gnperdue/singularity_imgs:py3_tf2gnt
$ singularity pull shub://gnperdue/singularity_imgs:py3_tf2gnt
Singularity Recipe
Bootstrap: docker
From: nvidia/cuda:10.0-devel-ubuntu16.04
%help
To start your container simply try
singularity exec THIS_CONTAINER.simg bash
To use GPUs, try
singularity exec --nv THIS_CONTAINER.simg bash
Container based on recipe by drinkingkazu
%labels
Maintainer gnperdue
Version ubuntu16.04-py3-tf2-alpha-gpu
#------------
# Global installation
#------------
%environment
export XDG_RUNTIME_DIR=/tmp/$USER
export LD_LIBRARY_PATH=/usr/local/cuda/lib64:/usr/local/cuda/lib64/stubs:${LD_LIBRARY_PATH}
export CUDA_DEVICE_ORDER=PCI_BUS_ID
%post
# add wilson cluster mount points
mkdir /scratch /data /lfstev
# apt-get
apt-get -y update
apt-get -y install dpkg-dev g++ gcc binutils libqt4-dev python3-dev python3-tk python3-pip
# pip
python3 -m pip install --upgrade setuptools pip
python3 -m pip install --no-cache-dir --upgrade tf-nightly-gpu-2.0-preview tf-agents-nightly tfp-nightly tensorflow-datasets
python3 -m pip install --no-cache-dir matplotlib seaborn scikit-image
python3 -m pip install --no-cache-dir 'gym[atari,box2d,classic_control]'
Collection
- Name: gnperdue/singularity_imgs
- License: None
View on Datalad
Metrics
key | value |
---|---|
id | /containers/gnperdue-singularity_imgs-py3_tf2gnt |
collection name | gnperdue/singularity_imgs |
branch | master |
tag | py3_tf2gnt |
commit | 7af2da88c575ea35408c3bb04b585c7b3937e583 |
version (container hash) | 3058b97fa940a932db3a66c24bf0e4dd |
build date | 2019-05-23T15:31:33.040Z |
size (MB) | 4240 |
size (bytes) | 1945292831 |
SIF | Download URL (please use pull with shub://) |
Datalad URL | View on Datalad |
Singularity Recipe | Singularity Recipe on Datalad |
Feedback
Was this page helpful?
Glad to hear it! Please tell us how we can improve.
Sorry to hear that. Please tell us how we can improve.