run nasfpn_1280 timeout

tusen-ai / simpledet

A Simple and Versatile Framework for Object Detection and Instance Recognition

Apache License 2.0

3.08k stars 488 forks source link

run nasfpn_1280 timeout #179

Closed Zenobia7 closed 5 years ago

Zenobia7 commented 5 years ago

hello , it always stay "Running performance tests to find the best convolution algorithm, this can take a while...". when i run retina_r50v1b_nasfpn_1280_7@384_25epoch.py with 2X 2080Ti .

RogerChern commented 5 years ago

This is generally because you are using a wrong CUDA/cuDNN lib. Please try to pull the CUDA10 docker to see if this problem persists.

Zenobia7 commented 5 years ago

@RogerChern thanks, I will try to pull the CUDA10 docker.