Abstractive Text Summarization On Cnn Daily

Metrics

ROUGE-1

ROUGE-2

ROUGE-L

Results

Performance results of various models on this benchmark

				Paper Title
Scrambled code + broken (alter)	48.18	19.84	45.35	Universal Evasion Attacks on Summarization Scoring
BRIO	47.78	23.55	44.57	BRIO: Bringing Order to Abstractive Summarization
Pegasus	47.36	24.02	44.45	Calibrating Sequence likelihood Improves Conditional Language Generation
PEGASUS + SummaReranker	47.16	22.61	43.87	SummaReranker: A Multi-Task Mixture-of-Experts Re-ranking Framework for Abstractive Summarization
Scrambled code + broken	46.71	20.39	43.56	Universal Evasion Attacks on Summarization Scoring
BART + SimCLS	46.67	22.15	43.54	SimCLS: A Simple Framework for Contrastive Learning of Abstractive Summarization
SEASON	46.27	22.64	43.08	Salience Allocation as Guidance for Abstractive Summarization
Fourier Transformer	44.76	21.55	41.34	Fourier Transformer: Fast Long Range Modeling by Removing Sequence Redundancy with FFT Operator
GLM-XXLarge	44.7	21.4	41.4	GLM: General Language Model Pretraining with Autoregressive Blank Infilling
BART + R-Drop	44.51	21.58	41.24	R-Drop: Regularized Dropout for Neural Networks
CoCoNet + CoCoPretrain	44.50	21.55	41.24	Learn to Copy from the Copying History: Correlational Copy Network for Abstractive Summarization
MUPPET BART Large	44.45	21.25	41.4	Muppet: Massive Multi-task Representations with Pre-Finetuning
CoCoNet	44.39	21.41	41.05	Learn to Copy from the Copying History: Correlational Copy Network for Abstractive Summarization
BART+R3F	44.38	21.53	41.17	Better Fine-Tuning by Reducing Representational Collapse
ERNIE-GENLARGE (large-scale text corpora)	44.31	21.35	41.60	ERNIE-GEN: An Enhanced Multi-Flow Pre-training and Fine-tuning Framework for Natural Language Generation
PALM	44.30	21.12	41.41	PALM: Pre-training an Autoencoding&Autoregressive Language Model for Context-conditioned Generation
ProphetNet	44.20	21.17	41.30	ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training
PEGASUS	44.17	21.47	41.11	PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization
BART	44.16	21.28	40.90	BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension
ERNIE-GENLARGE	44.02	21.17	41.26	ERNIE-GEN: An Enhanced Multi-Flow Pre-training and Fine-tuning Framework for Natural Language Generation

0 of 53 row(s) selected.

Command Palette

Abstractive Text Summarization On Cnn Daily

Metrics

Results