Fill-Mask2018en

GLUE

General Language Understanding Evaluation for masked language models

avg-score

#ModelScorePaper / CodeDate
1
DeBERTa-v3-largeOpen Source
Microsoft		91.37
DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing
Jan 2023
2
ALBERT-xxlarge-v2Open Source
Google		89.4
ALBERT: A Lite BERT for Self-supervised Learning of Language Representations
Feb 2020
3
RoBERTa-largeOpen Source
Facebook AI		88.5
RoBERTa: A Robustly Optimized BERT Pretraining Approach
Jul 2019

Related Papers3

DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing
Jan 2023Models: DeBERTa-v3-large
Paper
ALBERT: A Lite BERT for Self-supervised Learning of Language Representations
Feb 2020Models: ALBERT-xxlarge-v2
Paper
RoBERTa: A Robustly Optimized BERT Pretraining Approach
Jul 2019Models: RoBERTa-large
Paper