Skip to content

Latest commit

 

History

History
33 lines (19 loc) · 1.32 KB

README.md

File metadata and controls

33 lines (19 loc) · 1.32 KB

Stack Overflow label predictor

A simple NLP model that predicts the language of the question user has asked, The model has been trained on stack-overflow dataset which has 16,000 questions from 4 languages and the language the question its labelled for, See the Jupyter Notebook in the notebooks directory for more information and see the implementation of the model.

Before this was written with FastAPI and Jinja templates (see in src/fastapi.py)

But for deployement I switched to Streamlit Now the Website is deployed here - https://stack-overflow-label-predictor-rishabh.streamlit.app/

Now models only works for following language:

  • Csharp
  • Javascript
  • Java
  • Python

Models analysis

Loss and Validation Loss

image

Accuracy and Validation Accuracy

image

Accuracy and Loss on the Test set

image

Preview of the website

image