Skip to content

Latest commit

 

History

History
8 lines (7 loc) · 808 Bytes

README.md

File metadata and controls

8 lines (7 loc) · 808 Bytes

Kaggle-TSE

About

  1. The project was worked in a team of two while participating in Kaggle's Summer 2020 challenge named Tweet Sentiment Extraction.
  2. A model was required to be made which would extract phrases from a given tweet which reflects a given sentiment.
  3. Hugging Face's DistilBERT was used to train the model. The data for which was downloaded from Kaggle itself.
  4. A question answering model was made using the above-mentioned model while applying some pre-processing and post-processing methods inferred from the results and the given dataset.
  5. An accuracy of 70.636 was achieved in the final private leaderboard.