README Results shows wrong model for SED

[NumesSanguis]

Thank you for providing this tool!

In the README, you have the following output:

------ Sound event detection ------
Checkpoint path: /root/panns_data/Cnn14_mAP=0.431.pth
GPU number: 1
Save fig to appendixes/sed_result.pdf

However, it should not be possible to use SED with Cnn14, as the model only has 'clipwise_output'.

When I execute python3 example.py, it uses Cnn14_DecisionLevelMax for the SED task, which does have 'framewise_output'.

I assume that SED with Cnn14 is not possible right?

[qiuqiangkong]

Cnn14 is a tagging model and does not support SED, and the tagging result is not calculated from SED results.