List your company
ResearchFormerly WizWand

Natural Language Inference on 12 NLI tasks (CoNLL-2003, MIT-Restaurant, Amazon Reviews, Crowdflower) (test)

59.18Avg Test AccuracyPACMAML
46.793650.009353.22556.4407May 28, 2021
Updated 8mo ago

Evaluation results

MethodLinks
PACMAMLk (shots)=162021.0559.18
PACOHk (shots)=162021.0558.22
H-SMLMTk (shots)=162021.0557.9
MAMLk (shots)=162021.0557.38
BMAMLk (shots)=162021.0556.53
PACMAMLk (shots)=82021.0555.68
PACOHk (shots)=82021.0554.83
MAMLk (shots)=82021.0553.52
H-SMLMTk (shots)=82021.0552.92
BMAMLk (shots)=82021.0552.08
PACMAMLk (shots)=42021.0551.58
PACOHk (shots)=42021.0550.47
H-SMLMTk (shots)=42021.0548.61
MAMLk (shots)=42021.0548.21
BMAMLk (shots)=42021.0547.27