Spaces:
				
			
			
	
			
			
					
		Running
		
	
	
	
			
			
	
	
	
	
		
		
					
		Running
		
	Sync from GitHub repo
Browse filesThis Space is synced from the GitHub repo: https://github.com/SWivid/F5-TTS. Please submit contributions to the Space there
    	
        app.py
    CHANGED
    
    | 
         @@ -80,9 +80,9 @@ def generate_response(messages, model, tokenizer): 
     | 
|
| 80 | 
         | 
| 81 | 
         
             
            @gpu_decorator
         
     | 
| 82 | 
         
             
            def infer(
         
     | 
| 83 | 
         
            -
                ref_audio_orig, ref_text, gen_text, model, remove_silence, cross_fade_duration=0.15, speed=1
         
     | 
| 84 | 
         
             
            ):
         
     | 
| 85 | 
         
            -
                ref_audio, ref_text = preprocess_ref_audio_text(ref_audio_orig, ref_text, show_info= 
     | 
| 86 | 
         | 
| 87 | 
         
             
                if model == "F5-TTS":
         
     | 
| 88 | 
         
             
                    ema_model = F5TTS_ema_model
         
     | 
| 
         @@ -97,6 +97,7 @@ def infer( 
     | 
|
| 97 | 
         
             
                    vocoder,
         
     | 
| 98 | 
         
             
                    cross_fade_duration=cross_fade_duration,
         
     | 
| 99 | 
         
             
                    speed=speed,
         
     | 
| 
         | 
|
| 100 | 
         
             
                    progress=gr.Progress(),
         
     | 
| 101 | 
         
             
                )
         
     | 
| 102 | 
         | 
| 
         @@ -404,7 +405,7 @@ with gr.Blocks() as app_multistyle: 
     | 
|
| 404 | 
         | 
| 405 | 
         
             
                        # Generate speech for this segment
         
     | 
| 406 | 
         
             
                        audio, _ = infer(
         
     | 
| 407 | 
         
            -
                            ref_audio, ref_text, text, model_choice, remove_silence, 0
         
     | 
| 408 | 
         
             
                        )  # show_info=print no pull to top when generating
         
     | 
| 409 | 
         
             
                        sr, audio_data = audio
         
     | 
| 410 | 
         | 
| 
         @@ -726,4 +727,4 @@ if __name__ == "__main__": 
     | 
|
| 726 | 
         
             
                if not USING_SPACES:
         
     | 
| 727 | 
         
             
                    main()
         
     | 
| 728 | 
         
             
                else:
         
     | 
| 729 | 
         
            -
                    app.queue().launch()
         
     | 
| 
         | 
|
| 80 | 
         | 
| 81 | 
         
             
            @gpu_decorator
         
     | 
| 82 | 
         
             
            def infer(
         
     | 
| 83 | 
         
            +
                ref_audio_orig, ref_text, gen_text, model, remove_silence, cross_fade_duration=0.15, speed=1, show_info=gr.Info
         
     | 
| 84 | 
         
             
            ):
         
     | 
| 85 | 
         
            +
                ref_audio, ref_text = preprocess_ref_audio_text(ref_audio_orig, ref_text, show_info=show_info)
         
     | 
| 86 | 
         | 
| 87 | 
         
             
                if model == "F5-TTS":
         
     | 
| 88 | 
         
             
                    ema_model = F5TTS_ema_model
         
     | 
| 
         | 
|
| 97 | 
         
             
                    vocoder,
         
     | 
| 98 | 
         
             
                    cross_fade_duration=cross_fade_duration,
         
     | 
| 99 | 
         
             
                    speed=speed,
         
     | 
| 100 | 
         
            +
                    show_info=show_info,
         
     | 
| 101 | 
         
             
                    progress=gr.Progress(),
         
     | 
| 102 | 
         
             
                )
         
     | 
| 103 | 
         | 
| 
         | 
|
| 405 | 
         | 
| 406 | 
         
             
                        # Generate speech for this segment
         
     | 
| 407 | 
         
             
                        audio, _ = infer(
         
     | 
| 408 | 
         
            +
                            ref_audio, ref_text, text, model_choice, remove_silence, 0, show_info=print
         
     | 
| 409 | 
         
             
                        )  # show_info=print no pull to top when generating
         
     | 
| 410 | 
         
             
                        sr, audio_data = audio
         
     | 
| 411 | 
         | 
| 
         | 
|
| 727 | 
         
             
                if not USING_SPACES:
         
     | 
| 728 | 
         
             
                    main()
         
     | 
| 729 | 
         
             
                else:
         
     | 
| 730 | 
         
            +
                    app.queue().launch()
         
     |