Historical research checkpoint — SBERT-WK (2020)
No longer actively maintained. Retained for reproducibility of the original work.
This repository hosts the original 12-layer XLNet checkpoint used in the SBERT-WK experiments. It is a supporting checkpoint for the sentence embedding method.
For the paper, original software environment, and reproduction instructions, see SBERT-WK: A Sentence Embedding Method by Dissecting BERT-based Word Models.
Original model description
This model is pre-trained XLNET with 12 layers.
It comes with paper: SBERT-WK: A Sentence Embedding Method By Dissecting BERT-based Word Models
Project Page: SBERT-WK