1. 环境安装和数据准备
运行如下指令进行环境安装:
conda create -n opencompass python=3.10
conda activate opencompass
conda install pytorch==2.1.2 torchvision==0.16.2 torchaudio==2.1.2 pytorch-cuda=12.1 -c pytorch -c nvidia -y
# 注意:一定要先 cd /root
cd /root
git clone -b 0.2.4 https://github.com/open-compass/opencompass
cd opencompass
pip install -e .
apt-get update
apt-get install cmake
pip install -r requirements.txt
pip install protobuf
运行以下指令解压评测数据集到 /root/opencompass/data/ 处:
cp /share/temp/datasets/OpenCompassData-core-20231110.zip /root/opencompass/
unzip OpenCompassData-core-20231110.zip
解压指令的部分输出截图如下所示:

输入以下指令列出所有跟 InternLM 及 C-Eval 相关的配置:
python tools/list_configs.py internlm ceval
程序输出截图如下:


2. 启动评测
打开 opencompass文件夹下configs/models/hf_internlm/的hf_internlm2_chat_1_8b.py ,输入指定代码截图如下所示:

输入以下指令评测 InternLM2-Chat-1.8B 模型在 C-Eval 数据集上的性能:
#环境变量配置
export MKL_SERVICE_FORCE_INTEL=1
python run.py --datasets ceval_gen --models hf_internlm2_chat_1_8b --debug
运行结果如下所示:
08/31 11:42:34 - OpenCompass - INFO - time elapsed: 13944.92s
08/31 11:42:39 - OpenCompass - INFO - Partitioned into 52 tasks.
08/31 11:42:42 - OpenCompass - INFO - Task [internlm2-1.8b-hf/ceval-computer_network]: {'accuracy': 47.371421052631575}
08/31 11:42:44 - OpenCompass - INFO - Task [internlm2-1.8b-hf/ceval-operating_system]: {'accuracy': 47.364421052631575}
08/31 11:42:46 - OpenCompass - INFO - Task [internlm2-1.8b-hf/ceval-computer_architecture]: {'accuracy': 23.799523809523807}
08/31 11:42:48 - OpenCompass - INFO - Task [internlm2-1.8b-hf/ceval-college_programming]: {'accuracy': 27.019027027027028}
08/31 11:42:50 - OpenCompass - INFO - Task [internlm2-1.8b-hf/ceval-college_physics]: {'accuracy': 42.09526315789473}
08/31 11:42:52 - OpenCompass - INFO - Task [internlm2-1.8b-hf/ceval-college_chemistry]: {'accuracy': 37.5}
08/31 11:42:54 - OpenCompass - INFO - Task [internlm2-1.8b-hf/ceval-advanced_mathematics]: {'accuracy': 26.31578947368421}
08/31 11:42:56 - OpenCompass - INFO - Task [internlm2-1.8b-hf/ceval-probability_and_statistics]: {'accuracy': 22.22222222222222}
08/31 11:42:58 - Op

最低0.47元/天 解锁文章
395

被折叠的 条评论
为什么被折叠?



