الرئيسية الأخبار أحدث الأوراق البحثية الدروس مجموعات البيانات الموسوعة SOTA نماذج LLM لوحة الأداء GPU الفعاليات

العربية

Machine Translation On Iwslt2015 German

المقاييس

BLEU score

النتائج

نتائج أداء النماذج المختلفة على هذا المعيار القياسي

اسم النموذج	BLEU score	Paper Title	Repository
Word-level CNN w/attn, input feeding	24.0	Sequence-to-Sequence Learning as Beam-Search Optimization
Conv-LSTM (deep+pos)	30.4	A Convolutional Encoder Model for Neural Machine Translation
QRNN	19.41	Quasi-Recurrent Neural Networks
FlowSeq-base	24.75	FlowSeq: Non-Autoregressive Conditional Sequence Generation with Generative Flow
Denoising autoencoders (non-autoregressive)	32.43	Deterministic Non-Autoregressive Neural Sequence Modeling by Iterative Refinement
Word-level LSTM w/attn	20.2	Sequence Level Training with Recurrent Neural Networks
ConvS2S	32.31	Convolutional Sequence to Sequence Learning
Transformer with FRAGE	33.97	FRAGE: Frequency-Agnostic Word Representation
ConvS2S+Risk	32.93	Classical Structured Prediction Losses for Sequence to Sequence Learning
RNNsearch	29.98	An Actor-Critic Algorithm for Sequence Prediction
Bi-GRU (MLE+SLE)	28.53	Neural Machine Translation by Jointly Learning to Align and Translate
Pervasive Attention	34.18	Pervasive Attention: 2D Convolutional Neural Networks for Sequence-to-Sequence Prediction
NPMT + language model	30.08	Towards Neural Phrase-based Machine Translation
PS-KD	36.20	Self-Knowledge Distillation with Progressive Refinement of Targets
DCCL	29.56	Compressing Word Embeddings via Deep Compositional Code Learning

0 of 15 row(s) selected.