Started with bert-base-nli-stsb-mean-tokens
Trained using sentence transformers
Continue training of quora dataset
See quora.py for script used