Ming Li

Any plans to support DistributedDataParallel?

2

Hi @npielawski @amjltc295 , thanks for your contribution, are there any plans to support DistributedDataParallel training?

unrecognized arguments: --local_rank

1

When I run the example ``` mim train mmcls resnet101_b16x8_cifar10.py --work-dir tmp --gpus 4 --launcher pytorch ``` An error happened: ``` Training command is python -m torch.distributed.launch --nproc_per_node=4 --master_port=27803 /home/tiger/.local/lib/python3.7/site-packages/mmcls/.mim/tools/train.py...

Any plans to support DALI?

3

For some tasks (e.g. contrastive learning), data augmentation is very important and data augmentation on CPU is often very slow, are there any plans to support the use of DALI...

Hi, @chhluo @BIGWangYuDong @Czm369 By adding semantic_head, we can introduce semantic segmentation tasks at training time, however there seems to be no tutorial and config that let us output semantic...

How to perform semantic segmentation with semantic_head?

1

Hi, @chhluo @BIGWangYuDong By adding semantic_head, we can introduce semantic segmentation tasks at training time, however there seems to be no tutorial and config that let us output semantic segmentation...

How could I change the save dir?

2

I tried to use NNI to search hyper-parameters, however, after each experiment, **there are always some files or directories under the root path**. Like `trial_cfg_jGsUE.py` and `trial_cfg_FvJ5P` How could I...

FileNotFoundError: [Errno 2] No such file or directory: './data/imagenet/train_map.txt'

3

Hi @ancientmooner @zdaxie , Thanks for your contribution, but why there need a txt file to load ImageNet? How can I get it?

Hi @mrT23 @panchonoy Thanks for your contribution, but I still think it is inconvenience for users, could you please provide the `train.txt`, `val.txt` files like this ? (Constructed by image...

Where is the code for MC-Dropout?

1

Hi @layumi , I noticed that in your paper, the uncertainty is obtained by computing both MC-Dropout and difference between two classifiers, but I cannot find the code used for...