I have successfully trained a CNN using my own dataset. However, yesterday, after updating the caffe to its latest version, without altering the architecture of network, the training data is the same as previously, it reports segmentation fault when running the train.sh. Why it report segmentation fault on the latest caffe version? Thank you!
I have successfully trained a CNN using my own dataset. However, yesterday, after updating the caffe to its latest version, without altering the architecture of network, the training data is the same as previously, it reports segmentation fault when running the train.sh. Why it report segmentation fault on the latest caffe version? Thank you!