[Fix] fix Qwen2-Audio-7B-Instruct accuracy test (#4017)

### What this PR does / why we need it?

fix Qwen2-Audio-7B-Instruct accuracy test

### Does this PR introduce _any_ user-facing change?

### How was this patch tested?

- vLLM version: v0.11.0
- vLLM main:
https://github.com/vllm-project/vllm/commit/83f478bb19489b41e9d208b47b4bb5a95ac171ac

Signed-off-by: hfadzxy <starmoon_zhang@163.com>
This commit is contained in:
zhangxinyuehfad authored and GitHub committed 2025-11-10 11:54:18 +08:00
1 parent de49fb3deb
commit d40ba52454
3 files changed
+3 -4

No files matched your search

@@ -108,8 +108,7 @@ jobs:
model_list:
- Qwen3-8B
- Qwen2.5-VL-7B-Instruct
# TODO: This model has a bug that needs to be fixed and readded
# - Qwen2-Audio-7B-Instruct
- Qwen2-Audio-7B-Instruct
- Qwen3-8B-W8A8
- Qwen3-VL-8B-Instruct
- Qwen2.5-Omni-7B
@@ -64,8 +64,7 @@ jobs:
model_list:
- Qwen3-8B
- Qwen2.5-VL-7B-Instruct
# TODO: This model has a bug that needs to be fixed and readded
# - Qwen2-Audio-7B-Instruct
- Qwen2-Audio-7B-Instruct
- runner: linux-aarch64-a2-2
model_list:
- Qwen3-30B-A3B
@@ -8,3 +8,4 @@ tasks:
- name: "exact_match,flexible-extract"
value: 0.45
num_fewshot: 5
gpu_memory_utilization: 0.8