cpt-chewieeee / tensorflow2-cuda

Tensorflow 2 with Cuda (Docker Build)

tensorflow2-cuda

Tensorflow 2 with Cuda (Docker Build)

Workaround to handle CUDA 10.1 incompatibility

Requirements (Arch):

• Install latest nvidia graphics drivers on host: pacman -S nvidia
• Install nvidia-docker from aur: yay -S nvidia-docker

Usage:

docker run -d -P -p 8022:22 --runtime=nvidia twodarkmessiah/tensorflow2-cuda

default ssh password: 'changeme'

ssh [email protected] -p 8022

python3

python~ import tensorflow

python~ print("hello tensorflow")

Output should be:

Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 9359 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1080 Ti, pci bus id: 0000:01:00.0, compute capability: 6.1)

For Development

Configure Remote Python Interpreter to 127.0.0.1:8022.

This Docker Container is made to offer a workaround, to nvidia cuda problems with tensorflow2.

Failed to load the native TensorFlow runtime
ImportError: /usr/lib/libcusolver.so.10.0: version `libcusolver.so.10.0' not found