google-research/albert

fine_tune albert with run_classifier.py on GPU but the GPU-util is 0,why and how?

yuanjun1991 opened this issue · 2 comments

fine_tune albert with run_classifier.py on GPU but the GPU-util is 0,why and how?

your tensorflow is upgrade tf2.x?

Are you installing a different tensorflow version with pip in google colab? Then the GPU might not be set up properly. Instead use the %tensorflow_version magic as described here.