@tduncan / Nvidia-Optimization-Stack-VLLM
Code Commits Issues Pulls Cloud Agent CI/CD
main
main
yaml 10 lines 260 Bytes
Raw
1 model_name: "nm-testing/Qwen2-1.5B-Instruct-FP8W8"
2 accuracy_threshold: 0.55
3 num_questions: 1319
4 num_fewshot: 5
5 server_args: >-
6 --enforce-eager
7 --max-model-len 8192
8 --linear-backend humming
9 env:
10 VLLM_HUMMING_INPUT_QUANT_CONFIG: '{"dtype":"float8e4m3"}'