About the code

[Wz1h1NG]

Hi,

Thank you for your impressive work on Emotion-LLaMAv2. I've been reading the paper and I am trying to reproduce the evaluation results mentioned in the "Experiments" section.

I noticed that the paper describes a specific parsing method to extract emotion labels from the tags for calculating Accuracy and F1 scores. However, I cannot find the corresponding evaluation script (e.g., eval_emotion_llama_v2.py) in the current repository.

Could you please release the evaluation code or point me to the script that handles the parsing and metric calculation?

Thank you for your help!

[lizheng2267]

我也有同样的问题，在当前的代码库中找不到评估脚本，（例如 eval_emotion_llama_v2.py）。
能否提供评估代码脚本。
感谢您的帮助

[ZebangCheng]

我也有同样的问题，在当前的代码库中找不到评估脚本，（例如 eval_emotion_llama_v2.py）。 能否提供评估代码脚本。 感谢您的帮助

感谢您对emotion-llama-v2项目的关注！
有用我们团队的工作疏忽，相关文件未上传。已经补充上传评测代码：

evaluation/eval_mme_emotion_llama_v2.py
期待您展示您的复现结果。