RoBERTa-Large 355M

Reported on 3 benchmarks across 3 tasks · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing3 results

Question AnsweringonPIQA
Accuracy· 2019-07-26
79.4
best: 90.1 (Unicorn 11B (fine-tuned))
SOTA
RoBERTa: A Robustly Optimized BERT Pretraining Approach arXiv:1907.11692
Common Sense ReasoningonCommonsenseQA
Accuracy· 2019-07-26
72.1
best: 92.54 (GPT-4o (HPT))
SOTA
RoBERTa: A Robustly Optimized BERT Pretraining Approach arXiv:1907.11692
Sentence CompletiononHellaSwag
Accuracy· 2019-07-26
81.7
best: 96.1 (CompassMTL 567M with Tailor)
RoBERTa: A Robustly Optimized BERT Pretraining Approach arXiv:1907.11692