Don't fill tensor with 0

Signed-off-by: Hui Gao <[email protected]>

Author: HuiGao-NV

tensorrt_llm/_torch/memory_buffer_utils.py

@@ -91,7 +91,7 @@ def get_buffer(self, tensor_shape: list[int], dtype: torch.dtype,
         new_buffer_tensor = None
         try:
             with torch.cuda.memory.use_mem_pool(get_shared_pool()):
-                new_buffer_tensor = torch.zeros((required_memory_size, ),
+                new_buffer_tensor = torch.empty((required_memory_size, ),
                                                 device='cuda',
                                                 dtype=torch.uint8)
         except Exception as ex: