Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/uetchy/dockerfile-machinelearning
🐳🤖 Dockerfile for ML researchers.
https://github.com/uetchy/dockerfile-machinelearning
docker-image dockerfile machine-learning
Last synced: 19 days ago
JSON representation
🐳🤖 Dockerfile for ML researchers.
- Host: GitHub
- URL: https://github.com/uetchy/dockerfile-machinelearning
- Owner: uetchy
- License: mit
- Created: 2015-07-27T07:16:59.000Z (over 9 years ago)
- Default Branch: master
- Last Pushed: 2019-03-20T16:29:18.000Z (over 5 years ago)
- Last Synced: 2024-10-07T16:44:22.859Z (29 days ago)
- Topics: docker-image, dockerfile, machine-learning
- Language: Dockerfile
- Homepage:
- Size: 32.2 KB
- Stars: 45
- Watchers: 6
- Forks: 15
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Dockerfile for Machine Learning
A Dockerfile for accelerated research process, consists of major machine learning libraries.
- [DockerHub](https://registry.hub.docker.com/u/uetchy/ml/)
## Features
- Ubuntu 18.04
- Python 3.6 (Miniconda 3)
- GPU accelerated (CUDA 10.0, cuDNN 7)
- NCCL, CNMeM, Apex (PyTorch only) activated
- Jupyter and OpenCV 3.0 included
- Additional packages (Tensorboard, Hyperdash, etc)
## Available Deep Learning Frameworks
- TensorFlow `uetchy/ml:tensorflow`
- PyTorch `uetchy/ml:pytorch`
- Chainer `uetchy/ml:chainer`
- MXnet `uetchy/ml:mxnet`
- XGBoost `uetchy/ml:xgboost`
## Install
### System Requirements
- Docker
- CUDA-enabled GPUs
- CUDA Driver
- CUDA Toolkit
- nvidia-docker2 (https://github.com/NVIDIA/nvidia-docker)
### Pull the docker image from [DockerHub](https://registry.hub.docker.com/u/uetchy/ml/)
```bash
docker pull uetchy/ml:tensorflow
docker pull uetchy/ml:pytorch
docker pull uetchy/ml:chainer
...
```
### Launch Jupyter Lab on current directory
```bash
docker run --runtime=nvidia -v $PWD:/workspace -p 8888:8888 -it uetchy/ml:pytorch jupyter
open http://localhost:8888
```
### Open Python REPL
```bash
docker run --runtime=nvidia --rm -it uetchy/ml:base python
```
### Run Bash Shell
```bash
docker run --runtime=nvidia --rm -it uetchy/ml:tensorflow
```
## List of Docker images for Data Science
- [TensorFlow](https://hub.docker.com/r/tensorflow/tensorflow)
- [Caffe2](https://hub.docker.com/r/caffe2ai/caffe2)
- [MXNet](https://hub.docker.com/u/mxnet)
- [Caffe](https://github.com/BVLC/caffe/tree/master/docker)
# Contribution
PRs are accepted.
## Contributors
- Yasuaki Uechi
- UpmostScarab
- cyrusmvahid