Inconsistent results

[AIRobotZhang]

Hello, Table 1-2 is not consistent with results reported in https://github.com/WeOpenML/PandaLM/tree/main#test-data

[zhuohaoyu]

Hi, we are aware of the issue and it was fixed in #22. The results are different because we updated our inference methods with PandaLM/ChatGPT as reported in our paper to calculate more robust metrics. Please refer to our paper for latest results.

[AIRobotZhang]

Hi, we are aware of the issue and it was fixed in #22. The results are different because we updated our inference methods with PandaLM/ChatGPT as reported in our paper to calculate more robust metrics. Please refer to our paper for latest results.

How calculate the F1 score in Table 2. Is F1 score micro or macro?

[arjunbansal]

The results are different because we updated our inference methods with PandaLM/ChatGPT as reported in our paper to calculate more robust metrics. Please refer to our paper for latest results.

Do you plan to share the inference dataset files that accompany the updated methods?