UWU_Upgrad

• 

  Model in action

Inspiration

Our sheer shock after lookin at the dataset for problem statement 4.

What it does

It takes simple news paragraph as input on the web UI. Detects whether the news is fake or not. Then depending on the feedback of the user it improves itself.

How we built it

Frontend --> VannilaJS and CSS Backend --> Django ML Module --> Tensorflow Algorithm --> Bi-LSTM Deployment --> Railway

Challenges we ran into

• Basic RNN not enough
• Deployment was hard with large modules and high usage

Accomplishments that we're proud of

• The self trained model and custom implemented Retrain Pipeline
• Simplistic UI

What we learned

• In practice computation limitation while actually deploying.
• How to resolve Git merge conflicts while collaborating with multiple people.

What's next for UWU_Upgrad

• scraping newer data from web and then to actually make it practical for current usage.
• some method of detecting whether the user is spamming We will continue it later

Built With

• css
• django
• javascript
• railway
• tensorflow