fix cuddagraph issue

This commit is contained in:
sirius0xdev 2026-05-06 02:08:48 +00:00
parent 2102ea882f
commit bb86f0fc3f

View file

@ -37,7 +37,7 @@ spec:
- --port=8000
- --tensor-parallel-size=1
- --tokenizer-mode=hf
- --gpu-memory-utilization=0.98
- --gpu-memory-utilization=0.95
- --max-model-len=131072
- --enable-auto-tool-choice
- --kv-cache-dtype=fp8