GitHub topics: compute-efficient-ml
decile-team/cords
Reduce end to end training time from days to hours (or hours to minutes), and energy requirements/costs by an order of magnitude using coresets and data selection.
langage: Jupyter Notebook - taille: 57,5 Mo - dernière synchronisation: il y a environ un mois - enregistré: il y a environ 2 ans - étoiles: 330 - forks: 55
pittisl/Sparse-Activation
Code for paper "Achieving Sparse Activation in Small Language Models"
langage: Python - taille: 1,37 Mo - dernière synchronisation: il y a 12 mois - enregistré: il y a 12 mois - étoiles: 0 - forks: 0
