List your company

Common Sense Reasoning on ARC, BoolQ, RTE, Winogrande, and TruthfulQA

34ARC Challenge AccuracyOpenLLaMA (V1)
22.5625.5328.531.47Dec 28, 2023

Evaluation results

MethodLinks
OpenLLaMA (V1)Parameters=3B2023.1234696858622235
INCITE (V1)Parameters=3B2023.1232686752632133
MobileLLaMAParameters=2.7B2023.1232686159632336
GalacticaParameters=1.3B2023.1228626252552541
PythiaParameters=1.4B2023.1226616352572339
MobileLLaMAParameters=1.4B2023.1226615353592135
OPTParameters=1.3B2023.1224575651592439
TinyLLaMA (2T)Parameters=1.1B2023.1223575955572339