Multiple data set validation error

[studentyyh]

Prerequisite

• [X] I have searched Issues and Discussions but cannot get the expected help.
• [X] The bug has not been fixed in the latest version(https://github.com/open-mmlab/mmpose).

Environment

torch2.0

Reproduces the problem - code sample

Hello, there was an error when I used the multi-data set validation in mmpose. I used two validation sets, one of which was a self-defined data set with the same format as coco data set, and the other was coco data set, and only coco data set could be evaluated during model evaluation. The code indicates that the results of the second data set is empty. I found through debug that only the coco data set would be loaded when the verification set was loaded, so the results were empty when the second data set was verified because there was no data. I don't know what caused this.

Reproduces the problem - command or script

python tools/train.py

Reproduces the problem - error message

Additional information

No response

[Ben-Louis]

Hi @studentyyh, could you please share your config?

[studentyyh]

Here is the configuration file I used：

[studentyyh]

@Ben-Louis Thanks for your reply. Just now, through checking the source code, I found that the data set name in metainfo should be obtained before processing the input data set. The pipeline I wrote was missing this key, so it could not be read correctly, now I have fixed this problem, thank you.