Configuration for offline ASR. More...

#include <cxx-api.h>

Collaboration diagram for sherpa_onnx::cxx::OfflineRecognizerConfig:

Public Attributes
FeatureConfig	feat_config

OfflineModelConfig	model_config

OfflineLMConfig	lm_config

std::string	decoding_method = "greedy_search"

int32_t	max_active_paths = 4

std::string	hotwords_file

float	hotwords_score = 1.5

std::string	rule_fsts

std::string	rule_fars

float	blank_penalty = 0

HomophoneReplacerConfig	hr

Detailed Description

Example using SenseVoice:

OfflineRecognizerConfig config;
config.model_config.sense_voice.model =
    "./sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17-int8/model.int8.onnx";
config.model_config.sense_voice.language = "auto";
config.model_config.sense_voice.use_itn = true;
config.model_config.tokens =
    "./sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17-int8/tokens.txt";
config.model_config.num_threads = 1;

Example using Parakeet TDT v2:

OfflineRecognizerConfig config;
config.model_config.transducer.encoder =
    "./sherpa-onnx-nemo-parakeet-tdt-0.6b-v2-int8/encoder.int8.onnx";
config.model_config.transducer.decoder =
    "./sherpa-onnx-nemo-parakeet-tdt-0.6b-v2-int8/decoder.int8.onnx";
config.model_config.transducer.joiner =
    "./sherpa-onnx-nemo-parakeet-tdt-0.6b-v2-int8/joiner.int8.onnx";
config.model_config.tokens =
    "./sherpa-onnx-nemo-parakeet-tdt-0.6b-v2-int8/tokens.txt";
config.model_config.model_type = "nemo_transducer";
config.model_config.num_threads = 1;

Definition at line 723 of file cxx-api.h.

Member Data Documentation

◆ blank_penalty

float sherpa_onnx::cxx::OfflineRecognizerConfig::blank_penalty = 0

Optional blank penalty applied during decoding.

Definition at line 746 of file cxx-api.h.

◆ decoding_method

std::string sherpa_onnx::cxx::OfflineRecognizerConfig::decoding_method = "greedy_search"

Decoding method such as "greedy_search" or "modified_beam_search".

Definition at line 732 of file cxx-api.h.

◆ feat_config

FeatureConfig sherpa_onnx::cxx::OfflineRecognizerConfig::feat_config

Feature extraction configuration.

Definition at line 725 of file cxx-api.h.

◆ hotwords_file

std::string sherpa_onnx::cxx::OfflineRecognizerConfig::hotwords_file

Optional hotword file.

Definition at line 737 of file cxx-api.h.

◆ hotwords_score

float sherpa_onnx::cxx::OfflineRecognizerConfig::hotwords_score = 1.5

Hotword boost score.

Definition at line 740 of file cxx-api.h.

◆ hr

HomophoneReplacerConfig sherpa_onnx::cxx::OfflineRecognizerConfig::hr

Optional homophone replacement configuration.

Definition at line 748 of file cxx-api.h.

◆ lm_config

OfflineLMConfig sherpa_onnx::cxx::OfflineRecognizerConfig::lm_config

Optional LM configuration.

Definition at line 729 of file cxx-api.h.

◆ max_active_paths

int32_t sherpa_onnx::cxx::OfflineRecognizerConfig::max_active_paths = 4

Maximum number of active paths for beam-search-style decoding.

Definition at line 734 of file cxx-api.h.

◆ model_config

OfflineModelConfig sherpa_onnx::cxx::OfflineRecognizerConfig::model_config

Acoustic model configuration.

Definition at line 727 of file cxx-api.h.

◆ rule_fars

std::string sherpa_onnx::cxx::OfflineRecognizerConfig::rule_fars

Optional ITN rule FAR archive.

Definition at line 744 of file cxx-api.h.

◆ rule_fsts

std::string sherpa_onnx::cxx::OfflineRecognizerConfig::rule_fsts

Optional ITN rule FST archive.

Definition at line 742 of file cxx-api.h.

The documentation for this struct was generated from the following file:

c-api/cxx-api.h

Public Attributes