open-mmlab / mmselfsup

OpenMMLab Self-Supervised Learning Toolbox and Benchmark
https://mmselfsup.readthedocs.io/en/latest/
Apache License 2.0
3.2k stars 432 forks source link

train simclr on cumstom dataset #752

Open alaa-shubbak opened 1 year ago

alaa-shubbak commented 1 year ago

Checklist

  1. I have searched related issues but cannot get the expected help.yes
  2. I have read the FAQ documentation but cannot get the expected help.yse

i want to train the model simclr on my own custom dataset. my dataset contains of two directory: images annotation as meta data , only image name and class

I ran this command in linux terminal : bash tools/dist_train.sh configs/selfsup/simclr/simclr_resnet50_8xb32-coslr-200e_ACID.py 1 --work-dir work_dirs/simclr_resnet50_acid/

and got this error : RuntimeError: Address already in use

then i ran this command : python tools/train.py configs/selfsup/simclr/simclr_resnet50_8xb32-coslr-200e_ACID.py --work-dir work_dirs/simclr_resnet50_acid/

error was: KeyError: 'SelfSupVisualizer is not in the visualizer registry. Please check whether the value ofSelfSupVisualizer` is correct or it was registered as expected. More details can be found at https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#import-the-custom-module' as seen in this picture :

image

`any help please.

alaa-shubbak commented 1 year ago

is it fine if i have the mmcv version of 2.0.0 not 2.0.0rc1? what about the version of mmcls , mmdet , mmseg and mmpose?

alaa-shubbak commented 1 year ago

I try the config with densecl , unfortunately i got the same error message. 'SelfSupVisualizer' is only mentioned in the default_runtime.py .

would it be fine if i remove it , comment the line num 23 and 24 in this config file? or what shall i do?