Struct TranslationConfig

pub struct TranslationConfig {

    pub model_type: ModelType,
    pub model_resource: ModelResource,
    pub config_resource: Box<dyn ResourceProvider + Send>,
    pub vocab_resource: Box<dyn ResourceProvider + Send>,
    pub merges_resource: Option<Box<dyn ResourceProvider + Send>>,
    pub source_languages: HashSet<Language>,
    pub target_languages: HashSet<Language>,
    pub min_length: i64,
    pub max_length: Option<i64>,
    pub do_sample: bool,
    pub early_stopping: bool,
    pub num_beams: i64,
    pub temperature: f64,
    pub top_k: i64,
    pub top_p: f64,
    pub repetition_penalty: f64,
    pub length_penalty: f64,
    pub no_repeat_ngram_size: i64,
    pub num_return_sequences: i64,
    pub device: Device,
    pub num_beam_groups: Option<i64>,
    pub diversity_penalty: Option<f64>,
    pub kind: Option<Kind>,
}

Configuration for text translation

Contains information regarding the model to load, mirrors the GenerationConfig, with a different set of default parameters and sets the device to place the model on.

Fields

model_type: ModelType

Model type used for translation

model_resource: ModelResource

Model weights resource

config_resource: Box<dyn ResourceProvider + Send>

Config resource

vocab_resource: Box<dyn ResourceProvider + Send>

Vocab resource

merges_resource: Option<Box<dyn ResourceProvider + Send>>

Merges resource

source_languages: HashSet<Language>

Supported source languages

target_languages: HashSet<Language>

Supported target languages

min_length: i64

Minimum sequence length (default: 0)

max_length: Option<i64>

Maximum sequence length (default: 512)

do_sample: bool

Sampling flag. If true, will perform top-k and/or nucleus sampling on generated tokens, otherwise greedy (deterministic) decoding (default: true)

early_stopping: bool

Early stopping flag indicating if the beam search should stop as soon as num_beam hypotheses have been generated (default: false)

num_beams: i64

Number of beams for beam search (default: 5)

temperature: f64

Temperature setting. Values higher than 1 will improve originality at the risk of reducing relevance (default: 1.0)

top_k: i64

Top_k values for sampling tokens. Value higher than 0 will enable the feature (default: 0)

top_p: f64

Top_p value for Nucleus sampling, Holtzman et al.. Keep top tokens until cumulative probability reaches top_p (default: 0.9)

repetition_penalty: f64

Repetition penalty (mostly useful for CTRL decoders). Values higher than 1 will penalize tokens that have been already generated. (default: 1.0)

length_penalty: f64

Exponential penalty based on the length of the hypotheses generated (default: 1.0)

no_repeat_ngram_size: i64

Number of allowed repetitions of n-grams. Values higher than 0 turn on this feature (default: 3)

num_return_sequences: i64

Number of sequences to return for each prompt text (default: 1)

device: Device

Device to place the model on (default: CUDA/GPU when available)

num_beam_groups: Option<i64>

Number of beam groups for diverse beam generation. If provided and higher than 1, will split the beams into beam subgroups leading to more diverse generation.

diversity_penalty: Option<f64>

Diversity penalty for diverse beam search. High values will enforce more difference between beam groups (default: 5.5)

kind: Option<Kind>

Model weights precision. If not provided, will default to full precision on CPU, or the loaded weights precision otherwise

Implementations

impl TranslationConfig

pub fn new<RC, RV, S, T>( model_type: ModelType, model_resource: ModelResource, config_resource: RC, vocab_resource: RV, merges_resource: Option<RV>, source_languages: S, target_languages: T, device: impl Into<Option<Device>>, ) -> TranslationConfig

where RC: ResourceProvider + Send + 'static, RV: ResourceProvider + Send + 'static, S: AsRef<[Language]>, T: AsRef<[Language]>,

Create a new TranslationConfiguration from an available language.

Arguments

• language - Language enum value (e.g. Language::EnglishToFrench)
• device - Device to place the model on (CPU/GPU)

Example

use rust_bert::marian::{
    MarianConfigResources, MarianModelResources, MarianSourceLanguages, MarianSpmResources,
    MarianTargetLanguages, MarianVocabResources,
};
use rust_bert::pipelines::common::{ModelResource, ModelType};
use rust_bert::pipelines::translation::TranslationConfig;
use rust_bert::resources::RemoteResource;
use tch::Device;

let model_resource = ModelResource::Torch(Box::new(RemoteResource::from_pretrained(
    MarianModelResources::ROMANCE2ENGLISH,
)));
let config_resource = RemoteResource::from_pretrained(MarianConfigResources::ROMANCE2ENGLISH);
let vocab_resource = RemoteResource::from_pretrained(MarianVocabResources::ROMANCE2ENGLISH);
let spm_resource = RemoteResource::from_pretrained(MarianSpmResources::ROMANCE2ENGLISH);

let source_languages = MarianSourceLanguages::ROMANCE2ENGLISH;
let target_languages = MarianTargetLanguages::ROMANCE2ENGLISH;

let translation_config = TranslationConfig::new(
    ModelType::Marian,
    model_resource,
    config_resource,
    vocab_resource,
    Some(spm_resource),
    source_languages,
    target_languages,
    Device::cuda_if_available(),
);

Trait Implementations

impl From<TranslationConfig> for GenerateConfig

fn from(config: TranslationConfig) -> GenerateConfig

Converts to this type from the input type.