sentiment-api

A lightweight Flask API for sentiment analysis using Hugging Face's transformers library and the DistilBERT model. Packaged as docker container. Optional Nginx Reverse Proxy to redirect port 8000 to port 80

Using HTTP protocol you send a string to the service and it gives output on like/dislike example: run the program - asking about ""I love this movie!"

bash curl-clientusentiment_api.sh

output: [{"label":"POSITIVE","score":0.9998775720596313}]

Features

• RESTful API with a /predict endpoint
• Uses distilbert-base-uncased-finetuned-sst-2-english for binary sentiment classification
• Packaged as a Python module with CLI entry point
• Easy to install and extend

Server and client

Function	Name
'Docker image'	sentiment-api
'Docker container server'	sentiment-container
'Sample curl client'	curl-client-sentiment_api.sh
'Sample python client'	python-client-sentiment_package.py

Installation

Tested on windows and Linux

Make sure you have git installed.

git clone https://github.com/steve100/sentiment-api.git

• cd sentiment-api

File Name	Description
app.py	The Flask/gunicorn web application
build.sh	Builds the Docker image
curl-port80.sh	Tests the web service on port 80 (reverse proxy must be installed)
Dockerfile	Instructions on how to make the Docker image
requirements.txt	Python modules used
run.sh	Starts and runs the Docker container that runs the web service

• Container Save docker save > sentiment-api.tar

• Container Load docker load -i sentiment-api.tar

tests-tools

• cd tests-tools

File Name	Description
python-client-sentiment_package.py	Tests the web service on port 8000 with Python (run this in a venv)
curl-client-sentiment_api.sh	Tests the web service on port 8000 with Curl
curl-port80.sh	Tests the web service on port 80 (reverse proxy must be installed)
onequery-sentiment.py	Runs the model without a web service
requirements.txt	Python modules used

Run

Start the server

cd sentiment-api
bash run.sh

Test the client

cd tests-tools
bash curl-client-sentiment_api.sh

Test the Python client and write your application

Chose a location for the Python Virtual Environment HOW To Link: https://www.geeksforgeeks.org/python-virtual-environment/

python3 -m venv vpython
cd vpython
source ./bin/activate

#copy the tests and tools to your virtual environment.
#in this case vpython

cp  tests-tools/* . 
pip install --no-cache-dir --upgrade pip
pip install --no-cache-dir -r requirements.txt

python3 onequery-sentiment.py
python3 python-client-sentiment_package.py


#when done
deactivate 

How to Install GIT

General Git

https://git-scm.com/
https://git-scm.com/doc

Linux Ubuntu Git

• Information

Use your Distro
sudo apt upgrade -y  ; sudo   apt install git -y

or you can get the latest
https://git-scm.com/downloads/linux

Windows Git

• Information

https://git-scm.com/downloads/win

or same thing, more information
https://gitforwindows.org/

Git Usage

• Set the identity in the repo for push

  git config --global user.email youremail@yourdomain
  git config --global user.name  yourname

• Cheat sheet

  git status
  git add . 
  git commit -m "your commit"
  #you will now need a token because  userid/password depreciated
  git push

• tokens

https://git-scm.com/downloads/win

• Markdown

https://docs.github.com/en/get-started/writing-on-github/getting-started-with-writing-and-formatting-on-github/quickstart-for-writing-on-github

• Authentication

vim install-gh.sh ; bash install-gh.sh
gh auth login

Docker

Docker for Linux

Used the convience script
 https://docs.docker.com/engine/install/ubuntu/

It makes things easier if you add your user to the docker group

Docker for Windows

 https://docs.docker.com/desktop/setup/install/windows-install/

 Video Installing  Docker Desktop and optonal  with WSL2
 https://www.youtube.com/watch?v=ZyBBv1JmnWQ