harshavii/token_summarization

FLASK APPLICATION: PROMPT

Used flask framework along with flask-bootstrap. To run the flask application, all required packages are included in requirements.txt file.

For this project, I employed a summarization algorithm to enhance token count efficiency. I utilized the t5 model from Hugging Face's transformer library, which employs an abstractive approach to summarization. Additionally, I experimented with the bert model, which adopts an extractive approach, but the outcomes were not as optimal as those achieved using the abstractive approach. Basically, in this task, the utilization of a machine learning-based approach gave more optimal results when compared to nltk, spacy and rule-based approach.