WeNet CTC-based models

This page lists all offline CTC models from WeNet.

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10 (Cantonese, 粤语)

This model is converted from

https://huggingface.co/ASLP-lab/WSYue-ASR/tree/main/u2pp_conformer_yue

It uses 21.8k hours of training data.

Hint

If you want a Cantonese ASR model, please choose this model or sherpa-onnx-sense-voice-zh-en-ja-ko-yue-int8-2025-09-09 (Chinese, English, Japanese, Korean, Cantonese, 中英日韩粤语)

In the following, we describe how to use it.

Huggingface space

You can visit

https://huggingface.co/spaces/k2-fsa/automatic-speech-recognition

to try this model in your browser.

Hint

You need to first select the language Cantonese and then select the model csukuangfj/sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10.

Android APKs

Real-time speech recognition Android APKs can be found at

https://k2-fsa.github.io/sherpa/onnx/android/apk-simulate-streaming-asr.html

Please always download the latest version.

Hint

Please search for wenetspeech_yue_u2pconformer_ctc_2025_09_10_int8.apk in the above page, e.g., sherpa-onnx-1.12.11-arm64-v8a-simulated_streaming_asr-zh_en_yue-wenetspeech_yue_u2pconformer_ctc_2025_09_10_int8.apk.

Hint

For Chinese users, you can also visit https://k2-fsa.github.io/sherpa/onnx/android/apk-simulate-streaming-asr-cn.html

Download

Please use the following commands to download it:

cd /path/to/sherpa-onnx

wget https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10.tar.bz2
tar xf sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10.tar.bz2
rm sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10.tar.bz2

After downloading, you should find the following files:

ls -lh sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/

total 263264
-rw-r--r--   1 fangjun  staff   129B Sep 10 14:18 README.md
-rw-r--r--   1 fangjun  staff   128M Sep 10 14:18 model.int8.onnx
drwxr-xr-x  22 fangjun  staff   704B Sep 10 14:18 test_wavs
-rw-r--r--   1 fangjun  staff    83K Sep 10 14:18 tokens.txt

ls  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/

en.wav     yue-1.wav  yue-11.wav yue-13.wav yue-15.wav yue-17.wav yue-3.wav  yue-5.wav  yue-7.wav  yue-9.wav
yue-0.wav  yue-10.wav yue-12.wav yue-14.wav yue-16.wav yue-2.wav  yue-4.wav  yue-6.wav  yue-8.wav  zh.wav

In the following, we show how to decode the files sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-*.wav.

yue-0.wav

Wave filename	Content	Ground truth
yue-0.wav		两只小企鹅都有嘢食

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-0.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-0.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-0.wav
{"lang": "", "emotion": "", "event": "", "text": "两只小企鹅都有嘢食", "timestamps": [0.48, 0.68, 0.92, 1.16, 1.36, 1.84, 2.00, 2.20, 2.40], "tokens":["两", "只", "小", "企", "鹅", "都", "有", "嘢", "食"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 0.226 s
Real time factor (RTF): 0.226 / 3.072 = 0.074

yue-1.wav

Wave filename	Content	Ground truth
yue-1.wav		叫做诶诶直入式你个脑部里边咧记得呢一个嘅以前香港有一个广告好出名嘅佢乜嘢都冇噶净系影住喺弥敦道佢哋间铺头嘅啫但系就不停有人嗌啦平平吧平吧

