E-Sparse: Boosting the Large Language Model Inference through Entropy-based N:M Sparsity

MoE
Re-assign community