search

johnning2333 / M2Doc

28 stars 0 forks source link

Why is the MAP value very low when using the DINO-R50 weights for testing? #4

Closed sunshine-app closed 3 months ago

sunshine-app commented 3 months ago

bash mmdetection/tools/dist_test.sh mmdetection/m2doc_config/dino-4scale_w_m2doc_r50_doclaynet_12epoch.py dino_w_m2doc_r50_epoch_12.pth 1 20240614

johnning2333 commented 3 months ago

bash mmdetection/tools/dist_test.sh mmdetection/m2doc_config/dino-4scale_w_m2doc_r50_doclaynet_12epoch.py dino_w_m2doc_r50_epoch_12.pth 1 20240614

Please be sure to use ocr annotations of doclaynet. And convert dataset ocr annotations into specific format using /data/doclaynet/ocr_anno_convert.py.

JulioZhao97 commented 1 month ago

bash mmdetection/tools/dist_test.sh mmdetection/m2doc_config/dino-4scale_w_m2doc_r50_doclaynet_12epoch.py dino_w_m2doc_r50_epoch_12.pth 1 20240614

Hello!, can I ask you what version of pytorch, cuda, mmcv, mmengine, transformers did you use? I cannot run inference on my environment.

johnning2333 commented 1 month ago

bash mmdetection/tools/dist_test.sh mmdetection/m2doc_config/dino-4scale_w_m2doc_r50_doclaynet_12epoch.py dino_w_m2doc_r50_epoch_12.pth 1 20240614

Hello!, can I ask you what version of pytorch, cuda, mmcv, mmengine, transformers did you use? I cannot run inference on my environment.

you could create environment as:

conda create -n m2doc python=3.8 -y
conda activate m2doc
conda install pytorch==1.8.1 torchvision==0.9.1 torchaudio==0.8.1 cudatoolkit=10.2 -c pytorch
git clone https://github.com/johnning2333/M2Doc.git
cd M2Doc/mmdetection
pip install -v -e .
pip install transformers
pip install mmengine
mim install mmcv