Closed ElifErarslan1 closed 2 years ago
Hi, I have rearranged the code according to the custom dataset and when I ran the training code floating point error occurs. Although I changed the base learning rate, the issue still occurs.
what is your detectron2 version?
hi, I directly get it from instructions and its version is 0.4
hi, I directly get it from instructions and its version is 0.4
downgrade to 0.3 or 0.2.1
Thanks @AISoltani. At the time of release of this repository, we were using detectron2 v0.2.1. I will update the README to make it more clear, as it is quite outdated.
Hi, I downgrade it to v0.3 and it worked. Thank you for your support.
Hi, I have rearranged the code according to the custom dataset and when I ran the training code floating point error occurs. Although I changed the base learning rate, the issue still occurs.