Issue with Environment (mmdet)

[FrancoisPorcher]

Dear Co-DETR team,

I have the following issue with "mmdet":

I ran this command:

bash tools/dist_test.sh projects/configs/co_deformable_detr/co_dino_5scale_vit_large_lvis.py /private/home/francoisporcher/WeakSupSeg/francois-semi-supervised/Co-DETR/checkpoints/co_dino_5scale_vit_large_lvis.pth 2 --eval bbox

And get the following error:

Could you please explain what environment you use and how you set it up?

The commands I used are:

conda create --name codetr python=3.10 -y; conda activate codetr; conda install -y pytorch=1.11.0 torchvision torchaudio cudatoolkit=11.5 -c pytorch -c nvidia pip install mmcv-full==1.5.0 -f https://download.openmmlab.com/mmcv/dist/cu115/torch1.11.0/index.html

I did not install mmdet myself, as you advised, but it seems that it does not work for me.

Thanks!!

[FrancoisPorcher]

Also pip install terminal tables does not solve the issue because then there are other libraries (pycocotools, six, etc..)

[TempleX98]

Please try pip install -r requirements.txt

[FrancoisPorcher]

It does not work..

I still get the error:

Traceback (most recent call last): Traceback (most recent call last): File "/private/home/francoisporcher/WeakSupSeg/francois-semi-supervised/Co-DETR/tools/test.py", line 15, in File "/private/home/francoisporcher/WeakSupSeg/francois-semi-supervised/Co-DETR/tools/test.py", line 15, in from mmdet.apis import multi_gpu_test, single_gpu_test ModuleNotFoundError : from mmdet.apis import multi_gpu_test, single_gpu_testNo module named 'mmdet'

The sequence of commands I ran is:

conda create --name codetr python=3.10 -y; conda activate codetr; conda install -y pytorch=1.11.0 torchvision torchaudio cudatoolkit=11.5 -c pytorch -c nvidia

Then install the requirements pip install -r requirements.txt

pip install mmcv-full==1.5.0 -f https://download.openmmlab.com/mmcv/dist/cu115/torch1.11.0/index.html

Then:

bash tools/dist_test.sh projects/configs/co_deformable_detr/co_dino_5scale_vit_large_lvis.py /private/home/francoisporcher/WeakSupSeg/francois-semi-supervised/Co-DETR/checkpoints/co_dino_5scale_vit_large_lvis.pth 2 --eval bbox

[TempleX98]

I have never met this issue. Maybe you can install mmdet by pip install -v -e . to see if it works.

[FrancoisPorcher]

So I tried these commands (all ran from the Co-DETR/ folder, right?):

conda create --name codetr python=3.10 -y; conda activate codetr; conda install -y pytorch=1.11.0 torchvision torchaudio cudatoolkit=11.5 -c pytorch -c nvidia

pip install -r requirements.txt

pip install mmcv-full==1.5.0 -f https://download.openmmlab.com/mmcv/dist/cu115/torch1.11.0/index.html

pip install -v -e .

And I get this error:

I tried pip install fair scale, but then its the library fvcore (also called by mmdet), which leads me to think that its mmdet the problem but I am not sure. What do you think?

If it's not too much to ask, could you please try to create the same virtual environment on conda to check? Thanks!

[Code-of-Liujie]

The same question, speechless

[Code-of-Liujie]

The bug cannot be solved at all. What is the reason that delays my day?

[MihaiDavid05]

Hello, @Code-of-Liujie, @FrancoisPorcher .

I was able to make it work.

1. git clone https://github.com/Sense-X/Co-DETR.git
2. conda create --name codetr python=3.8 -y
3. conda activate codetr
4. pip install torch==1.11.0+cu113 torchvision==0.12.0+cu113 torchaudio==0.11.0 --extra-index-url https://download.pytorch.org/whl/cu113 (mandatory to install it with pip)
5. pip install mmcv-full==1.5.0 -f https://download.openmmlab.com/mmcv/dist/cu113/torch1.11.0/index.html
6. pip install -r requirements.txt
7. pip install yapf==0.40.1 (this is beacuse of this error)