[NAACL 2021] This is the code for our paper `Fine-Tuning Pre-trained Language Model with Weak Supervision: A Contrastive-Regularized Self-Training Approach'.

transform multi-label classification as sentence pair task, with more training data and information

BERT: Sentence-Pair Classification for Natural Language Understanding

Gathers machine learning and Tensorflow deep learning models for NLP problems

A binary classification task for identifying speakers in a dialogue, training using a RNN with attention and BERT on data from the British parliment.

Example application for applying QLoRA-based Parameter-Efficient Fine-Tuning (PEFT) to a sentence pair classification task using Mistral-7B and Llama3-8B