Forgetting in domain-adapted NMT is strongly linked to the adaptation dataset's target vocabulary coverage, and a minimal coverage-based mix-in of generic data prevents most of the loss.
Costa-juss \`a , Fethi Bougares, and Olivier Galibert
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Domain adapted machine translation: What does catastrophic forgetting forget and why?
Forgetting in domain-adapted NMT is strongly linked to the adaptation dataset's target vocabulary coverage, and a minimal coverage-based mix-in of generic data prevents most of the loss.