HyperAIHyperAI초신경
홈플랫폼문서뉴스연구 논문튜토리얼데이터셋백과사전SOTALLM 모델GPU 랭킹컨퍼런스
전체 검색
소개
HyperAIHyperAI초신경

Command Palette

Search for a command to run...

Machine Translation On Wmt2016 English 1

평가 지표

BLEU score

평가 결과

이 벤치마크에서 각 모델의 성능 결과

Paper TitleRepository
DeLighT34.7DeLighT: Deep and Light-weight Transformer
CMLM+LAT+4 iterations32.87Incorporating a Local Translation Mechanism into Non-autoregressive Translation
FlowSeq-large (NPD n = 30)32.35FlowSeq: Non-Autoregressive Conditional Sequence Generation with Generative Flow
FlowSeq-large (NPD n=15)31.97FlowSeq: Non-Autoregressive Conditional Sequence Generation with Generative Flow
FlowSeq-large (IWD n = 15)31.08FlowSeq: Non-Autoregressive Conditional Sequence Generation with Generative Flow
CMLM+LAT+1 iterations30.74Incorporating a Local Translation Mechanism into Non-autoregressive Translation
ConvS2S BPE40k29.9Convolutional Sequence to Sequence Learning
FlowSeq-large29.86FlowSeq: Non-Autoregressive Conditional Sequence Generation with Generative Flow
NAT +FT + NPD29.79Non-Autoregressive Neural Machine Translation
Denoising autoencoders (non-autoregressive)29.66Deterministic Non-Autoregressive Neural Sequence Modeling by Iterative Refinement
FlowSeq-base29.26FlowSeq: Non-Autoregressive Conditional Sequence Generation with Generative Flow
GRU BPE90k28.9--
BiGRU28.1Edinburgh Neural Machine Translation Systems for WMT 16
Deep Convolutional Encoder; single-layer decoder27.8A Convolutional Encoder Model for Neural Machine Translation
BiLSTM27.5A Convolutional Encoder Model for Neural Machine Translation
PBSMT + NMT25.13Phrase-Based & Neural Unsupervised Machine Translation
Unsupervised PBSMT21.33Phrase-Based & Neural Unsupervised Machine Translation
Unsupervised NMT + Transformer21.18Phrase-Based & Neural Unsupervised Machine Translation
FLAN 137B (few-shot, k=9)20.5Finetuned Language Models Are Zero-Shot Learners
FLAN 137B (zero-shot)18.9Finetuned Language Models Are Zero-Shot Learners
0 of 21 row(s) selected.