Error in running train.py on GPU machine

[Hizafa-Nadeem]

I am facing issue while running the train.py on GPU linux machine. train.sh: line 1: ./darknet directory is not found.

Please, inform how can I resolve the error?

[laurelrr]

Please verify that you have followed the installation instructions in the manual. See https://github.com/ZexinChen/AlphaTracker/blob/main/Manual/Tracking.md

If you have followed these instructions, it is possible that the automatic install.sh file did not work for you. In this case, you should open install.sh and manually enter the commands listed there in your terminal window. In particular, the command to install darknet is:

Please first modify Makefile according to cuda version (this is covered in the manual).
cd ./Tracking/AlphaTracker/train_yolo/darknet/

make You should now see the darknet folder contains a darkenet executable file. If this is not present, then you probably need to correct the -gencode arch setting you used.

cd ../../../../