[Bug]the Results of 31 classifications on the HRSC2016 dataset

[foreverfire01]

Prerequisite

• I have searched Issues and Discussions but cannot get the expected help.
• I have read the FAQ documentation but cannot get the expected help.
• The bug has not been fixed in the latest version (master) or latest version (1.x).

Task

I'm using the official example scripts/configs for the officially supported tasks/models/datasets.

Branch

master branch https://github.com/open-mmlab/mmrotate

Environment

Has anyone tested the results of 31 classifications on the HRSC2016 dataset using MMRotate? I tested it with the original program and the results were quite outrageous

Reproduces the problem - code sample

The result of R3De:
classwise=False, num_classes=1, lr=0.0025, 36 epochs:
mAP=47.94 AP50=87 AP75=46.7

classwise=True, num_classes=31,  lr=0.0025, 36 epochs:
    mAP=8.76  AP50=17.3  AP75=7.8

Reproduces the problem - command or script

No script.

Reproduces the problem - error message

No Error.

Additional information

No response