采用 Openai server 测试，能够产生 inference 结果，但是 eval 输出结果为空

[HaoWuSR]

问题描述 / Issue Description

请简要描述您遇到的问题。 / Please briefly describe the issue you encountered.

使用的工具 / Tools Used

• Native / 原生框架
• [✅] Opencompass backend
• VLMEvalKit backend
• RAGEval backend
• Perf / 模型推理压测工具
• Arena /竞技场模式

执行的代码或指令 / Code or Commands Executed

请提供您执行的主要代码或指令。 / Please provide the main code or commands you executed. 例如 / For example:

from evalscope.run import run_task
from evalscope.summarizer import Summarizer

task_cfg_dict = dict(
    eval_backend='OpenCompass',
    eval_config={
        'datasets': ['gsm8k'],
        'models': [
            {'path': '/workspace/models/Llama-2-13b-chat-hf', 
            'openai_api_base': 'http://127.0.0.1:8008/v1/chat/completions', 
            'is_chat': True,
            'batch_size': 16},
        ],
        'work_dir': 'outputs/llama-2-13b-chat-hf',
        'limit': None,
        },
    )

def run_eval():
    # 选项 1: python 字典
    task_cfg = task_cfg_dict

    # 选项 2: yaml 配置文件
    # task_cfg = 'eval_openai_api.yaml'

    # 选项 3: json 配置文件
    # task_cfg = 'eval_openai_api.json'
    # print(task_cfg)

    run_task(task_cfg=task_cfg)

    print('>> Start to get the report with summarizer ...')
    report_list = Summarizer.get_report_from_cfg(task_cfg)
    print(f'\n>> The report list: {report_list}')

run_eval()

# 例如：在终端中执行的指令 / Terminal command executed
python script.py

错误日志 / Error Log

请粘贴完整的错误日志或控制台输出。 / Please paste the full error log or console output. 例如 / For example:

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
dataset                                 version    metric    mode    /workspace/models/Llama-2-13b-chat-hf
--------------------------------------  ---------  --------  ------  ---------------------------------------
--------- 考试 Exam ---------           -          -         -       -
ceval                                   -          -         -       -
agieval                                 -          -         -       -
mmlu                                    -          -         -       -
GaokaoBench                             -          -         -       -
ARC-c                                   -          -         -       -
--------- 语言 Language ---------       -          -         -       -
WiC                                     -          -         -       -
summedits                               -          -         -       -
chid-dev                                -          -         -       -
afqmc-dev                               -          -         -       -
bustm-dev                               -          -         -       -
cluewsc-dev                             -          -         -       -
WSC                                     -          -         -       -
winogrande                              -          -         -       -
flores_100                              -          -         -       -
--------- 知识 Knowledge ---------      -          -         -       -
BoolQ                                   -          -         -       -
commonsense_qa                          -          -         -       -
nq                                      -          -         -       -
triviaqa                                -          -         -       -
--------- 推理 Reasoning ---------      -          -         -       -
cmnli                                   -          -         -       -
ocnli                                   -          -         -       -
ocnli_fc-dev                            -          -         -       -
AX_b                                    -          -         -       -
AX_g                                    -          -         -       -
CB                                      -          -         -       -
RTE                                     -          -         -       -
story_cloze                             -          -         -       -
COPA                                    -          -         -       -
ReCoRD                                  -          -         -       -
hellaswag                               -          -         -       -
piqa                                    -          -         -       -
siqa                                    -          -         -       -
strategyqa                              -          -         -       -
math                                    -          -         -       -
gsm8k                                   -          -         -       -
TheoremQA                               -          -         -       -
openai_humaneval                        -          -         -       -
mbpp                                    -          -         -       -
bbh                                     -          -         -       -
--------- 理解 Understanding ---------  -          -         -       -
C3                                      -          -         -       -
CMRC_dev                                -          -         -       -
DRCD_dev                                -          -         -       -
MultiRC                                 -          -         -       -
race-middle                             -          -         -       -
race-high                               -          -         -       -
openbookqa_fact                         -          -         -       -
csl_dev                                 -          -         -       -
lcsts                                   -          -         -       -
Xsum                                    -          -         -       -
eprstmt-dev                             -          -         -       -
lambada                                 -          -         -       -
tnews-dev                               -          -         -       -

运行环境 / Runtime Environment

• 操作系统 / Operating System:

  • Windows
  • macOS
  • [✅] Ubuntu

• Python版本 / Python Version:

  • 3.11
  • [ ✅] 3.10
  • 3.9

其他信息 / Additional Information

如果有其他相关信息，请在此处提供。 / If there is any other relevant information, please provide it here.

