Skip to content
/ QAML Public
forked from saptab/QAML

Web Interface for Adversarial Question Answering

Notifications You must be signed in to change notification settings

Pinafore/QAML

 
 

Repository files navigation

QAML

GitHub repo size Lines of code GitHub language count GitHub contributors

QAML (/ˈkæməl/, Question-writing Aided by Machine Learning) is an interface for adversarial question-writing enhanced by the QANTA QA dataset.

Front end: Vue.js

Back end: Flask

Database: MySQL

Preparations

  1. Clone the Repo
  2. cd into the repo in your terminal

Model Files

Please try our new model download script first! If there are any problems, you can revert to installing them manually.

With Script:

cd Flask
pip install -r requirements.txt
python3 download.py

Without Script:

Sample Questions

Please consider using sample questions in this Google Doc to test the interface https://docs.google.com/document/d/1hJZDDQGI3GuTFqO1_Ogya_OK7yRx-mSiB0Cpp7dt3yg/edit?usp=sharing

Vue.js ( 2.6.14 )

npm install 
npm install -g @vue/cli
npm install -g [email protected]

If this does not work check out the guide below to try and install another way.

  • Follow this link to install vue.js we recommend the production version as it has proper warnings for development.

Python Packages

We recomend that use a seperate Conda envorinment or pipenv environment for installing packages, however this is only a recommendation.

(2.5 Gb needed)

If Conda:

  • To install conda visit this link.
  • conda create --name QAML python=3.9
  • conda activate QAML
  • cd Flask
  • python3 -m spacy download en_core_web_sm
  • pip3 install -r requirements.txt

If no Conda:

  • cd Flask
  • python3 -m spacy download en_core_web_sm
  • pip3 install -r requirements.txt

Running the Code

Run the Flask App (Back End)

  1. Navigate to the root directory
  2. cd Flask
  3. python3 run.py

Run the Vue Framework (Front End)

  1. (In a seperate terminal) Navigate to the root directory
  2. cd Vue
  3. npm install
  4. npm run dev

Setup Error

About

Web Interface for Adversarial Question Answering

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 35.1%
  • Vue 34.3%
  • JavaScript 17.7%
  • Jupyter Notebook 9.7%
  • SCSS 2.0%
  • CSS 0.7%
  • Other 0.5%