mrsayan / docwiz

DocWiz

An automatic document generator for generating customizable Machine Learning risk management documents by just

• uploading your models and validation dataset
• writing very bare minimum about the model performance and general information
• Watch different charts and choose which are required to export
• Using chatGPT API DocWiz can generate professional reports from just bare minimum information
• Get your Documentation with customizable templates within seconds

---

How to run the project

First clone the project by the following command:

git clone https://github.com/mrsayan/DocWiz.git

Linux

You can either locally install everything by create your virtual environment first by:

make virtualenv

and then just install it by:

make install 

Additionally you can also run other make commands get started with linting and formatting.

Then just run the following command for running the project

streamlit run main.py 

Windows

In windows if you are in WSL then doing the same as above will be just fine. Else you can install cmake using choco with this command:

choco install cmake

And then do the samee thing as followd.

Docker

If you have docker installed then you can build the image by the following:

docker build -t doc-wiz:v1 . 

And then run the following command to start the container

docker run doc-wiz:v1

---

Future Scopes and Roadmap

Web

• Building a dedicated backend using FastAPI / Flask
• Better UI with React/Angular

ML/MLOps

• Tracking data and model drift and visualization
• Using XAI (Explainable AI) for providing better root cause analysis
• Using automatics CI/CD for generating major data/model version control information on the doc.
• Also showing other information metric required for deployement including CPU Usages, disk usage and latency