Skip to content

[Bug] qpqa 测试卡住 #580

Closed
Closed

Description

@shiyuan680

Operating System and Version

ubuntu

Python Environment for Tool Installation

Python virtual environment created with Anaconda/Miniconda

Python Version

3.10

AISBench Tool Version

v3.1-20260912-master

AISBench Execution Command

ais_bench --models vllm_api_stream_chat --datasets gpqa_gen_0_shot_cot_chat_prompt --num-warmups 0

Model Configuration File or Custom Configuration File Content

vllm_api_stream_chat.py
from ais_bench.benchmark.models import VLLMCustomAPIChat
from ais_bench.benchmark.utils.postprocess.model_postprocessors import extract_non_reasoning_content

models = [
dict(
attr="service",
type=VLLMCustomAPIChat,
abbr="vllm-api-stream-chat",
path="/mnt/share/DeepSeek-V4.1-Flash-bf16/",
model="deepseek-v41-engram",
stream=True,
request_rate=0,
use_timestamp=False,
retry=2,
api_key="",
host_ip="localhost",
host_port=8000,
url="",
max_out_len=98304,
batch_size=1024,
trust_remote_code=False,
generation_kwargs=dict(
temperature=1,
top_p=0.95,
chat_template_kwargs=dict(reasoning_effort="low", enable_thinking=True),
),
pred_postprocessor=dict(type=extract_non_reasoning_content),
)
]

Expected Behavior

No response

Actual Behavior

一直卡在145不动,然后前端服务有推理
Base path of result&log : outputs/default/20260929_113620
Task Progress Table (Updated at: 2026-09-29 12:59:41)
Page: 1/1 Total 2 rows of data
Press Up/Down arrow to page, 'P' to PAUSE/RESUME screen refresh, 'Ctrl + C' to exit

+-----------------------------------+-----------+-----------------------------------------------------+-------------+-------------+--------------------------------------------
| Task Name | Process | Progress | Time Cost | Status | Log Path
+===================================+===========+=====================================================+=============+=============+============================================
| vllm-api-stream-chat/GPQA_diamond | 1142811 | [##################### ] 145/198 [0.3 it/s] | 1:22:59 | inferencing | logs/infer/vllm-api-stream-chat/GPQA_diamon
+-----------------------------------+-----------+-----------------------------------------------------+-------------+-------------+--------------------------------------------
------+------------------------------------------------------+

Pre-Checks

  • I have read the Quick Start section in the homepage documentation and cannot resolve the issue
  • I have searched the FAQ and found no duplicate issues
  • I have searched existing Issues and found no duplicate issues
  • I have updated to the latest version, and the issue still persists

Activity

  1. github-actions commented on Sep 29, 2026

    @github-actions

    🌟欢迎给AISBench benchmark测评工具提issue,您可以先尝试以下途径寻求issue的解决方法:

    1. 【强烈推荐❤️‍🔥】确保issue描述完整后,可以试着将issue交给Ask DeepWiki回答,deepwiki包含了和工具相关的所有知识库
    2. 在 工具历史issue 中通过问题的关键日志片段或关键词检索相似issue
    3. 在 官方文档 的搜索栏中通过问题的关键日志片段或关键词检索
    4. 寻求仓库维护者的帮助(issue评论区中 @SJTUyh)
  2. github-actions commented on Sep 29, 2026

    @github-actions

    🤖 基于AI机器人的issue内容完整性检查结果:

    问题内容检查结果

    检查状态

    PASS

    详细检查

    • 操作系统及版本: COMPLETE
    • 安装工具的python环境: COMPLETE
    • python版本: COMPLETE
    • AISBench工具版本: COMPLETE
    • AISBench执行命令: COMPLETE
    • 模型配置文件或自定义配置文件内容: COMPLETE
    • 实际行为: COMPLETE

    改进建议

    所有内容已完备!

    补充说明

    该问题报告包含了所有必需的章节,并且每个章节都提供了清晰、完整且相关的信息,有助于理解和复现问题。

    👉 如果想重新检查,在评论区@issue_checker即可。

    【强烈推荐❤️‍🔥】确保issue描述完整后,可以试着将issue交给Ask DeepWiki回答,deepwiki包含了和工具相关的所有知识库

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't workingcontent_check_passedissue content check passed

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions