[Misc] Use torch.Tensor for type annotation (vllm-project#6505)

benchmarks/cutlass_benchmarks/w8a8_benchmarks.py

@@ -20,18 +20,18 @@
 # helpers
 
 
-def to_fp8(tensor: torch.tensor) -> torch.tensor:
+def to_fp8(tensor: torch.Tensor) -> torch.Tensor:
     finfo = torch.finfo(torch.float8_e4m3fn)
     return torch.round(tensor.clamp(
         min=finfo.min, max=finfo.max)).to(dtype=torch.float8_e4m3fn)
 
 
-def to_int8(tensor: torch.tensor) -> torch.tensor:
+def to_int8(tensor: torch.Tensor) -> torch.Tensor:
     return torch.round(tensor.clamp(min=-128, max=127)).to(dtype=torch.int8)
 
 
 def make_rand_tensors(dtype: torch.dtype, m: int, n: int,
-                      k: int) -> Tuple[torch.tensor, torch.tensor]:
+                      k: int) -> Tuple[torch.Tensor, torch.Tensor]:
 
     a = torch.randn((m, k), device='cuda') * 5
     b = torch.randn((n, k), device='cuda').t() * 5
@@ -47,25 +47,25 @@ def make_rand_tensors(dtype: torch.dtype, m: int, n: int,
 # impl
 
 
-def pytorch_mm_impl(a: torch.tensor, b: torch.tensor, scale_a: torch.tensor,
-                    scale_b: torch.tensor,
-                    out_dtype: torch.dtype) -> torch.tensor:
+def pytorch_mm_impl(a: torch.Tensor, b: torch.Tensor, scale_a: torch.Tensor,
+                    scale_b: torch.Tensor,
+                    out_dtype: torch.dtype) -> torch.Tensor:
     return torch.mm(a, b)
 
 
-def pytorch_fp8_impl(a: torch.tensor, b: torch.tensor, scale_a: torch.tensor,
-                     scale_b: torch.tensor,
-                     out_dtype: torch.dtype) -> torch.tensor:
+def pytorch_fp8_impl(a: torch.Tensor, b: torch.Tensor, scale_a: torch.Tensor,
+                     scale_b: torch.Tensor,
+                     out_dtype: torch.dtype) -> torch.Tensor:
     return torch._scaled_mm(a,
                             b,
                             scale_a=scale_a,
                             scale_b=scale_b,
                             out_dtype=out_dtype)
 
 
-def pytorch_fp8_impl_fast_accum(a: torch.tensor, b: torch.tensor,
-                                scale_a: torch.tensor, scale_b: torch.tensor,
-                                out_dtype: torch.dtype) -> torch.tensor:
+def pytorch_fp8_impl_fast_accum(a: torch.Tensor, b: torch.Tensor,
+                                scale_a: torch.Tensor, scale_b: torch.Tensor,
+                                out_dtype: torch.dtype) -> torch.Tensor:
     return torch._scaled_mm(a,
                             b,
                             scale_a=scale_a,
@@ -74,15 +74,15 @@ def pytorch_fp8_impl_fast_accum(a: torch.tensor, b: torch.tensor,
                             use_fast_accum=True)
 
 
-def cutlass_impl(a: torch.tensor, b: torch.tensor, scale_a: torch.tensor,
-                 scale_b: torch.tensor,
-                 out_dtype: torch.dtype) -> torch.tensor:
+def cutlass_impl(a: torch.Tensor, b: torch.Tensor, scale_a: torch.Tensor,
+                 scale_b: torch.Tensor,
+                 out_dtype: torch.dtype) -> torch.Tensor:
     return ops.cutlass_scaled_mm(a, b, scale_a, scale_b, out_dtype=out_dtype)
 
 
 # bench
-def bench_fn(a: torch.tensor, b: torch.tensor, scale_a: torch.tensor,
-             scale_b: torch.tensor, out_dtype: torch.dtype, label: str,
+def bench_fn(a: torch.Tensor, b: torch.Tensor, scale_a: torch.Tensor,
+             scale_b: torch.Tensor, out_dtype: torch.dtype, label: str,
              sub_label: str, fn: Callable, description: str) -> TMeasurement:
 
     min_run_time = 1

vllm/worker/worker.py

@@ -105,7 +105,7 @@ def __init__(
         # initialize_cache.
         self.cache_engine: List[CacheEngine]
         # Initialize gpu_cache as embedding models don't initialize kv_caches
-        self.gpu_cache: Optional[List[List[torch.tensor]]] = None
+        self.gpu_cache: Optional[List[List[torch.Tensor]]] = None
 
     def init_device(self) -> None:
         if self.device_config.device.type == "cuda":