WongKinYiu / yolov7

Implementation of paper - YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors
GNU General Public License v3.0
13.22k stars 4.17k forks source link

yolov7-mask cannot run #587

Open futureflsl opened 2 years ago

futureflsl commented 2 years ago
from detectron2.modeling.poolers import ROIPooler

File "/home/fut/Downloads/detectron2/detectron2/modeling/init.py", line 20, in from .meta_arch import ( File "/home/fut/Downloads/detectron2/detectron2/modeling/meta_arch/init.py", line 6, in from .panoptic_fpn import PanopticFPN File "/home/fut/Downloads/detectron2/detectron2/modeling/meta_arch/panoptic_fpn.py", line 14, in from .rcnn import GeneralizedRCNN File "/home/fut/Downloads/detectron2/detectron2/modeling/meta_arch/rcnn.py", line 18, in from ..roi_heads import build_roi_heads File "/home/fut/Downloads/detectron2/detectron2/modeling/roi_heads/init.py", line 15, in from .roi_heads import ( File "/home/fut/Downloads/detectron2/detectron2/modeling/roi_heads/roi_heads.py", line 17, in from ..poolers import ROIPooler File "/home/fut/Downloads/detectron2/detectron2/modeling/poolers.py", line 10, in from detectron2.utils.tracing import assert_fx_safe File "/home/fut/Downloads/detectron2/detectron2/utils/tracing.py", line 4, in from torch.fx._symbolic_trace import _orig_module_call ModuleNotFoundError: No module named 'torch.fx._symbolic_trace'

I use envronment:

torch==1.8.0 torchvsion==0.9.0 detectron2==2.0.6

I changed torch==1.7.0 and torch==1.9.0 error is the same

WongKinYiu commented 2 years ago

Our environment: torch==1.11.0+cu113 torchvsion==0.12.0+cu113 detectron2==0.6