HyperAI초신경

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

한국어

HyperAI초신경

Discourse Parsing On Instructional Dt Instr

평가 지표

Standard Parseval (Full)

Standard Parseval (Nuclearity)

Standard Parseval (Relation)

Standard Parseval (Span)

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름	Standard Parseval (Full)	Standard Parseval (Nuclearity)	Standard Parseval (Relation)	Standard Parseval (Span)	Paper Title	Repository
Top-down (XLNet)	40.2	55.2	47.0	74.3	A Simple and Strong Baseline for End-to-End Neural RST-style Discourse Parsing
Bottom-up (DeBERTa)	44.4	60.0	51.4	77.8	A Simple and Strong Baseline for End-to-End Neural RST-style Discourse Parsing
Top-down (BERT)	30.9	44.6	37.6	65.3	A Simple and Strong Baseline for End-to-End Neural RST-style Discourse Parsing
Top-down (SpanBERT)	36.7	54.5	42.7	73.7	A Simple and Strong Baseline for End-to-End Neural RST-style Discourse Parsing
Top-down (DeBERTa)	43.4	57.9	50.0	77.3	A Simple and Strong Baseline for End-to-End Neural RST-style Discourse Parsing
Guz et al. (2020)	-	44.41	-	64.55	Unleashing the Power of Neural Discourse Parsers -- A Context and Structure Aware Approach Using Large Scale Pretraining	-
Bottom-up (SpanBERT)	40.5	53.8	46.0	72.9	A Simple and Strong Baseline for End-to-End Neural RST-style Discourse Parsing
Top-down (RoBERTa)	41.5	56.1	48.7	75.7	A Simple and Strong Baseline for End-to-End Neural RST-style Discourse Parsing
Bottom-up (BERT)	32.9	46.3	39.5	66.6	A Simple and Strong Baseline for End-to-End Neural RST-style Discourse Parsing
Guz et al. (2020) (pretrained)	-	46.59	-	65.41	Unleashing the Power of Neural Discourse Parsers -- A Context and Structure Aware Approach Using Large Scale Pretraining	-
Bottom-up (XLNet)	40.7	56.4	47.4	73.6	A Simple and Strong Baseline for End-to-End Neural RST-style Discourse Parsing
Bottom-up (RoBERTa)	41.4	55.5	47.9	73.2	A Simple and Strong Baseline for End-to-End Neural RST-style Discourse Parsing

0 of 12 row(s) selected.