Differentiable Scheduled Sampling for Credit Assignment
원문 초록
We demonstrate that a continuous relaxation of the argmax operation can be used to create a differentiable approximation to greedy decoding for sequence-to-sequence (seq2seq) models. By incorporating this approximation into the scheduled sampling training procedure (Bengio et al., 2015)--a well-known technique for correcting exposure bias--we introduce a new training objective that is continuous and differentiable everywhere and that can provide informative gradients near points where previous decoding decisions change their value. In addition, by using a related approximation, we demonstrate a similar approach to sampled-based training. Finally, we show that our approach outperforms cross-entropy training and scheduled sampling procedures in two sequence prediction tasks: named entity recognition and machine translation.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Selecting and Merging: Towards Adaptable and Scalable Named Entity Recognition with Large Language Models
Flippi: End To End GenAI Assistant for E-Commerce
Transformer-Based Named Entity Recognition for Automated Server Provisioning
HEProto: A Hierarchical Enhancing ProtoNet based on Multi-Task Learning for Few-shot Named Entity Recognition
Legal Entity Extraction using a Pointer Generator Network
An end to end approach for brand recognition in product titles with BI-LSTM-CRF
토론
이 논문과 연결된 포럼 글이 아직 없습니다.