ORKG Shapes

go to ORKG: http://orkg.org/orkg/predicate/P119133

Pretraining compute [PetaFLOPs-days]

FLOPs (Floating Point Operations) is a measure of computational performance, and PetaFLOPs represents one quadrillion (10^15) floating-point operations per second. In the context of language models like GPT-3, pretraining involves a significant amount of computation to train the model on a large corpus of text data, and the "PF-days" metric quantifies the total computational power used for pretraining, typically measured in PetaFLOPs multiplied by the number of days the computation took.

• http://orkg.org/orkg/shapes/orkgc_Resource
• http://orkg.org/orkg/shapes/orkgc_Contribution