FIMA-Q: Post-Training Quantization for Vision Transformers by Fisher Information Matrix Approximation

Devs

FIMA-Q: Post-Training Quantization for Vision Transformers by Fisher Information Matrix Approximation | Read Paper on Bytez