flex_smooth_quant Configuration Reference

1. Configuration Overview

FlexSmoothQuant smooth quantization processor configuration.

Item	Content
Config Class	FlexSmoothQuantProcessorConfig
Source	processor.py

2. Parameter List

2.1 FlexSmoothQuantProcessorConfig

Field Path	Type	Required/Optional	Default	Value Range or Format	Description	Referenced Config
type	string	Optional	flex_smooth_quant	flex_smooth_quant	Processor type, fixed to flex_smooth_quant.	None
alpha	float	Optional	null	—	Strength of the activation-to-weight smooth migration (0-1); a higher value migrates more outliers to the weights.	None
beta	float	Optional	null	—	Additional smoothing tuning coefficient (0-1), used together with alpha.	None
enable_subgraph_type	list[any]	Optional	['norm-linear', 'linear-linear', 'ov', 'up-down']	—	List of subgraph types to smooth; the default is norm-linear, linear-linear, ov, up-down.	None
include	list[string] / null	Optional	null	—	Module name patterns to include; if unset, all modules match.	None
exclude	list[string] / null	Optional	null	—	Module name patterns to exclude; takes precedence over include.	None

Configuration Constraints

• None.

3. Complete Configuration Reference

apiversion: modelslim_v1
spec:
  process:
  - type: flex_smooth_quant
    enable_subgraph_type:
    - norm-linear
    - linear-linear
    - ov
    - up-down