Past year
All results
- All results
- Verbatim
Oct 9, 2023 · Be- low, we present two distinct strategies to try harder to promote the terminology constraints, via auto- matic post-editing through constrained beam search.
Missing: Diverse | Show results with:Diverse
Feb 26, 2024 · We augment the ground-truth solutions of our seed data and train a back-translation model to translate the augmented solutions back into new questions.
Mar 22, 2024 · In this paper, we present Energy-based Constrained Decoding with Langevin Dynamics (COLD), a decoding framework which unifies constrained generation as ...
Dec 6, 2023 · We perform decoding experiments on document continua- tion and story generation, and demonstrate that Look-back is able to generate more fluent and coherent ...
Mar 14, 2024 · To combat label bias, we propose two constrained HMMs: 1) Adaptive Window HMM, which explicitly balances the number of outgoing transitions at different states; ...
Missing: Diverse | Show results with:Diverse
Dec 7, 2023 · Linguistic Decoder: Generates a phoneme sequence for the target speech ... The back-translation involves encoding the source input spectrogram, decoding ...
Missing: Diverse via
Sep 22, 2023 · Traditionally, pseudo-parallel corpora are generated by randomly replacing words in low-resource language machine translation. However, this method can ...
Feb 7, 2024 · We introduce Bi-SimCut: a simple but effective training strategy to boost neural machine translation (NMT) performance. It consists of two procedures: ...
Nov 22, 2023 · The paper proposes a decoding framework for language models that frames it as an optimization problem and performs decoding by optimizing towards desired ...
Nov 29, 2023 · A basic. NMT model is trained using adversarial training to create accurate translations that outperform those produced by the generator (G) and discriminator ( ...