Closed dragon515 closed 4 years ago
@huangzehao #224 caused this.
@dragon515
Hotfix: add the following codes in class RpnParam
will solve this issue
class anchor_assign:
image_anchor = 256
A PR will be created to solve this issue. The config files of fpn based cascade rcnn will be updated. Feel free to reopen the issue if it persists. @dragon515
I use costom dataset, and change the class number, the gpu index parameters.
python detection_train.py --config config_cancer/cascade_r101v1_fpn_1x.py Traceback (most recent call last): File "detection_train.py", line 307, in
train_net(parse_args())
File "detection_train.py", line 21, in train_net
transform, data_name, label_name, metric_list = config.get_config(is_train=True)
File "/home/www/simpledet/config_cancer/cascade_r101v1_fpn_1x.py", line 198, in get_config
bbox_head_3rd
File "/home/www/simpledet/models/cascade_rcnn/builder.py", line 25, in get_train_symbol
rpn_loss = rpn_head.get_loss(rpn_feat, gt_bbox, im_info)
File "/home/www/simpledet/models/FPN/builder.py", line 159, in get_loss
image_anchor = p.anchor_assign.image_anchor
AttributeError: 'NoneType' object has no attribute 'image_anchor