[HaoWuSR]

下面是模型路径下的 out 文件
`11/21 15:20:16 - OpenCompass - INFO - Task [/workspace/models/Llama-2-13b-chat-hf/gsm8k]
11/21 15:20:24 - OpenCompass - WARNING - Max Completion tokens for /workspace/models/Llama-2-13b-chat-hf is :16384
11/21 15:20:26 - OpenCompass - INFO - Start inferencing [/workspace/models/Llama-2-13b-chat-hf/gsm8k]
[2024-11-21 15:20:27,222] [opencompass.openicl.icl_inferencer.icl_gen_inferencer] [INFO] Starting build dataloader
[2024-11-21 15:20:27,222] [opencompass.openicl.icl_inferencer.icl_gen_inferencer] [INFO] Starting inference process...

0%| | 0/83 [00:00<?, ?it/s]
1%| | 1/83 [00:12<17:28, 12.78s/it]
2%|▏ | 2/83 [00:24<16:26, 12.18s/it]
4%|▎ | 3/83 [00:36<16:00, 12.01s/it]
5%|▍ | 4/83 [00:48<15:58, 12.14s/it]
6%|▌ | 5/83 [01:01<16:10, 12.44s/it]
7%|▋ | 6/83 [01:14<16:03, 12.51s/it]
8%|▊ | 7/83 [01:26<15:44, 12.43s/it]
10%|▉ | 8/83 [01:38<15:17, 12.24s/it]
11%|█ | 9/83 [01:50<15:09, 12.29s/it]
12%|█▏ | 10/83 [02:03<15:08, 12.45s/it]
13%|█▎ | 11/83 [02:16<15:01, 12.53s/it]
14%|█▍ | 12/83 [02:28<14:42, 12.43s/it]
16%|█▌ | 13/83 [02:41<14:47, 12.68s/it]
17%|█▋ | 14/83 [02:53<14:09, 12.31s/it]
18%|█▊ | 15/83 [03:05<13:47, 12.17s/it]
19%|█▉ | 16/83 [03:17<13:36, 12.19s/it]
20%|██ | 17/83 [03:29<13:24, 12.19s/it]
22%|██▏ | 18/83 [03:43<13:41, 12.64s/it]
23%|██▎ | 19/83 [03:56<13:32, 12.70s/it]
24%|██▍ | 20/83 [04:08<13:07, 12.50s/it]
25%|██▌ | 21/83 [04:20<12:49, 12.41s/it]
27%|██▋ | 22/83 [04:32<12:31, 12.32s/it]
28%|██▊ | 23/83 [04:44<12:10, 12.18s/it]
29%|██▉ | 24/83 [04:56<12:05, 12.29s/it]
30%|███ | 25/83 [05:10<12:09, 12.59s/it]
31%|███▏ | 26/83 [05:22<12:00, 12.65s/it]
33%|███▎ | 27/83 [05:36<12:00, 12.86s/it]
34%|███▎ | 28/83 [05:49<11:48, 12.87s/it]
35%|███▍ | 29/83 [06:02<11:40, 12.97s/it]
36%|███▌ | 30/83 [06:15<11:24, 12.91s/it]
37%|███▋ | 31/83 [06:27<11:05, 12.79s/it]
39%|███▊ | 32/83 [06:39<10:38, 12.52s/it]
40%|███▉ | 33/83 [06:51<10:21, 12.44s/it]
41%|████ | 34/83 [07:04<10:15, 12.56s/it]
42%|████▏ | 35/83 [07:17<10:06, 12.64s/it]
43%|████▎ | 36/83 [07:29<09:42, 12.38s/it]
45%|████▍ | 37/83 [07:40<09:18, 12.15s/it]
46%|████▌ | 38/83 [07:53<09:16, 12.37s/it]
47%|████▋ | 39/83 [08:06<09:11, 12.52s/it]
