neural_compressor.onnxrt.algorithms.layer_wise

Package Contents

Functions

layer_wise_quant(...)

Quantize model layer by layer to save memory.