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-1.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-1.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-1.wav
{"lang": "", "emotion": "", "event": "", "text": "叫做诶诶直入式你个脑部里边咧记得呢一个嘅以前香港有一个广告好出名嘅佢乜嘢都冇噶净系影住喺弥敦道佢哋间铺头嘅啫但系就不停有人嗌啦平平吧平吧", "timestamps": [0.04, 0.16, 0.36, 0.84, 1.16, 1.40, 1.64, 1.88, 2.00, 2.24, 2.56, 2.76, 2.92, 3.08, 3.28, 3.44, 3.60, 3.68, 3.80, 4.00, 4.20, 4.36, 4.52, 4.64, 4.76, 4.84, 4.92, 5.04, 5.16, 5.32, 5.48, 5.64, 5.88, 6.48, 6.64, 6.80, 6.92, 7.08, 7.24, 7.60, 7.72, 7.88, 8.04, 8.16, 8.36, 8.52, 8.72, 8.88, 9.00, 9.20, 9.36, 9.48, 9.64, 9.80, 10.12, 10.20, 10.32, 10.52, 10.64, 10.80, 10.88, 11.04, 11.24, 12.04, 12.84, 13.08, 13.96, 14.20], "tokens":["叫", "做", "诶", "诶", "直", "入", "式", "你", "个", "脑", "部", "里", "边", "咧", "记", "得", "呢", "一", "个", "嘅", "以", "前", "香", "港", "有", "一", "个", "广", "告", "好", "出", "名", "嘅", "佢", "乜", "嘢", "都", "冇", "噶", "净", "系", "影", "住", "喺", "弥", "敦", "道", "佢", "哋", "间", "铺", "头", "嘅", "啫", "但", "系", "就", "不", "停", "有", "人", "嗌", "啦", "平", "平", "吧", "平", "吧"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 1.185 s
Real time factor (RTF): 1.185 / 15.104 = 0.078

yue-2.wav

Wave filename	Content	Ground truth
yue-2.wav		忽然从光线死角嘅阴影度窜出一只大猫

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-2.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-2.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-2.wav
{"lang": "", "emotion": "", "event": "", "text": "忽然从光线死角嘅阴影度传出一只大猫", "timestamps": [0.44, 0.56, 1.16, 1.36, 1.64, 1.92, 2.12, 2.24, 2.36, 2.56, 2.80, 3.16, 3.36, 3.52, 3.64, 3.80, 3.96], "tokens":["忽", "然", "从", "光", "线", "死", "角", "嘅", "阴", "影", "度", "传", "出", "一", "只", "大", "猫"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 0.369 s
Real time factor (RTF): 0.369 / 4.608 = 0.080

yue-3.wav

Wave filename	Content	Ground truth
yue-3.wav		今日我带大家去见识一位九零后嘅靓仔咧

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-3.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-3.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-3.wav
{"lang": "", "emotion": "", "event": "", "text": "今日我带大家去见识一位九零后嘅靓仔咧", "timestamps": [0.32, 0.48, 0.60, 0.72, 0.92, 1.08, 1.56, 1.76, 1.96, 2.12, 2.24, 2.56, 2.80, 3.04, 3.20, 3.36, 3.56, 3.80], "tokens":["今", "日", "我", "带", "大", "家", "去", "见", "识", "一", "位", "九", "零", "后", "嘅", "靓", "仔", "咧"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 0.380 s
Real time factor (RTF): 0.380 / 4.352 = 0.087

yue-4.wav

Wave filename	Content	Ground truth
yue-4.wav		香港嘅消费市场从此不一样

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-4.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-4.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-4.wav
{"lang": "", "emotion": "", "event": "", "text": "香港嘅消费市场从此不一样", "timestamps": [0.44, 0.64, 0.80, 0.96, 1.16, 1.44, 1.64, 1.96, 2.16, 2.44, 2.64, 2.80], "tokens":["香", "港", "嘅", "消", "费", "市", "场", "从", "此", "不", "一", "样"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 0.228 s
Real time factor (RTF): 0.228 / 3.200 = 0.071

yue-5.wav

Wave filename	Content	Ground truth
yue-5.wav		景天谂唔到呢个守门嘅弟子竟然咁无礼霎时间面色都变埋

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-5.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-5.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-5.wav
{"lang": "", "emotion": "", "event": "", "text": "景天谂唔到呢个守门嘅弟子竟然咁无礼霎时间面色都变埋", "timestamps": [0.52, 0.72, 1.00, 1.12, 1.24, 1.40, 1.52, 1.68, 1.92, 2.08, 2.20, 2.40, 3.12, 3.28, 3.52, 3.92, 4.12, 5.00, 5.24, 5.40, 5.72, 5.92, 6.08, 6.28, 6.52], "tokens":["景", "天", "谂", "唔", "到", "呢", "个", "守", "门", "嘅", "弟", "子", "竟", "然", "咁", "无", "礼", "霎", "时", "间", "面", "色", "都", "变", "埋"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 0.551 s
Real time factor (RTF): 0.551 / 7.168 = 0.077

