Struct QuantizationConfig 

pub struct QuantizationConfig {
    pub quant_method: String,
    pub quant_algo: String,
    pub format: String,
    pub ignore_modules: Vec<String>,
    pub precision: DeclaredPrecisionPlan,
}

2026-09-26: The weight-quantization block of a checkpoint config (quantization_config, into which the server merges a sibling hf_quant_config.json), as read by parse_quantization_config. That parser returns None when the block declares no method, algorithm or ignore list.

Fields

quant_method: String

2026-09-26: quant_method as written, modelopt for a ModelOpt sidecar that omits it, or empty.

quant_algo: String

2026-09-26: quant_algo, else NVFP4/FP8 derived from config_groups.group_0.weights, else empty.

format: String

2026-09-26: format, or empty.

ignore_modules: Vec<String>

2026-09-26: Module-path patterns from ignore, then those from exclude_modules not already listed.

precision: DeclaredPrecisionPlan

2026-09-28: What the block declares per linear layer, weights and activations: the default precision every loader and kernel dispatch follows.

Trait Implementations

impl Clone for QuantizationConfig

fn clone(&self) -> QuantizationConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for QuantizationConfig

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.