Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/bobbui/tensorflow-serving-cuda-docker

Docker image for tensorflow serving with Nvidia CUDA, CuDNN
https://github.com/bobbui/tensorflow-serving-cuda-docker

cuda cudnn docker docker-image tensorflow tensorflow-serving ubuntu1604

Last synced: 6 days ago
JSON representation

Docker image for tensorflow serving with Nvidia CUDA, CuDNN

Awesome Lists containing this project

README

        

OBSOLETE: now either TFS install via package manager like apt-get or using official dockerfile from here https://github.com/tensorflow/serving/tree/master/tensorflow_serving/tools/docker

# tensorflow-serving-cuda-docker
Docker image for tensorflow serving with CUDA which suitable for development purpose

This image is based from nvidia/cuda:8.0-cudnn5-devel-ubuntu14.04 include:
- ubuntu 14.04
- CUDA 8 with development packages
- cuDNN 5 with development packages
- bazel 0.4.4
- tensorflow 1.0.0
- tensorflow serving source code that is ready to build
- other setting: proxy,enviroment variable, etc .

You can using either following dockerfile:
- Dockerfile: no tensorflow serving build happen:
- Dockerfile.GUI extended from above Dockerfile with GUI and VNC server:
connect to: vnc://:5901 via VNC client. The VNC password is **password**.
After build the docker image, run tensorflow serving compliation with the desired setting.

```
cd /serving/tensorflow
./configure
cd /serving
bazel build -c opt --config=cuda tensorflow_serving/...
bazel test tensorflow_serving/...
```