Enum CandleDevice 

pub enum CandleDevice {
    Cpu,
    Cuda,
    Metal,
    Auto,
}

Compute device for the Candle local-inference backend.

Determines which hardware accelerator is used for inference. Feature flags candle/metal and candle/cuda must be enabled at compile time for the corresponding variants to succeed at runtime.

Variants

Cpu

Run inference on the CPU.

Cuda

Run inference on an NVIDIA CUDA GPU (requires cuda feature).

Metal

Run inference on Apple Silicon via Metal (requires metal feature).

Auto

Auto-detect the best available device: Metal → CUDA → CPU.

Requires the corresponding metal or cuda feature to be enabled at compile time. Falls back to CPU when no GPU features are compiled in.

Trait Implementations

impl Clone for CandleDevice

fn clone(&self) -> CandleDevice

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for CandleDevice

impl Debug for CandleDevice

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

Formats the value using the given formatter.

impl Default for CandleDevice

fn default() -> CandleDevice

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for CandleDevice

fn deserialize<__D>( __deserializer: __D, ) -> Result<CandleDevice, <__D as Deserializer<'de>>::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Eq for CandleDevice

impl PartialEq for CandleDevice

fn eq(&self, other: &CandleDevice) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl Serialize for CandleDevice

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

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl StructuralPartialEq for CandleDevice