yue-6.wav

Wave filename	Content	Ground truth
yue-6.wav		六个星期嘅课程包括六堂课同两个测验你唔掌握到基本嘅十九个声母五十六个韵母同九个声调我哋仲针对咗广东话学习者会遇到嘅大樽颈啊以国语为母语人士最难掌握嘅五大韵母教课书唔会教你嘅七种变音同十种变调说话生硬唔自然嘅根本性问题提供全新嘅学习方向等你突破难关

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-6.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-6.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-6.wav
{"lang": "", "emotion": "", "event": "", "text": "六个星期嘅课程包括六堂课同两个测验你只掌握到基本嘅十九个声母五十六个韵母同九个声调我哋仲针对咗广东话学习者会遇到嘅大樽颈啊以国语为母语人士最难掌握嘅五大韵母教课书唔会教你嘅七种变音同十种变调说话生硬唔自然嘅根本性问题提供全新嘅学习方向等你突破难关", "timestamps": [0.52, 0.68, 0.92, 1.12, 1.32, 1.44, 1.64, 2.20, 2.40, 2.60, 2.80, 3.04, 3.48, 3.68, 3.84, 4.08, 4.28, 4.92, 5.20, 5.36, 5.52, 5.68, 5.92, 6.12, 6.36, 6.64, 6.84, 7.00, 7.12, 7.32, 7.68, 7.88, 8.04, 8.16, 8.28, 8.52, 8.96, 9.20, 9.40, 9.56, 9.72, 10.16, 10.32, 10.48, 10.60, 10.76, 10.92, 11.16, 11.36, 11.56, 11.72, 11.88, 12.08, 12.44, 12.64, 12.84, 13.04, 13.56, 13.80, 14.04, 14.28, 14.68, 14.84, 15.04, 15.24, 15.48, 15.60, 15.76, 15.96, 16.44, 16.68, 16.92, 17.12, 17.32, 17.76, 17.92, 18.08, 18.32, 18.80, 19.08, 19.28, 19.52, 19.68, 19.84, 20.04, 20.20, 20.40, 20.60, 20.84, 21.04, 21.40, 21.64, 21.80, 22.04, 22.20, 23.16, 23.32, 23.56, 23.80, 24.24, 24.44, 24.64, 24.84, 25.24, 25.48, 25.72, 25.92, 26.08, 26.60, 26.76, 27.04, 27.28, 27.44, 27.56, 27.72, 27.88, 28.08, 28.60, 28.76, 29.32, 29.52, 29.76, 29.96], "tokens":["六", "个", "星", "期", "嘅", "课", "程", "包", "括", "六", "堂", "课", "同", "两", "个", "测", "验", "你", "只", "掌", "握", "到", "基", "本", "嘅", "十", "九", "个", "声", "母", "五", "十", "六", "个", "韵", "母", "同", "九", "个", "声", "调", "我", "哋", "仲", "针", "对", "咗", "广", "东", "话", "学", "习", "者", "会", "遇", "到", "嘅", "大", "樽", "颈", "啊", "以", "国", "语", "为", "母", "语", "人", "士", "最", "难", "掌", "握", "嘅", "五", "大", "韵", "母", "教", "课", "书", "唔", "会", "教", "你", "嘅", "七", "种", "变", "音", "同", "十", "种", "变", "调", "说", "话", "生", "硬", "唔", "自", "然", "嘅", "根", "本", "性", "问", "题", "提", "供", "全", "新", "嘅", "学", "习", "方", "向", "等", "你", "突", "破", "难", "关"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 2.590 s
Real time factor (RTF): 2.590 / 30.592 = 0.085

