Closed fang-lg closed 5 months ago
- Your error log seems to be caused by the mmcv and mmdetection module. Check the issue of FormatCode() got an unexpected keyword argument 'verify' #3 (comment)
- Please provide more details in your question; Specify which config file you used and what commands you executed.
run: python3 -u tools/train.py configs/OA-DG/dwd/faster_rcnn_r101_dc5_1x_dwd_oadg.py --work-dir /out
yapf==0.40.2,mmcv-full==1.5.0
We are currently using yapf==0.32.0 and mmcv-full==1.5.0. The attached file lists all our installed packages. We will organize and upload the requirements file later. requiremets_oa-dg.txt
We are currently using yapf==0.32.0 and mmcv-full==1.5.0. The attached file lists all our installed packages. We will organize and upload the requirements file later. requiremets_oa-dg.txt
上述问题已经解决,目前已经训练成功了。但是验证的时候报以下错误
Traceback (most recent call last):
File "/home/ailab/HDD/fl/ws/external/tools/test_dwd.py", line 258, in
The error seems to occur when the dataset_type
is not specified in 'configs/base/datasets/s-dgod.py'. Currently, dataset_type
is set to SdgodDataset
.
The error seems to occur when the
dataset_type
is not specified in 'configs/base/datasets/s-dgod.py'. Currently,dataset_type
is set toSdgodDataset
.
已经正确设置了
We run OA-DG directly from the source code. But it seems that you run the code from the library. How about running OA-DG directly from the source code?
We run OA-DG directly from the source code. But it seems that you run the code from the library. How about running OA-DG directly from the source code?
我是直接在程序中运行的,他这个从源码中加载还是从依赖中加载如何区分?
Error log indicates that build_dataset
is run in the "/home/ailab/miniconda3/envs/oadg/lib/python3.9/site-packages/mmdet/datasets/builder.py". So, you are running the code from the library.
Traceback (most recent call last): File "/home/ailab/HDD/fl/ws/external/tools/train.py", line 227, in
main()
File "/home/ailab/HDD/fl/ws/external/tools/train.py", line 98, in main
cfg = Config.fromfile(args.config)
File "/home/ailab/miniconda3/envs/fl/lib/python3.9/site-packages/mmcv/utils/config.py", line 337, in fromfile
cfg_dict, cfg_text = Config._file2dict(filename,
File "/home/ailab/miniconda3/envs/fl/lib/python3.9/site-packages/mmcv/utils/config.py", line 267, in _file2dict
base_cfg_dict = Config._merge_a_into_b(cfg_dict, base_cfg_dict)
File "/home/ailab/miniconda3/envs/fl/lib/python3.9/site-packages/mmcv/utils/config.py", line 326, in _merge_a_into_b
b[k] = Config._merge_a_into_b(v, b[k], allow_list_keys)
File "/home/ailab/miniconda3/envs/fl/lib/python3.9/site-packages/mmcv/utils/config.py", line 326, in _merge_a_into_b
b[k] = Config._merge_a_into_b(v, b[k], allow_list_keys)
File "/home/ailab/miniconda3/envs/fl/lib/python3.9/site-packages/mmcv/utils/config.py", line 320, in _merge_a_into_b
raise TypeError(
TypeError: init_cfg={'type': 'Pretrained', 'checkpoint': 'open-mmlab://detectron2/resnet101_caffe'} in child config cannot inherit from base because init_cfg is a dict in the child config but is of type <class 'NoneType'> in base config. You may set
_delete_=True
to ignore the base config.