Running the Code for Object Detection
sazed49 opened this issue · comments
I am completely new in this field. I have assigned to reproduce this code for my class project but I am facing several errors for the past two months. Sometime it's with the version of pytorch, cuda and all. Sometimes it's with the code. At this point I am totally stressed up. i don't know how many environments I created just to make it run ! Thus can anyone help me out how to accomplish a simple task of "training a dataset(COCO format available) on this method?
Hi, sorry for the inconvenience. I assume you are facing issues of creating mmdetection environments?
A kind reminder is that you probably need to exactly to follow the packages version I mentioned in the README file (mmdet changes a lot in different versions). If that is not clear, I would recommand read instructions in PVT and PoolFormer, which share some environments.