vancouver-datajam / news_processor Goto Github PK

Run any transformations necessary to our supervised Learning models

Optionals:

1. Label the dataset if necessary
2. Tag metadata

• Create Presentation
• Deliver Presentation

Explore Datasets that we can use. (Kaggle, Github, Open Source)
If not available or does not fulfill our requirements scrape data.

Put Models on Docker Files
Flaks WebServer

Jupyter Notebook with all the bells and whistles.

Should be Clean, Engaging and show meaningful insights

• Decision on how to show our models/system
• Front-end Development ( Js,Gradio, Powerpoint, API)

Try out MVP sentiment analysis using out-of-the-box models a snippet of text (no dependency on getting data ready yet)

Train Models or fine-tune them

Do the pre-training required transformations.
Include Embedding, Tokenizations, Lemmatization, ETC.

Label the data