no code implementations • 27 Sep 2023 • Zihao Deng, Benjamin Ghaemmaghami, Ashish Kumar Singh, Benjamin Cho, Leo Orshansky, Mattan Erez, Michael Orshansky
At constant model quality, MLET allows embedding dimension, and model size, reduction by up to 16x, and 5. 8x on average, across the models.
no code implementations • 10 Jun 2020 • Benjamin Ghaemmaghami, Zihao Deng, Benjamin Cho, Leo Orshansky, Ashish Kumar Singh, Mattan Erez, Michael Orshansky
Increasing the dimension of embedding vectors improves model accuracy but comes at a high cost to model size.