Joint Source-Target Self Attention with Locality Constraints

José A. R. FonollosaNoe CasasMarta R. Costa-jussà

   Papers with code   Abstract  PDF

The dominant neural machine translation models are based on the encoder-decoder structure, and many of them rely on an unconstrained receptive field over source and target sequences. In this paper we study a new architecture that breaks with both conventions... (read more)

Benchmarked Models

RANK
MODEL
REPO
CODE RESULT
PAPER RESULT
ε-REPRODUCED
BUILD
1
Local Joint Self-attention
42.51
43.30
RANK
MODEL
REPO
CODE RESULT
PAPER RESULT
ε-REPRODUCED
BUILD
1
Local Joint Self-attention
29.59
29.70
RANK
MODEL
REPO
CODE RESULT
PAPER RESULT
ε-REPRODUCED
BUILD
1
Local Joint Self-attention
37.70
--