We have used a pretrained model, and trained it on our custom dataset created using raw webcam pictures and annotated using Albumentation library.

Created using python and dash library, this is a dynamic Avocado Sales visualization dashboard, which uses python decorators for dyanmic behavior.

Performed Sentiment Analysis with BERT Transformer

CSV messy data extracted transformed and loaded in to SQLite db, mini ETL pipeline

This is a handwritten-equations to latex-equations converter, built using trocrbasestage1 model + CROHME dataset + custom-built SentencePiece Tokenizer (latex)