Login / Signup
A Provably Effective Method for Pruning Experts in Fine-tuned Sparse Mixture-of-Experts.
Mohammed Nowaz Rabbani Chowdhury
Meng Wang
Kaoutar El Maghraoui
Naigang Wang
Pin-Yu Chen
Christopher D. Carothers
Published in:
CoRR (2024)
Keyphrases
</>
clustering method
synthetic data
fine tuning
high precision
significant improvement
dynamic programming
data sets
preprocessing
cost function
experimental evaluation
detection method
high accuracy
pruning method
dirichlet distribution
segmentation method
worst case
pairwise
similarity measure