HyperAI초신경

홈 뉴스 최신 연구 논문 튜토리얼 데이터셋 백과사전 SOTA LLM 모델 GPU 랭킹 컨퍼런스

한국어

HyperAI초신경

Conversational Response Selection On Rrs

평가 지표

MAP

MRR

P@1

R10@1

R10@2

R10@5

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름	MAP	MRR	P@1	R10@1	R10@2	R10@5	Paper Title	Repository
BERT-FP	0.702	0.712	0.543	0.488	0.708	0.927	Fine-grained Post-training for Improving Retrieval-based Dialogue Systems
BERT	0.625	0.639	0.453	0.404	0.606	0.875	An Effective Domain Adaptive Post-Training Method for BERT in Response Selection
DAM	0.511	0.534	0.347	0.308	0.457	0.751	Multi-Turn Response Selection for Chatbots with Deep Attention Matching Network
SMN	0.487	0.501	0.309	0.281	0.442	0.723	Sequential Matching Network: A New Architecture for Multi-turn Response Selection in Retrieval-based Chatbots
SA-BERT+BERT-FP	0.701	0.715	0.555	0.497	0.685	0.931	Speaker-Aware BERT for Multi-Turn Response Selection in Retrieval-Based Chatbots
MSN	0.550	0.563	0.383	0.343	0.498	0.798	Multi-hop Selector Network for Multi-turn Response Selection in Retrieval-based Chatbots
SA-BERT+HCL	0.671	0.683	0.503	0.454	0.659	0.917	Dialogue Response Selection with Hierarchical Curriculum Learning

0 of 7 row(s) selected.