HyperAI

Open Domain Question Answering On Kilt 1

Metriken

EM
F1
KILT-EM
KILT-F1
R-Prec
Recall@5

Ergebnisse

Leistungsergebnisse verschiedener Modelle zu diesem Benchmark

Modellname
EM
F1
KILT-EM
KILT-F1
R-Prec
Recall@5
Paper TitleRepository
Sphere31.6441.550.00.00.00.0--
T5-base12.6419.570.00.00.00.0KILT: a Benchmark for Knowledge Intensive Language Tasks
Multi-task DPR0.00.00.00.042.9228.39--
BART15.3721.970.00.00.00.0--
GENRE0.00.00.00.051.2734.03--
BERT + DPR11.2917.350.741.2625.0410.4--
Multitask DPR + BART31.7741.569.5311.2742.9228.39--
intersect40.4651.4418.0621.4258.8351.03--
TABi0.00.00.00.053.1235.48--
Wikipedia36.947.6611.7113.8845.3835.75--
RAG26.9736.033.214.130.5912.59--
BART + DPR25.1834.071.962.5325.0410.4--
multi-task small3.296.840.00.00.00.0--
chriskuei0.00.00.00.051.834.57--
0 of 14 row(s) selected.