exampe_repo_from_scutbds

Author: Yirong Chen

README: English | 中文

Introduction

This is an open source project specification for Research Center of Body Data Science from South China University of Technology.

Tools

• Git: Git is an open source distributed version control system for agile and efficient processing of any small or large project.
• Github: GitHub is a code hosting cloud service website that helps developers store and manage their project source code, and can track, record, and control users' modifications to their code.
• shields: Shields is an icon generator, often used to add some icons to the project to improve readability.
• Markdown: Markdown is a text-to-HTML conversion tool used on the Web. It can generate structured HTML documents in a simple, easy-to-read and easy-to-write text format. Currently github, Stackoverflow and other websites support this format.

Structure of the Whole Repository

• ./config: Store the model's hyperparameter configuration, vocabulary, etc.
• ./data: Store the data set or data sample of the training model in .txt, .json or .csv format
• ./eval: Store the code used to verify the model or evaluate the results of the model output
• ./figure: Storage related pictures of the project
• ./model: Store model code, the folder name can also be baselines or model name, etc.
• ./utils: Store the code to load the data set
• LICENSE: The copyright permission selected when github created the repository, the automatically generated file
• README.md: English version description of the project
• README-zh.md: Chinese version description of the project
• requirements.txt: the requirement file for running the model
• run_train_model.sh： the bash command file for running the model

Note: The above is not the only structure. You can adjust the name of directory/file according to the actual situation, and you can also add or delete the unneeded directory/file.

Structure of the README.md file

• Introduction: One sentence is usually used to indicate the purpose of the project. If there is a corresponding published paper or data set, the corresponding link needs to be given. For example, The text below is an introduction from the project BERT.

BERT, or Bidirectional Encoder Representations from Transformers, is a new method of pre-training language representations which obtains state-of-the-art results on a wide array of Natural Language Processing (NLP) tasks.

Our academic paper which describes BERT in detail and provides full results on a number of tasks can be found here: https://arxiv.org/abs/1810.04805.