@tduncan / Nvidia-Optimization-Stack-VLLM
Code Commits Issues Pulls Cloud Agent CI/CD
main
main
yaml 10 lines 229 Bytes
Raw
1 model_name: "Qwen/Qwen3.5-35B-A3B-FP8"
2 accuracy_threshold: 0.79
3 tolerance: 0.03
4 num_questions: 1319
5 num_fewshot: 5
6 server_args: >-
7 --max-model-len 4096
8 --data-parallel-size 2
9 --enable-expert-parallel
10 --kv-cache-dtype fp8