Update app.py

app.py

@@ -8,7 +8,8 @@ token = os.environ.get('HF_TOKEN')
 caption = gr.Blocks.load(name="spaces/SRDdev/Image-Caption")
 audio_gen = gr.Blocks.load(name="spaces/fffiloni/audioldm-text-to-audio-generation-clone", api_key=token)
 
-def infer(image_input, manual_caption):
+def infer(image_input, manual_caption, duration_in):
+    print(duration_in)
     if manual_caption == "":
         cap = caption(image_input, fn_index=0)
         print("gpt2 caption: " + cap)
@@ -76,7 +77,7 @@ with gr.Blocks(css="style.css") as demo:
     
         input_img = gr.Image(type="filepath", elem_id="input-img")
         manual_cap = gr.Textbox(label="Manual Image description (optional)", lines=2, placeholder="If you're not happy with sound result, you can manually describe the scene depicted in your image :)")
-        #duration_in = gr.Slider(minimum=2, maximum=12, step=1, value=6, label="Duration")
+        duration_in = gr.Slider(minimum=2, maximum=12, step=1, value=10, label="Duration")
         caption_output = gr.Textbox(label="Caption", visible=False, elem_id="text-caption")
         sound_output = gr.Audio(label="Result", elem_id="sound-output")
         
@@ -89,7 +90,7 @@ with gr.Blocks(css="style.css") as demo:
 
         gr.HTML(article)
         
-    generate.click(infer, inputs=[input_img, manual_cap], outputs=[caption_output, sound_output, share_group], api_name="i2fx")
+    generate.click(infer, inputs=[input_img, manual_cap, duration_in], outputs=[caption_output, sound_output, share_group], api_name="i2fx")
     share_button.click(None, [], [], _js=share_js)
 
 demo.queue(max_size=32).launch(debug=True)