Evaluation of Self Supervised Pre training Model: "Detecting" Problems in Downstream Tasks

[majintao1994]

After I have extracted the weights of the pre training model, I need to ask the following questions when using this backone.pth to train the downstream detection tasks. What's wrong? Thank you

error：
Traceback (most recent call last):
File "/home/work/mmselftest/mmself/tools/train.py", line 187, in
main()
File "/home/work/mmselftest/mmself/tools/train.py", line 161, in main
test_cfg=cfg.get('test_cfg'))
File "/home/work/mmselftest/mmself/mmdet/models/builder.py", line 77, in build_detector
return build(cfg, DETECTORS, dict(train_cfg=train_cfg, test_cfg=test_cfg))
File "/home/work/mmselftest/mmself/mmdet/models/builder.py", line 34, in build
return build_from_cfg(cfg, registry, default_args)
File "/home/ty/anaconda3/envs/trt2/lib/python3.7/site-packages/mmcv/utils/registry.py", line 182, in build_from_cfg
raise type(e)(f'{obj_cls.name}: {e}')
TypeError: FasterRCNN: ResNet: init() got an unexpected keyword argument 'init_cfg'
my config file;

base = 'faster_rcnn_r50_fpn_1x_coco.py'

model = dict(
    backbone=dict(
        frozen_stages=-1,
        init_cfg=dict(type='Pretrained', checkpoint='checkpoints/backbone_200.pth')
))

# optimizer
optimizer = dict(
lr=0.02 * (1 / 8)
)

# Runner type
runner = dict(delete=True, type='IterBasedRunner', max_iters=9000)

checkpoint_config = dict(interval=3000)

evaluation = dict(interval=3000)

[fangyixiao18]

Hi, could you please provide more information? like command, package version, etc. Thanks

[mzr1996]

It looks like your mmdetection is outdated