Spaces:
				
			
			
	
			
			
		Runtime error
		
	
	
	
			
			
	
	
	
	
		
		
		Runtime error
		
	Fix cache max_seq_len
Browse files- app.py +2 -2
 - tools/llama/generate.py +4 -2
 
    	
        app.py
    CHANGED
    
    | 
         @@ -414,7 +414,7 @@ def build_app(): 
     | 
|
| 414 | 
         
             
                                        label="Maximum tokens per batch, 0 means no limit",
         
     | 
| 415 | 
         
             
                                        minimum=0,
         
     | 
| 416 | 
         
             
                                        maximum=2048,
         
     | 
| 417 | 
         
            -
                                        value= 
     | 
| 418 | 
         
             
                                        step=8,
         
     | 
| 419 | 
         
             
                                    )
         
     | 
| 420 | 
         | 
| 
         @@ -640,7 +640,7 @@ if __name__ == "__main__": 
     | 
|
| 640 | 
         
             
                        reference_audio=None,
         
     | 
| 641 | 
         
             
                        reference_text="",
         
     | 
| 642 | 
         
             
                        max_new_tokens=0,
         
     | 
| 643 | 
         
            -
                        chunk_length= 
     | 
| 644 | 
         
             
                        top_p=0.7,
         
     | 
| 645 | 
         
             
                        repetition_penalty=1.2,
         
     | 
| 646 | 
         
             
                        temperature=0.7,
         
     | 
| 
         | 
|
| 414 | 
         
             
                                        label="Maximum tokens per batch, 0 means no limit",
         
     | 
| 415 | 
         
             
                                        minimum=0,
         
     | 
| 416 | 
         
             
                                        maximum=2048,
         
     | 
| 417 | 
         
            +
                                        value=0,  # 0 means no limit
         
     | 
| 418 | 
         
             
                                        step=8,
         
     | 
| 419 | 
         
             
                                    )
         
     | 
| 420 | 
         | 
| 
         | 
|
| 640 | 
         
             
                        reference_audio=None,
         
     | 
| 641 | 
         
             
                        reference_text="",
         
     | 
| 642 | 
         
             
                        max_new_tokens=0,
         
     | 
| 643 | 
         
            +
                        chunk_length=200,
         
     | 
| 644 | 
         
             
                        top_p=0.7,
         
     | 
| 645 | 
         
             
                        repetition_penalty=1.2,
         
     | 
| 646 | 
         
             
                        temperature=0.7,
         
     | 
    	
        tools/llama/generate.py
    CHANGED
    
    | 
         @@ -250,9 +250,11 @@ def generate( 
     | 
|
| 250 | 
         
             
                device, dtype = prompt.device, prompt.dtype
         
     | 
| 251 | 
         
             
                with torch.device(device):
         
     | 
| 252 | 
         
             
                    model.setup_caches(
         
     | 
| 253 | 
         
            -
             
     | 
| 
         | 
|
| 
         | 
|
| 254 | 
         
             
                    )
         
     | 
| 255 | 
         
            -
             
     | 
| 256 | 
         
             
                codebook_dim = 1 + model.config.num_codebooks
         
     | 
| 257 | 
         
             
                # create an empty tensor of the expected final shape and fill in the current tokens
         
     | 
| 258 | 
         
             
                empty = torch.empty((codebook_dim, T_new), dtype=dtype, device=device)
         
     | 
| 
         | 
|
| 250 | 
         
             
                device, dtype = prompt.device, prompt.dtype
         
     | 
| 251 | 
         
             
                with torch.device(device):
         
     | 
| 252 | 
         
             
                    model.setup_caches(
         
     | 
| 253 | 
         
            +
                            max_batch_size=1,
         
     | 
| 254 | 
         
            +
                            max_seq_len=model.config.max_seq_len,
         
     | 
| 255 | 
         
            +
                            dtype=next(model.parameters()).dtype,
         
     | 
| 256 | 
         
             
                    )
         
     | 
| 257 | 
         
            +
                    
         
     | 
| 258 | 
         
             
                codebook_dim = 1 + model.config.num_codebooks
         
     | 
| 259 | 
         
             
                # create an empty tensor of the expected final shape and fill in the current tokens
         
     | 
| 260 | 
         
             
                empty = torch.empty((codebook_dim, T_new), dtype=dtype, device=device)
         
     |