Downloads · 30 days
96
2% of all-time downloads
mirror-bert-base-uncased-word
cambridgeltl/mirror-bert-base-uncased-word
mirror-bert-base-uncased-word is a feature extraction model from cambridgeltl. Use it when you need embeddings to search or compare text. It is set up for transformers.
All-time downloads
4.2K
Public
Parameters
109M
876 MB on disk
Likes
1
Public
Repo makeup
Click a slice to open those files.
.bin438 MB · 50%
Parameter types
How the weights are stored.
F32109M · 100%
Model card
From the Hugging Face model README
language: en
tags:
- word-embeddings
- word-similarity
mirror-bert-base-uncased-word
An unsupervised word encoder proposed by Liu et al. (2021). Trained with a set of unlabelled words, using bert-base-uncased as the base model. Please use [CLS] as the representation of the input.
Citation
@inproceedings{
liu2021fast,
title={Fast, Effective and Self-Supervised: Transforming Masked LanguageModels into Universal Lexical and Sentence Encoders},
author={Liu, Fangyu and Vuli{\'c}, Ivan and Korhonen, Anna and Collier, Nigel},
booktitle={EMNLP 2021},
year={2021}
}
- Task
- Feature Extraction
- Library
- transformers
- Type
- bert
- Created
- Mar 2, 2022
- Updated
- Apr 5, 2023