This is a proejct for L101 Machine Learning for Language Processing module at the University of Cambridge, MPhil ACS 2022/23 academic year.
It is requied to have SNLI, MNLI, and ChaosNLI data, stored at./data/snli
, ./data/mnli
and./data/chaosNLI
respectively.
dataset.py
-- Custom dataset object to handle soft labelsfine_tune.py
-- Pipeline for fine-tuning sentence transformersmodel.py
-- Model file for classifier moduleMy....py
-- Classes inherenting default sentence transformer classes to support soft labelspreprocessing.py
-- Preprocessingtrain.py
Training pipeline for the classifier moduleutils.py
-- Utilities