions of the woman's voice and whistling."]
                     ],
                     inputs=[audio_in, caption],  # 对应上面列表的顺序：第一个是 Audio，第二个是 Textbox