https://github.com/hsm207/docker-nlp-architect
A Dockerfile to get started with Intel's NLP Architect
https://github.com/hsm207/docker-nlp-architect
deep-learning deeplearning dockerfile natural-language-processing nlp nlu
Last synced: about 1 month ago
JSON representation
A Dockerfile to get started with Intel's NLP Architect
- Host: GitHub
- URL: https://github.com/hsm207/docker-nlp-architect
- Owner: hsm207
- License: gpl-3.0
- Created: 2019-04-29T12:32:18.000Z (about 7 years ago)
- Default Branch: master
- Last Pushed: 2019-05-01T12:31:31.000Z (about 7 years ago)
- Last Synced: 2025-03-04T07:15:03.549Z (over 1 year ago)
- Topics: deep-learning, deeplearning, dockerfile, natural-language-processing, nlp, nlu
- Language: Python
- Size: 18.6 KB
- Stars: 0
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Introduction
This repository contains a script to generate a Dockerfile that contains Intel's [NLP Architect](https://github.com/NervanaSystems/nlp-architect)
with the option for CPU or GPU backend. Running the image will start a Jupyter server which
you can use to explore the library's repository.
# Requirements
This script requires Python 3.6 or above.
# Usage
## Create the Dockerfile
1. Clone this repository and navigate to the root folder.
2. Execute:
```
python create_nlp_architect_dockerfile.py --backend=gpu
```
to create an NLP Architect Dockerfile with a gpu backend (replace `gpu` with `cpu` if you
want a cpu backend). Note the location of the Dockerfile.
## Run the image
Following the example in the previous section, execute the following command after you
have built the image with `docker build` (assuming you tagged the image as `nlp_architect`):
```
docker run --runtime=nvidia -it -p 8888:8888 -p 5006:5006 nlp_architect
```
You can now open your web browser and navigate to your machine's ip address at port 8888 to
explore the NLP Architect repository. The `examples` and `tutorials` folder are a good
place to start.
Note: You will need [nvidia-docker](https://github.com/NVIDIA/nvidia-docker) installed if you
plan on running the gpu version of the image.
# Contributions
Feel free to raise a pull request if you have any questions, feedback or contributions to share.