biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Doc]: fix typos in various files (#28811)

Browse Source 
Signed-off-by: Didier Durand <durand.didier@gmail.com>
This commit is contained in:
Didier Durand
2025-11-16 15:30:06 +01:00
committed by GitHub
parent 8d259fad6c
commit 63fed55506
7 changed files with 10 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
docs/design/cuda_graphs.md
View File

@@ -128,7 +128,7 @@ A [CUDAGraphWrapper][vllm.compilation.cuda_graph.CUDAGraphWrapper] instance wrap
3. Otherwise, i.e., the runtime_mode matches the mode of the wrapper, the wrapper will perform CUDA Graphs capture (if key does not exist, create
a new entry and cache it) or replay (if key exists in the cache).
The above steps are based on the assumption that the CUDA Graphs wrapper would directly trust what’s in the forward context (controlled by the dispatcher). This lets us simplify and cenralize the logic, reducing the complexity as well as the risk of mismatched state between the wrappers and the dispatcher. It also allows reusing the wrapper class for both `FULL` and `PIECEWISE` runtime modes. See the implementation [here](https://github.com/vllm-project/vllm/blob/f751e50b7a2aae3110d83ed0d88202fc91b3e78a/vllm/compilation/cuda_graph.py#L106).
The above steps are based on the assumption that the CUDA Graphs wrapper would directly trust what’s in the forward context (controlled by the dispatcher). This lets us simplify and centralize the logic, reducing the complexity as well as the risk of mismatched state between the wrappers and the dispatcher. It also allows reusing the wrapper class for both `FULL` and `PIECEWISE` runtime modes. See the implementation [here](https://github.com/vllm-project/vllm/blob/f751e50b7a2aae3110d83ed0d88202fc91b3e78a/vllm/compilation/cuda_graph.py#L106).
#### Nested Wrapper design