Struct InferenceConfig 

pub struct InferenceConfig {
    pub compute_backend: ComputeBackend,
    pub quantization: Quantization,
    pub batch_size: usize,
    pub thread_count: usize,
    pub block_size: usize,
    pub use_flash_attention: bool,
    pub device_type: DeviceType,
    pub dtype: DType,
}

Configuration generated by auto-detection

Fields

compute_backend: ComputeBackend

Selected compute backend

quantization: Quantization

Recommended quantization

batch_size: usize

Recommended batch size

thread_count: usize

Recommended thread count for CPU inference

block_size: usize

Optimal block size for attention

use_flash_attention: bool

Whether to use flash attention

device_type: DeviceType

Device type for the backend crate

dtype: DType

Data type for tensors

Implementations

impl InferenceConfig

pub fn auto() -> Self

Create an auto-configured inference config

pub fn low_memory() -> Self

Create a config optimized for low memory usage

pub fn high_throughput() -> Self

Create a config optimized for high throughput

pub fn low_latency() -> Self

Create a config optimized for low latency

pub fn estimated_tokens_per_second(&self) -> f32

Get estimated tokens per second for this configuration

pub fn power_efficient() -> Self

Create a config optimized for power efficiency (uses ANE when available)

Trait Implementations

impl Clone for InferenceConfig

fn clone(&self) -> InferenceConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for InferenceConfig

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

Formats the value using the given formatter.

impl Default for InferenceConfig

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for InferenceConfig

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for InferenceConfig

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

where __S: Serializer,

Serialize this value into the given Serde serializer.