From 52ae03b614d5720b2a57daf371fc8083cd1e9193 Mon Sep 17 00:00:00 2001 From: Fangjun Kuang Date: Sun, 12 May 2024 11:50:26 +0800 Subject: [PATCH 1/2] Install naudiodon2 manually. It is needed only when using a microphone. The CI tests don't need it. --- nodejs-addon-examples/README.md | 7 +++++++ nodejs-addon-examples/package.json | 1 - 2 files changed, 7 insertions(+), 1 deletion(-) diff --git a/nodejs-addon-examples/README.md b/nodejs-addon-examples/README.md index 81eea42eb9..058a19a745 100644 --- a/nodejs-addon-examples/README.md +++ b/nodejs-addon-examples/README.md @@ -36,6 +36,9 @@ rm sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20.tar.bz2 node ./test_asr_streaming_transducer.js +# To run the test with microphone, you need to install the package naudiodon2 +npm install naudiodon2 + node ./test_asr_streaming_transducer_microphone.js ``` @@ -44,6 +47,10 @@ node ./test_asr_streaming_transducer_microphone.js ```bash wget https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/silero_vad.onnx + +# To run the test with microphone, you need to install the package naudiodon2 +npm install naudiodon2 + node ./test_vad_microphone.js ``` diff --git a/nodejs-addon-examples/package.json b/nodejs-addon-examples/package.json index 82958bd7d5..70d74d429c 100644 --- a/nodejs-addon-examples/package.json +++ b/nodejs-addon-examples/package.json @@ -1,6 +1,5 @@ { "dependencies": { - "naudiodon2": "^2.4.0", "perf_hooks": "*", "sherpa-onnx-node": "*" } From 469b08bf419393af77e56f98aa6f2068f2925baa Mon Sep 17 00:00:00 2001 From: Fangjun Kuang Date: Sun, 12 May 2024 11:56:36 +0800 Subject: [PATCH 2/2] small fixes --- scripts/apk/generate-tts-apk-script.py | 2 +- scripts/node-addon-api/test/test_asr_streaming_transducer.js | 4 ++-- sherpa-onnx/csrc/online-ctc-fst-decoder.cc | 2 +- 3 files changed, 4 insertions(+), 4 deletions(-) diff --git a/scripts/apk/generate-tts-apk-script.py b/scripts/apk/generate-tts-apk-script.py index 8c19a5e4e9..2bf7b74d44 100755 --- a/scripts/apk/generate-tts-apk-script.py +++ b/scripts/apk/generate-tts-apk-script.py @@ -107,7 +107,7 @@ def get_piper_models() -> List[TtsModel]: TtsModel(model_dir="vits-piper-de_DE-eva_k-x_low"), TtsModel(model_dir="vits-piper-de_DE-karlsson-low"), TtsModel(model_dir="vits-piper-de_DE-kerstin-low"), - TtsModel(model_dir="vits-piper-de_DE-mls-medium"), + # TtsModel(model_dir="vits-piper-de_DE-mls-medium"), TtsModel(model_dir="vits-piper-de_DE-pavoque-low"), TtsModel(model_dir="vits-piper-de_DE-ramona-low"), TtsModel(model_dir="vits-piper-de_DE-thorsten-high"), diff --git a/scripts/node-addon-api/test/test_asr_streaming_transducer.js b/scripts/node-addon-api/test/test_asr_streaming_transducer.js index 67684993ef..908ac74773 100644 --- a/scripts/node-addon-api/test/test_asr_streaming_transducer.js +++ b/scripts/node-addon-api/test/test_asr_streaming_transducer.js @@ -35,10 +35,10 @@ console.log('Started') let start = performance.now(); const stream = recognizer.createStream(); const wave = sherpa_onnx.readWave(waveFilename); -stream.acceptWaveform(wave.samples, wave.sampleRate); +stream.acceptWaveform({samples: wave.samples, sampleRate: wave.sampleRate}); const tailPadding = new Float32Array(wave.sampleRate * 0.4); -stream.acceptWaveform(tailPadding, wave.sampleRate); +stream.acceptWaveform({samples: tailPadding, sampleRate: wave.sampleRate}); while (recognizer.isReady(stream)) { recognizer.decode(stream); diff --git a/sherpa-onnx/csrc/online-ctc-fst-decoder.cc b/sherpa-onnx/csrc/online-ctc-fst-decoder.cc index 7619e0db58..93e4c103b5 100644 --- a/sherpa-onnx/csrc/online-ctc-fst-decoder.cc +++ b/sherpa-onnx/csrc/online-ctc-fst-decoder.cc @@ -62,7 +62,7 @@ static void DecodeOne(const float *log_probs, int32_t num_rows, std::ostringstream os; int32_t prev_id = -1; - int32_t num_trailing_blanks = 0; + int32_t &num_trailing_blanks = result->num_trailing_blanks; int32_t f = 0; // frame number for (auto i : isymbols_out) {