mistralrs

Trait Pipeline

source 
pub trait Pipeline:
    Send
    + Sync
    + PreProcessingMixin
    + IsqPipelineMixin
    + CacheManagerMixin
    + AdapterActivationMixin
    + MetadataMixin
    + AnyMoePipelineMixin {
    // Required methods
    fn forward_inputs(
        &mut self,
        inputs: Box<dyn Any>,
        return_raw_logits: bool,
    ) -> Result<ForwardInputsResult, Error>;
    fn sample_causal_gen<'life0, 'life1, 'life2, 'life3, 'async_trait>(
        &'life0 self,
        seqs: &'life1 mut [&'life2 mut Sequence],
        logits: Vec<Tensor>,
        prefix_cacher: &'life3 mut PrefixCacheManagerV2,
        disable_eos_stop: bool,
        rng: Arc<Mutex<Isaac64Rng>>,
    ) -> Pin<Box<dyn Future<Output = Result<(), Error>> + Send + 'async_trait>>
       where 'life0: 'async_trait,
             'life1: 'async_trait,
             'life2: 'async_trait,
             'life3: 'async_trait,
             Self: 'async_trait;
    fn category(&self) -> ModelCategory;

    // Provided method
    fn step<'life0, 'life1, 'life2, 'life3, 'life4, 'async_trait>(
        &'life0 mut self,
        input_seqs: &'life1 mut [&'life2 mut Sequence],
        is_prompt: bool,
        return_raw_logits: bool,
        prefix_cacher: &'life3 mut PrefixCacheManagerV2,
        disable_eos_stop: bool,
        rng: Arc<Mutex<Isaac64Rng>>,
        backend_metadata: CacheBackendMetadata<'life4>,
    ) -> Pin<Box<dyn Future<Output = Result<Duration, Error>> + Send + 'async_trait>>
       where 'life0: 'async_trait,
             'life1: 'async_trait,
             'life2: 'async_trait,
             'life3: 'async_trait,
             'life4: 'async_trait,
             Self: 'async_trait { ... }
}

Required Methods§

source

fn forward_inputs( &mut self, inputs: Box<dyn Any>, return_raw_logits: bool, ) -> Result<ForwardInputsResult, Error>

source

fn sample_causal_gen<'life0, 'life1, 'life2, 'life3, 'async_trait>( &'life0 self, seqs: &'life1 mut [&'life2 mut Sequence], logits: Vec<Tensor>, prefix_cacher: &'life3 mut PrefixCacheManagerV2, disable_eos_stop: bool, rng: Arc<Mutex<Isaac64Rng>>, ) -> Pin<Box<dyn Future<Output = Result<(), Error>> + Send + 'async_trait>>
where 'life0: 'async_trait, 'life1: 'async_trait, 'life2: 'async_trait, 'life3: 'async_trait, Self: 'async_trait,

source

fn category(&self) -> ModelCategory

Provided Methods§

source

fn step<'life0, 'life1, 'life2, 'life3, 'life4, 'async_trait>( &'life0 mut self, input_seqs: &'life1 mut [&'life2 mut Sequence], is_prompt: bool, return_raw_logits: bool, prefix_cacher: &'life3 mut PrefixCacheManagerV2, disable_eos_stop: bool, rng: Arc<Mutex<Isaac64Rng>>, backend_metadata: CacheBackendMetadata<'life4>, ) -> Pin<Box<dyn Future<Output = Result<Duration, Error>> + Send + 'async_trait>>
where 'life0: 'async_trait, 'life1: 'async_trait, 'life2: 'async_trait, 'life3: 'async_trait, 'life4: 'async_trait, Self: 'async_trait,

Returns the total of model execution time.

Implementors§