Anna Sun
		
	commited on
		
		
					Commit 
							
							·
						
						f67796b
	
1
								Parent(s):
							
							d760758
								
update welcome
Browse files
    	
        streaming-react-app/src/StreamingInterface.tsx
    CHANGED
    
    | @@ -769,9 +769,9 @@ export default function StreamingInterface() { | |
| 769 | 
             
                              Check out the seamless_communication <a target="_blank" rel="noopener noreferrer" href="https://github.com/facebookresearch/seamless_communication/tree/main">README</a> for more information.
         | 
| 770 | 
             
                              <br/>
         | 
| 771 | 
             
                              SeamlessStreaming model is a research model and is not released
         | 
| 772 | 
            -
                              for production deployment.  | 
| 773 | 
            -
                               | 
| 774 | 
            -
                              every couple of sentences, or you may wish adjust the VAD threshold
         | 
| 775 | 
             
                              in the model config. The real-time performance will degrade
         | 
| 776 | 
             
                              if you try streaming multiple speakers at the same time.
         | 
| 777 | 
             
                            </Typography>
         | 
|  | |
| 769 | 
             
                              Check out the seamless_communication <a target="_blank" rel="noopener noreferrer" href="https://github.com/facebookresearch/seamless_communication/tree/main">README</a> for more information.
         | 
| 770 | 
             
                              <br/>
         | 
| 771 | 
             
                              SeamlessStreaming model is a research model and is not released
         | 
| 772 | 
            +
                              for production deployment. It is important to use a microphone with 
         | 
| 773 | 
            +
                              noise cancellation (for e.g. a smartphone), otherwise you may see model hallucination on noises. 
         | 
| 774 | 
            +
                              It works best if you pause every couple of sentences, or you may wish adjust the VAD threshold
         | 
| 775 | 
             
                              in the model config. The real-time performance will degrade
         | 
| 776 | 
             
                              if you try streaming multiple speakers at the same time.
         | 
| 777 | 
             
                            </Typography>
         | 
