EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[Doc] Replace deprecated full_cuda_graph with cudagraph_mode in Qwen2.5-Omni (#7286)

Browse Source 
## Summary
- Replace `full_cuda_graph: 1` with `cudagraph_mode: FULL_DECODE_ONLY`
in both single-NPU and multi-NPU examples
- `full_cuda_graph` is deprecated and falls back to `NONE` on NPU

Fixes #4696
- vLLM version: v0.17.0
- vLLM main:
4034c3d32e

Signed-off-by: bazingazhou233-hub <bazingazhou233-hub@users.noreply.github.com>
Co-authored-by: bazingazhou233-hub <bazingazhou233-hub@users.noreply.github.com>
This commit is contained in:
bazingazhou233-hub
2026-03-14 22:38:36 +08:00
committed by GitHub
parent bb506a1c99
commit 9e6c547d98
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
docs/source/tutorials/models/Qwen2.5-Omni.md
View File

@@ -82,7 +82,7 @@ vllm serve "${MODEL_PATH}" \
--served-model-name Qwen-Omni \
--allowed-local-media-path ${LOCAL_MEDIA_PATH} \
--trust-remote-code \
--compilation-config '{"full_cuda_graph": 1}' \
--compilation-config '{"cudagraph_mode": "FULL_DECODE_ONLY"}' \
--no-enable-prefix-caching
```
@@ -113,7 +113,7 @@ vllm serve ${MODEL_PATH}\
--served-model-name Qwen-Omni \
--allowed-local-media-path ${LOCAL_MEDIA_PATH} \
--trust-remote-code \
--compilation-config {"full_cuda_graph": 1} \
--compilation-config '{"cudagraph_mode": "FULL_DECODE_ONLY"}' \
--data-parallel-size ${DP_SIZE} \
--no-enable-prefix-caching
```