EngineX-Iluvatar/enginex_bi_series-sherpa-onnx
Archived
16
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

RNNLM model support lm_num_thread and lm_provider setting (#173)

Browse Source 
* rnnlm model inference supports num_threads setting

* rnnlm params decouple num_thread and provider with Transducer.

* fix python csrc bug which offline-lm-config.cc and online-lm-config.cc arguments problem

* lm_num_threads and lm_provider set default values

---------

Co-authored-by: cuidongcai1035 <cuidongcai1035@wezhuiyi.com>
This commit is contained in:
keanu
2023-06-12 15:51:27 +08:00
committed by GitHub
parent 13b33fcc08
commit 1a1b9fd236
18 changed files with 67 additions and 31 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
sherpa-onnx/csrc/offline-recognizer-transducer-impl.h
View File

@@ -59,7 +59,7 @@ class OfflineRecognizerTransducerImpl : public OfflineRecognizerImpl {
std::make_unique<OfflineTransducerGreedySearchDecoder>(model_.get());
} else if (config_.decoding_method == "modified_beam_search") {
if (!config_.lm_config.model.empty()) {
lm_ = OfflineLM::Create(config);
lm_ = OfflineLM::Create(config.lm_config);
}
decoder_ = std::make_unique<OfflineTransducerModifiedBeamSearchDecoder>(