tensorflow ml nlp tf2

Please refer to the link below for the first book on natural language processing starting with TensorFlow and machine learning.

First book link: https://github.com/NLP-kr/tensorflow-ml-nlp

* Notice: GPT2 model is no longer supported by Dropbox. To download the model, please follow the steps below.

 wget https://github.com/NLP-kr/tensorflow-ml-nlp-tf2/releases/download/v1.0/gpt_ckpt.zip -O gpt_ckpt.zip

NLPBOOK

Natural language processing starting with TensorFlow 2 and machine learning (from logistic regression to BERT and GPT3)

Introduction

This is a repository of natural language processing examples included in the book.

This repository will be more helpful if you use it in conjunction with a natural language processing book that starts with TensorFlow and machine learning.

Installation method (Environments)

A Docker guide has been added to support various development environments for the natural language processing example.

We recommend using Docker to respond to changes in lab package updates for future examples.

Run Jupyter Docker (recommended)

When using a Docker environment, we recommend using version 19.03 or later.

• Running bash build_jupyter_for_cpu.sh or bash build_jupyter_for_gpu.sh creates a docker image.
• When you run bash exec_jupyter_for_cpu.sh or bash exec_jupyter_for_gpu.sh jupyter runs in the Docker environment.
• The jupyter execution port number is 8889, so that port number must be available.

Anaconda Installation

If you cannot find the Anaconda installation file, find and download the next version from the following link.

 https://repo.anaconda.com/archive
# windows 권장 버전: Anaconda3-5.3.0-Windows-x86_64.exe
# linux 권장 버전: Anaconda3-5.3.0-Linux-x86_64.sh
# MacOS 권장 버전: Anaconda3-5.3.0-MacOSX-x86_64.pkg
# 해당 버전을 받지 않아도 환경 구성을 하는데 큰 문제는 없습니다.

• Anaconda Python virtual environment configuration

 conda create -n {사용할 환경 이름} python=3.6
conda activate {사용할 환경 이름} # 경우에 따라 동작이 안될 수 있습니다. 아래의 명령어를 입력해보세요.
# source activate {사용할 환경 이름} 
pip install -r requirements.txt

If the installation does not proceed properly, please install python 3.6 and proceed.

 conda install python=3.6

Table of Contents

Preparatory stage - This chapter provides background on natural language processing and preparation for development.

1. Entering
2. Preparing for natural language processing development
3. Natural Language Processing Overview

Natural Language Processing Basics - This is a practice chapter on basic models in natural language processing.

4. Text classification
5. Text Similarity

Advanced Natural Language Processing - This is a practice chapter on more advanced natural language processing through chatbot models.

6. Create a chatbot
7. Fine-tune learning
8. GPT3

Colab practice

A separate repository has been released for Colab practice only for Chapters 7 and 8.

• repository link

Inquiries

Pull requests are always welcome. If you have any problems, bugs, or questions, please leave a message in the issue.

**Before checking the issue, please look at the document on Wiki first and leave a comment on the issue!

Authors

ChangWookJun / @changwookjun ([email protected])
Taekyoon / @taekyoon ([email protected])
JungHyun Cho / @JungHyunCho ([email protected])
Ryan S. Shin / @aiscientist ([email protected])