aPaperADay
2021 / 08
47 CANINE, Pre-training an Efficient Tokenization-Free Encoder for Language Representation
08-13-2021
46 ByT5, Towards a token-free future with pre-trained byte-to-byte models
08-10-2021
45 Space-Time Correspondence as a Contrastive Random Walk
08-09-2021
44 🤗 Transformers
08-06-2021
43 The Evolved Transformer
08-05-2021
42 BlenderBot, Recipes for building an open-domain chatbot
08-04-2021
41 Big Bird, Transformers for Longer Sequences
08-03-2021
2021 / 07
40 ELECTRA
07-31-2021
39 GAUSSIAN ERROR LINEAR UNITS (GELUS)
07-27-2021
38 Are Sixteen Heads Really Better than One?
07-23-2021