RuntimeError: GET was unable to find an engine to execute this computation

HarborYuan / ovsam

[ECCV 2024] The official code of paper "Open-Vocabulary SAM".

https://www.mmlab-ntu.com/project/ovsam

Other

914 stars 27 forks source link

RuntimeError: GET was unable to find an engine to execute this computation #27

Closed Danbinabo closed 3 months ago

Danbinabo commented 3 months ago

I prepare envirement and want to Inference,but run bash tools/dist.sh test seg/configs/ovsam/ovsam_coco_rn50x16_point.py 8 happend GRT false, its: File "/mnt/NewDataShare/D4/common/wbzhou/MLLM/ovsam/seg/models/data_preprocessor/ovsam_preprocessor.py", line 193, in forward gt_instances.point_coords = get_center_coords( File "/mnt/NewDataShare/D4/common/wbzhou/MLLM/ovsam/seg/models/data_preprocessor/ovsam_preprocessor.py", line 24, in ovsam_bug

Is this issue due to incorrect environment configuration? torch==2.1.2+cu121 torchvison==0.16.2+cu121 mmcv==2.1.0 mmdet==3.3.0

HarborYuan commented 3 months ago

Hi @Danbinabo ,

I have a brief check of your screenshot. It is not expected to have the shown error with your current versions. The error seems like there is no suitable compiled CUDA op. I suspect that it is because of the way you install your env.

Please give more descriptions of how you install the environment if possible.

HarborYuan commented 3 months ago

You need to modify this (file) config file to support more dataset.

To write such a config, you may need to write a new dataset class starting from COCO and import it to your config.

HarborYuan commented 3 months ago

BTW, could you please provide some details about how you fix the env bug? It would be great for other guys to run the code.