Compression of Generative Pre-trained Language Models via Quantization

39050
Back to Articles Dashboard

Knowledge Triples

Source Entity Relationship Target Entity
BART

Method

 achieves
GPT-2

Material

 achieves