yue-7.wav

Wave filename	Content	Ground truth
yue-7.wav		同意嘅累积唔系阴同阳嘅累积可以讲三既融合咗一同意融合咗阴同阳

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-7.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-7.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-7.wav
{"lang": "", "emotion": "", "event": "", "text": "同意嘅累积唔系阴同阳嘅累积可以讲三既融合咗一同二融合咗阴同阳", "timestamps": [0.64, 0.92, 1.24, 1.40, 1.64, 2.60, 2.80, 3.04, 3.44, 3.76, 4.04, 4.20, 4.40, 5.60, 5.80, 6.08, 6.96, 8.00, 8.24, 8.48, 8.80, 9.36, 9.88, 10.16, 11.28, 11.48, 11.76, 12.16, 12.64, 12.88], "tokens":["同", "意", "嘅", "累", "积", "唔", "系", "阴", "同", "阳", "嘅", "累", "积", "可", "以", "讲", "三", "既", "融", "合", "咗", "一", "同", "二", "融", "合", "咗", "阴", "同", "阳"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 1.078 s
Real time factor (RTF): 1.078 / 13.900 = 0.078

yue-8.wav

Wave filename	Content	Ground truth
yue-8.wav		而较早前已经复航嘅氹仔北安码头星期五开始增设夜间航班不过两个码头暂时都冇凌晨班次有旅客希望尽快恢复可以留喺澳门长啲时间

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-8.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-8.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-8.wav
{"lang": "", "emotion": "", "event": "", "text": "而较早前已经复航嘅氹仔北安码头星期五开始增设夜间航班不过两个码头暂时都冇凌晨班次有旅客希望尽快恢复可以留喺澳门长啲时间", "timestamps": [0.40, 0.56, 0.76, 0.92, 1.16, 1.28, 1.52, 1.68, 1.92, 2.12, 2.32, 2.52, 2.72, 2.92, 3.12, 3.48, 3.64, 3.80, 3.96, 4.16, 4.48, 4.68, 4.92, 5.08, 5.24, 5.40, 6.24, 6.40, 6.68, 6.84, 7.04, 7.20, 7.44, 7.68, 7.88, 8.04, 8.24, 8.40, 8.60, 8.80, 9.60, 9.80, 9.96, 10.12, 10.28, 10.52, 10.72, 10.88, 11.12, 11.68, 11.80, 11.96, 12.12, 12.32, 12.52, 12.76, 12.96, 13.20, 13.40], "tokens":["而", "较", "早", "前", "已", "经", "复", "航", "嘅", "氹", "仔", "北", "安", "码", "头", "星", "期", "五", "开", "始", "增", "设", "夜", "间", "航", "班", "不", "过", "两", "个", "码", "头", "暂", "时", "都", "冇", "凌", "晨", "班", "次", "有", "旅", "客", "希", "望", "尽", "快", "恢", "复", "可", "以", "留", "喺", "澳", "门", "长", "啲", "时", "间"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 1.138 s
Real time factor (RTF): 1.138 / 14.080 = 0.081

yue-9.wav

Wave filename	Content	Ground truth
yue-9.wav		刘备仲马鞭一指蜀兵一齐掩杀过去打到吴兵大败唉刘备八路兵马以雷霆万钧之势啊杀到吴兵啊尸横遍野血流成河

