Struct QuantizedModel 

pub struct QuantizedModel {
    pub config_json: String,
    pub layers: Vec<QuantizedWeights>,
    pub scheme: String,
    pub total_quantized_bytes: usize,
    pub total_fp32_bytes: usize,
}

Collection of quantised weights for an entire model.

Fields

config_json: String

Model configuration JSON (serialised crate::config::SensorLMConfig).

layers: Vec<QuantizedWeights>

Quantised weights for every Linear layer in the model.

scheme: String

Quantisation scheme used.

total_quantized_bytes: usize

Total size of quantised weights in bytes.

total_fp32_bytes: usize

Total size of original FP32 weights in bytes.

Implementations

impl QuantizedModel

pub fn compression_ratio(&self) -> f32

Overall compression ratio (FP32 → INT8).

pub fn save(&self, path: &Path) -> Result<()>

Save the quantised model to a JSON file.

pub fn load(path: &Path) -> Result<Self>

Load a quantised model from a JSON file.

Trait Implementations

impl Clone for QuantizedModel

fn clone(&self) -> QuantizedModel

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for QuantizedModel

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for QuantizedModel

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for QuantizedModel

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.