search

jshilong / GroupRCNN

Group R-CNN for Point-based Weakly Semi-supervised Object Detection (CVPR2022)
Apache License 2.0
138 stars 10 forks source link

Use 2 GPUs #9

Open suekarry opened 1 year ago

suekarry commented 1 year ago

Sorry to bother you, but how to set the file parameters when using two Gpus or one GPU? Is it set directly through the command? When I tried this method, I got an error as shown in the figure below: image Hope to receive your reply, thank you!

jshilong commented 1 year ago

Please check your environment first and confirm your CUDA Pytorch and MMCV installed correctly Then run

sh tools/dist_train.sh projects/configs/10_coco/group_rcnn_24e_10_percent_coco_detr_augmentation.py 8 --work-dir ./exp/group_rcnn

suekarry commented 1 year ago

Please check your environment first and confirm your CUDA Pytorch and MMCV installed correctly Then run

sh tools/dist_train.sh projects/configs/10_coco/group_rcnn_24e_10_percent_coco_detr_augmentation.py 8 --work-dir ./exp/group_rcnn

Sorry to bother you. How is part E (VOC2007, and VOC2012) in the appendix realized? At present, I directly converted VOC into COCO format and ran it, but the evaluation index was much different from AP 72.6 mentioned in the appendix. May I ask how you handled the data set?