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-9.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-9.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-9.wav
{"lang": "", "emotion": "", "event": "", "text": "刘备仲马鞭一指蜀兵一齐掩杀过去打到吴兵大败嘿刘备八路兵马以雷霆万军之势啊杀到吴兵啊尸横遍野血流成河", "timestamps": [0.44, 0.64, 0.80, 1.00, 1.20, 1.36, 1.48, 2.44, 2.64, 2.88, 3.20, 3.44, 3.68, 3.88, 4.04, 4.36, 4.56, 4.80, 5.00, 5.28, 5.48, 6.24, 6.72, 6.96, 7.40, 7.64, 7.84, 8.08, 8.76, 9.00, 9.24, 9.48, 9.68, 9.92, 10.12, 10.28, 10.44, 10.64, 10.84, 11.04, 11.24, 11.80, 12.20, 12.48, 12.76, 13.00, 13.20, 13.40, 13.60], "tokens":["刘", "备", "仲", "马", "鞭", "一", "指", "蜀", "兵", "一", "齐", "掩", "杀", "过", "去", "打", "到", "吴", "兵", "大", "败", "嘿", "刘", "备", "八", "路", "兵", "马", "以", "雷", "霆", "万", "军", "之", "势", "啊", "杀", "到", "吴", "兵", "啊", "尸", "横", "遍", "野", "血", "流", "成", "河"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 1.116 s
Real time factor (RTF): 1.116 / 14.336 = 0.078

yue-10.wav

Wave filename	Content	Ground truth
yue-10.wav		原来王力宏咧系佢家中里面咧成就最低个吓哇

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-10.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-10.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-10.wav
{"lang": "", "emotion": "", "event": "", "text": "原来王力宏呢系佢家中里边咧成就最低个吓", "timestamps": [0.44, 0.60, 0.92, 1.28, 1.52, 1.68, 1.84, 1.96, 2.20, 2.44, 2.60, 2.76, 2.88, 3.08, 3.32, 3.60, 3.80, 4.20, 5.00], "tokens":["原", "来", "王", "力", "宏", "呢", "系", "佢", "家", "中", "里", "边", "咧", "成", "就", "最", "低", "个", "吓"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 0.481 s
Real time factor (RTF): 0.481 / 6.656 = 0.072

yue-11.wav

Wave filename	Content	Ground truth
yue-11.wav		无论你提出任何嘅要求

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-11.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-11.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-11.wav
{"lang": "", "emotion": "", "event": "", "text": "无论你提出任何嘅要求", "timestamps": [0.56, 0.68, 0.84, 1.00, 1.16, 1.36, 1.56, 1.72, 1.88, 2.08], "tokens":["无", "论", "你", "提", "出", "任", "何", "嘅", "要", "求"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 0.225 s
Real time factor (RTF): 0.225 / 2.688 = 0.084

yue-12.wav

Wave filename	Content	Ground truth
yue-12.wav		咁咁多样材料咁我哋首先第一步处理咗一件

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-12.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-12.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-12.wav
{"lang": "", "emotion": "", "event": "", "text": "咁咁多样材料咁我哋首先第一步处理咗一件", "timestamps": [0.52, 0.76, 0.96, 1.16, 1.36, 1.60, 2.00, 2.12, 2.24, 2.36, 2.60, 2.84, 3.00, 3.24, 3.68, 3.88, 4.04, 4.16, 4.28], "tokens":["咁", "咁", "多", "样", "材", "料", "咁", "我", "哋", "首", "先", "第", "一", "步", "处", "理", "咗", "一", "件"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 0.355 s
Real time factor (RTF): 0.355 / 4.864 = 0.073

yue-13.wav

Wave filename	Content	Ground truth
yue-13.wav		啲点样对于佢哋嘅服务态度啊不透过呢一年左右嘅时间啦其实大家都静一静啦咁你就会见到香港嘅经济其实

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-13.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-13.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-13.wav
{"lang": "", "emotion": "", "event": "", "text": "啲点样对于佢哋嘅服务态度啊当透过呢一年左右嘅时间啦其实大家都静一静啦咁你就会见到香港嘅经济其实", "timestamps": [0.04, 0.24, 0.44, 0.72, 0.88, 1.08, 1.28, 1.88, 2.16, 2.36, 2.60, 2.84, 3.04, 3.32, 3.52, 3.76, 4.04, 4.32, 4.60, 4.80, 5.04, 5.24, 5.36, 5.56, 5.76, 6.16, 6.32, 6.48, 6.68, 6.84, 7.08, 7.24, 7.40, 7.60, 8.08, 8.24, 8.40, 8.52, 8.68, 8.84, 9.04, 9.24, 9.40, 9.52, 9.72, 10.00, 10.20], "tokens":["啲", "点", "样", "对", "于", "佢", "哋", "嘅", "服", "务", "态", "度", "啊", "当", "透", "过", "呢", "一", "年", "左", "右", "嘅", "时", "间", "啦", "其", "实", "大", "家", "都", "静", "一", "静", "啦", "咁", "你", "就", "会", "见", "到", "香", "港", "嘅", "经", "济", "其", "实"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 0.817 s
Real time factor (RTF): 0.817 / 10.624 = 0.077

yue-14.wav

Wave filename	Content	Ground truth
yue-14.wav		就即刻会同贵正两位八代长老带埋五名七代弟子前啲灵蛇岛想话生擒谢信抢咗屠龙宝刀翻嚟献俾帮主嘅

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-14.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-14.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-14.wav
{"lang": "", "emotion": "", "event": "", "text": "就即刻会同贵正两位八代长老带埋五零七代弟子前啲灵蛇岛想话生擒谢信抢咗屠龙堡都翻嚟献俾帮主嘅", "timestamps": [0.28, 0.40, 0.52, 0.72, 0.96, 1.24, 1.52, 1.80, 1.92, 2.12, 2.32, 2.60, 2.84, 3.72, 3.88, 4.20, 4.44, 4.64, 4.84, 5.08, 5.28, 6.00, 6.12, 6.32, 6.56, 6.84, 7.80, 8.00, 8.36, 8.64, 9.00, 9.24, 10.12, 10.28, 10.52, 10.72, 10.92, 11.12, 11.28, 11.48, 11.76, 11.92, 12.16, 12.40, 12.64], "tokens":["就", "即", "刻", "会", "同", "贵", "正", "两", "位", "八", "代", "长", "老", "带", "埋", "五", "零", "七", "代", "弟", "子", "前", "啲", "灵", "蛇", "岛", "想", "话", "生", "擒", "谢", "信", "抢", "咗", "屠", "龙", "堡", "都", "翻", "嚟", "献", "俾", "帮", "主", "嘅"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 1.000 s
Real time factor (RTF): 1.000 / 13.056 = 0.077

yue-15.wav

Wave filename	Content	Ground truth
yue-15.wav		我知道我的观众大部分都是对广东话有兴趣想学广东话的人

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-15.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-15.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-15.wav
{"lang": "", "emotion": "", "event": "", "text": "我知道我嘅观众大部分都系对广东话有兴趣想学广东话嘅人", "timestamps": [0.44, 0.56, 0.72, 0.84, 1.00, 1.12, 1.36, 2.08, 2.28, 2.48, 2.68, 2.80, 2.96, 3.12, 3.32, 3.48, 3.64, 3.84, 4.04, 4.80, 5.00, 5.20, 5.40, 5.56, 5.76, 5.92], "tokens":["我", "知", "道", "我", "嘅", "观", "众", "大", "部", "分", "都", "系", "对", "广", "东", "话", "有", "兴", "趣", "想", "学", "广", "东", "话", "嘅", "人"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 0.453 s
Real time factor (RTF): 0.453 / 6.400 = 0.071

yue-16.wav

Wave filename	Content	Ground truth
yue-16.wav		诶原来啊我哋中国人呢讲究物极必反

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-16.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-16.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-16.wav
{"lang": "", "emotion": "", "event": "", "text": "啊原来啊我哋中国人呢讲究密极必反", "timestamps": [1.80, 1.92, 2.08, 2.24, 2.72, 2.84, 3.00, 3.20, 3.40, 3.56, 3.72, 3.88, 4.08, 4.28, 4.48, 4.76], "tokens":["啊", "原", "来", "啊", "我", "哋", "中", "国", "人", "呢", "讲", "究", "密", "极", "必", "反"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 0.467 s
Real time factor (RTF): 0.467 / 5.700 = 0.082

yue-17.wav

Wave filename	Content	Ground truth
yue-17.wav		如果东边道建成咁丹东呢就会成为最近嘅出海港同埋经过哈大线出海相比绥分河则会减少运渠三百五十六公里

