Skip to content

Latest commit

 

History

History
12 lines (9 loc) · 1.45 KB

TROUBLESHOOTING.md

File metadata and controls

12 lines (9 loc) · 1.45 KB

Troubleshooting and Frequently Asked Questions

Troubleshooting

Error Resolution
ImportError: libpython3.7m.so.1.0: cannot open shared object file: No such file or directory or Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; Ensure you have download cuda and set your LD_LIBRARY_PATH, e.g. export LD_LIBRARY_PATH=[ENV_PATH]:$LD_LIBRARY_PATH , where [ENV_PATH] is where your python virtual environment is located or nvidai instructions.
GPU crashes immeditely due to a lack of memory. Add this to the top of the main file, before any other code. import tensorflow as tf
import os
physical_devices = tf.config.list_physical_devices("GPU")
if physical_devices:
tf.config.experimental.set_memory_growth(physical_devices[0], True)
os.environ["TF_FORCE_GPU_ALLOW_GROWTH"] = "true"
make: 'build' is up to date. Use -B option
Docker container not using gpu. Download latest nvidia cuda drivers.

Frequently Asked Questions