fix: resolve IndexError when using spk model and the audio contains only 1 ...

Atomie CHEN

2024-03-25 ff83116369ba6afe56430cb0b9d977fbe7f4bced

 funasr/datasets/llm_datasets_vicuna/datasets.py

@@ -140,7 +140,7 @@
        
        for key, data_list in outputs.items():
            if isinstance(data_list[0], torch.Tensor):
                if data_list[0].dtype == torch.int64:
                if data_list[0].dtype == torch.int64 or data_list[0].dtype == torch.int32:
                    
                    pad_value = self.int_pad_value
                else: