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

Fix modified beam search for iOS and android (#76)

Browse Source 
* Use Int type for sampling rate

* Fix swift

* Fix iOS
This commit is contained in:
Fangjun Kuang
2023-03-03 15:18:31 +08:00
committed by GitHub
parent 7f72c13d9a
commit 5f31b22c12
15 changed files with 125 additions and 93 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
sherpa-onnx/csrc/sherpa-onnx.cc
View File

@@ -61,7 +61,7 @@ for a list of pre-trained models to download.
sherpa_onnx::OnlineRecognizer recognizer(config);
float expected_sampling_rate = config.feat_config.sampling_rate;
int32_t expected_sampling_rate = config.feat_config.sampling_rate;
bool is_ok = false;
std::vector<float> samples =
@@ -72,7 +72,7 @@ for a list of pre-trained models to download.
return -1;
}
float duration = samples.size() / expected_sampling_rate;
float duration = samples.size() / static_cast<float>(expected_sampling_rate);
fprintf(stderr, "wav filename: %s\n", wav_filename.c_str());
fprintf(stderr, "wav duration (s): %.3f\n", duration);