anaconda-tensorflow-gpu
2.4K
You have to write cudnn-7.5-linux-x64-v5.0-ga path into cuDNN_path.txt.
During Docker build, it will be downloaded and extracted.
default endpoint is jupyter notebook
DOCKER_NVIDIA_DEVICES="--device /dev/nvidia0:/dev/nvidia0 --device /dev/nvidiactl:/dev/nvidiactl --device /dev/nvidia-uvm:/dev/nvidia-uvm"
sudo docker run --name tensorflow -v $PWD/notebook:/notebook -p 8888:8888 -p 6006:6006 -d $DOCKER_NVIDIA_DEVICES drunkar/anaconda-tensorflow-gpu /bin/bash
If you want to run bash on this container, docker exec command is a good way.
After docker run,
sudo docker exec -it tensorflow bash
You can also use TensorBoard in exec process
tensorboard --logdir=/data/dir
Content type
Image
Digest
sha256:ed7e0751f…
Size
1.6 GB
Last updated
over 10 years ago
docker pull drunkar/anaconda-tensorflow-gpu