A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 14, 05:30 UTC
[Misc] method name typo fix (#23042)
Signed-off-by: Andy Xie <andy.xning@gmail.com>
Ning Xie committed
Aug 17, 2025 at 12:49 UTC
87f48623a537d379284bb3e3d1b23ab0ee2af1c1
1 file changed
+2
-2
vllm/v1/worker/cpu_model_runner.py
+2
-2
@@ -29,7 +29,7 @@ class CPUModelRunner(GPUModelRunner):
29
self.use_cuda_graph = False
30
self.cascade_attn_enabled = False
31
32
- self._postprocess_tenosrs()
32
+ self._postprocess_tensors()
33
34
def _may_reorder_batch(self, scheduler_output: "SchedulerOutput") -> None:
35
"""
@@ -59,7 +59,7 @@ class CPUModelRunner(GPUModelRunner):
59
self.attn_groups[0][0].metadata_builder.reorder_batch(
60
self.input_batch, scheduler_output)
61
62
- def _postprocess_tenosrs(self) -> None:
62
+ def _postprocess_tensors(self) -> None:
63
# Note: replace device tensors with cpu tensors
64
def replace_tensor(obj: Any, cpu_attr_name: str,
65
device_attr_name) -> None: