Ad astra or astray: Exploring linguistic knowledge of multilingual BERT through NLI task
Sign Up to like & getrecommendations! Published in 2022 at "Natural Language Engineering"
DOI: 10.1017/s1351324922000225
Abstract: Abstract Recent research has reported that standard fine-tuning approaches can be unstable due to being prone to various sources of randomness, including but not limited to weight initialization, training data order, and hardware. Such brittleness… read more here.
Keywords: training data; astra astray; task; fine tuning ... See more keywords