48%|████▊ | 40/83 [08:18<08:53, 12.41s/it]
49%|████▉ | 41/83 [08:31<08:42, 12.43s/it]
51%|█████ | 42/83 [08:44<08:38, 12.64s/it]
52%|█████▏ | 43/83 [08:55<08:14, 12.35s/it]
53%|█████▎ | 44/83 [09:08<08:06, 12.47s/it]
54%|█████▍ | 45/83 [09:20<07:47, 12.30s/it]
55%|█████▌ | 46/83 [09:33<07:43, 12.51s/it]
57%|█████▋ | 47/83 [09:46<07:31, 12.54s/it]
58%|█████▊ | 48/83 [09:59<07:24, 12.71s/it]
59%|█████▉ | 49/83 [10:11<07:03, 12.45s/it]
60%|██████ | 50/83 [10:23<06:53, 12.54s/it]
61%|██████▏ | 51/83 [10:37<06:46, 12.71s/it]
63%|██████▎ | 52/83 [10:49<06:28, 12.52s/it]
64%|██████▍ | 53/83 [11:02<06:22, 12.73s/it]
65%|██████▌ | 54/83 [11:14<06:04, 12.57s/it]
66%|██████▋ | 55/83 [11:26<05:44, 12.31s/it]
67%|██████▋ | 56/83 [11:38<05:35, 12.42s/it]
69%|██████▊ | 57/83 [11:50<05:18, 12.27s/it]
70%|██████▉ | 58/83 [12:03<05:08, 12.33s/it]
71%|███████ | 59/83 [12:16<05:01, 12.56s/it]
72%|███████▏ | 60/83 [12:28<04:44, 12.36s/it]
73%|███████▎ | 61/83 [12:40<04:30, 12.30s/it]
75%|███████▍ | 62/83 [12:52<04:16, 12.21s/it]
76%|███████▌ | 63/83 [13:05<04:07, 12.39s/it]
77%|███████▋ | 64/83 [13:17<03:56, 12.47s/it]
78%|███████▊ | 65/83 [13:30<03:46, 12.59s/it]
80%|███████▉ | 66/83 [13:42<03:28, 12.29s/it]
81%|████████ | 67/83 [13:54<03:16, 12.26s/it]
82%|████████▏ | 68/83 [14:07<03:06, 12.46s/it]
83%|████████▎ | 69/83 [14:20<02:55, 12.51s/it]
84%|████████▍ | 70/83 [14:32<02:42, 12.53s/it]
86%|████████▌ | 71/83 [14:44<02:29, 12.43s/it]
87%|████████▋ | 72/83 [14:57<02:17, 12.53s/it]
88%|████████▊ | 73/83 [15:09<02:03, 12.39s/it]
89%|████████▉ | 74/83 [15:21<01:51, 12.34s/it]
90%|█████████ | 75/83 [15:35<01:40, 12.59s/it]
92%|█████████▏| 76/83 [15:47<01:28, 12.65s/it]
93%|█████████▎| 77/83 [16:01<01:16, 12.78s/it]
94%|█████████▍| 78/83 [16:14<01:04, 12.99s/it]
95%|█████████▌| 79/83 [16:27<00:52, 13.05s/it]
96%|█████████▋| 80/83 [16:40<00:39, 13.01s/it]
98%|█████████▊| 81/83 [16:52<00:25, 12.69s/it]
99%|█████████▉| 82/83 [17:04<00:12, 12.59s/it]
100%|██████████| 83/83 [17:12<00:00, 11.18s/it]
100%|██████████| 83/83 [17:12<00:00, 12.44s/it]
11/21 15:37:40 - OpenCompass - INFO - time elapsed: 1043.18s
`

[Yunnglin]

请执行pip list | grep ms-opencompass看一下opencomapss版本