谢昕辰
Refer to #1883 And might take a look at this discussion. https://github.com/open-mmlab/mmsegmentation/issues/427
Hi, thanks for your nice pr. I've left some comments and we coiuld continue to revise and discuss.
Hi @Jinshan99, Might try `cfg.evaluation.save_best='mIoU'`. #1650
Sorry, I don't understand your question very well, could you describe it more clearly? If your machine doesn't support training with an auxiliary head, you might remove it and it...
Hi @HaoGuo98, You might add `--show-dir` argument to the test command. https://mmsegmentation.readthedocs.io/en/latest/inference.html#test-a-dataset
We specify in the document that `--show-dir` only supports the Single GPU test, you could try it again with a single GPU. ```shell python tools/test.py configs/dnlnet/dnlnet_r101-d8_512x512_80k_ade20k.py work_dirs/dnl_ade.pth --show-dir ade_test ```
> We specify in the document that `--show-dir` only supports the Single GPU test, you could try it again with a single GPU. > > ```shell > python tools/test.py configs/dnlnet/dnlnet_r101-d8_512x512_80k_ade20k.py...
Hi @cm0561, Could you provide the model and the command you used if possible? I can't reproduce this error.
Hi @RunningLeon, @zhouzaida, Could you take a look at this issue if you're available?
Hi @cunese24, The annotation files are the training label. If you would like to use TXT files as the label, you need to customize a new Dataset Class.