./build/bin/sherpa-onnx-offline \
  --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt \
  --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx \
  --num-threads=1 \
  sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-17.wav

/Users/fangjun/open-source/sherpa-onnx/sherpa-onnx/csrc/parse-options.cc:Read:372 ./build/bin/sherpa-onnx-offline --tokens=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt --wenet-ctc-model=./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx --num-threads=1 sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-17.wav 

OfflineRecognizerConfig(feat_config=FeatureExtractorConfig(sampling_rate=16000, feature_dim=80, low_freq=20, high_freq=-400, dither=0, normalize_samples=True, snip_edges=False), model_config=OfflineModelConfig(transducer=OfflineTransducerModelConfig(encoder_filename="", decoder_filename="", joiner_filename=""), paraformer=OfflineParaformerModelConfig(model=""), nemo_ctc=OfflineNemoEncDecCtcModelConfig(model=""), whisper=OfflineWhisperModelConfig(encoder="", decoder="", language="", task="transcribe", tail_paddings=-1), fire_red_asr=OfflineFireRedAsrModelConfig(encoder="", decoder=""), tdnn=OfflineTdnnModelConfig(model=""), zipformer_ctc=OfflineZipformerCtcModelConfig(model=""), wenet_ctc=OfflineWenetCtcModelConfig(model="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/model.int8.onnx"), sense_voice=OfflineSenseVoiceModelConfig(model="", language="auto", use_itn=False), moonshine=OfflineMoonshineModelConfig(preprocessor="", encoder="", uncached_decoder="", cached_decoder=""), dolphin=OfflineDolphinModelConfig(model=""), canary=OfflineCanaryModelConfig(encoder="", decoder="", src_lang="", tgt_lang="", use_pnc=True), telespeech_ctc="", tokens="./sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/tokens.txt", num_threads=1, debug=False, provider="cpu", model_type="", modeling_unit="cjkchar", bpe_vocab=""), lm_config=OfflineLMConfig(model="", scale=0.5, lodr_scale=0.01, lodr_fst="", lodr_backoff_id=-1), ctc_fst_decoder_config=OfflineCtcFstDecoderConfig(graph="", max_active=3000), decoding_method="greedy_search", max_active_paths=4, hotwords_file="", hotwords_score=1.5, blank_penalty=0, rule_fsts="", rule_fars="", hr=HomophoneReplacerConfig(dict_dir="", lexicon="", rule_fsts=""))
Creating recognizer ...
Started
Done!

sherpa-onnx-wenetspeech-yue-u2pp-conformer-ctc-zh-en-cantonese-int8-2025-09-10/test_wavs/yue-17.wav
{"lang": "", "emotion": "", "event": "", "text": "如果东边道建城咁丹东呢就会成为最近嘅出海港同埋经过哈大线出海相比绥分河将会减少运距三百五十六公里", "timestamps": [0.52, 0.64, 0.84, 1.04, 1.28, 1.52, 1.80, 2.72, 2.96, 3.20, 3.44, 3.96, 4.08, 4.24, 4.40, 4.60, 4.76, 4.92, 5.08, 5.20, 5.44, 6.48, 6.60, 6.80, 6.96, 7.12, 7.36, 7.56, 7.76, 7.96, 8.16, 8.36, 9.40, 9.60, 9.88, 10.40, 10.52, 10.68, 10.92, 11.16, 11.36, 11.92, 12.16, 12.32, 12.48, 12.64, 12.80, 13.00], "tokens":["如", "果", "东", "边", "道", "建", "城", "咁", "丹", "东", "呢", "就", "会", "成", "为", "最", "近", "嘅", "出", "海", "港", "同", "埋", "经", "过", "哈", "大", "线", "出", "海", "相", "比", "绥", "分", "河", "将", "会", "减", "少", "运", "距", "三", "百", "五", "十", "六", "公", "里"], "words": []}
----
num threads: 1
decoding method: greedy_search
Elapsed seconds: 1.039 s
Real time factor (RTF): 1.039 / 13.800 = 0.075