EngineX-Iluvatar/enginex_bi_series-sherpa-onnx
Archived
15
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity
This repository has been archived on 2025-08-26. You can view files and clone it, but cannot push or open issues or pull requests.
Files
3bf986d08d245e1ec482f0c203b85f3d6501c0ed
enginex_bi_series-sherpa-onnx/scripts/apk/generate-vad-asr-apk-script.py
Fangjun Kuang 3bf986d08d Support non-streaming zipformer CTC ASR models (#2340) 
This PR adds support for non-streaming Zipformer CTC ASR models across 
multiple language bindings, WebAssembly, examples, and CI workflows.

- Introduces a new OfflineZipformerCtcModelConfig in C/C++, Python, Swift, Java, Kotlin, Go, Dart, Pascal, and C# APIs
- Updates initialization, freeing, and recognition logic to include Zipformer CTC in WASM and Node.js
- Adds example scripts and CI steps for downloading, building, and running Zipformer CTC models

Model doc is available at
https://k2-fsa.github.io/sherpa/onnx/pretrained_models/offline-ctc/icefall/zipformer.html
2025-07-04 15:57:07 +08:00

625 lines
14 KiB
Python
Executable File
Raw Blame History

#!/usr/bin/env python3
import argparse
from dataclasses import dataclass
from pathlib import Path
import jinja2
def get_args():
parser = argparse.ArgumentParser()
parser.add_argument(
"--total",
type=int,
default=1,
help="Number of runners",
)
parser.add_argument(
"--index",
type=int,
default=0,
help="Index of the current runner",
)
return parser.parse_args()
@dataclass
class Model:
# We will download
# https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/{model_name}.tar.bz2
model_name: str
# The type of the model, e..g, 0, 1, 2. It is hardcoded in the kotlin code
idx: int
# e.g., zh, en, zh_en
lang: str
lang2: str
# e.g., whisper, paraformer, zipformer
short_name: str = ""
# cmd is used to remove extra file from the model directory
cmd: str = ""
rule_fsts: str = ""
use_hr: bool = False
# See get_2nd_models() in ./generate-asr-2pass-apk-script.py
def get_models():
models = [
Model(
model_name="sherpa-onnx-whisper-tiny.en",
idx=2,
lang="en",
lang2="English",
short_name="whisper_tiny",
cmd="""
pushd $model_name
rm -fv tiny.en-encoder.onnx
rm -fv tiny.en-decoder.onnx
rm -rf test_wavs
rm -fv *.py
rm -fv requirements.txt
rm -fv .gitignore
rm -fv README.md
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-paraformer-zh-2023-09-14",
idx=0,
lang="zh_en",
lang2="Chinese,English",
short_name="paraformer",
rule_fsts="itn_zh_number.fst",
cmd="""
if [ ! -f itn_zh_number.fst ]; then
curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/itn_zh_number.fst
fi
pushd $model_name
rm -fv README.md
rm -rfv test_wavs
rm -fv model.onnx
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17",
idx=15,
lang="zh_en_ko_ja_yue",
lang2="中英粤日韩",
short_name="sense_voice",
use_hr=True,
cmd="""
pushd $model_name
rm -rfv test_wavs
rm -fv model.onnx
rm -fv *.py
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-paraformer-zh-small-2024-03-09",
idx=14,
lang="zh_en",
lang2="Chinese,English",
short_name="small_paraformer",
rule_fsts="itn_zh_number.fst",
cmd="""
if [ ! -f itn_zh_number.fst ]; then
curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/itn_zh_number.fst
fi
pushd $model_name
rm -fv README.md
rm -fv *.py
rm -fv *.yaml
rm -fv *.mvn
rm -rfv test_wavs
ls -lh
popd
""",
),
Model(
model_name="icefall-asr-zipformer-wenetspeech-20230615",
idx=4,
lang="zh",
lang2="Chinese",
short_name="zipformer",
rule_fsts="itn_zh_number.fst",
cmd="""
if [ ! -f itn_zh_number.fst ]; then
curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/itn_zh_number.fst
fi
pushd $model_name
rm -rfv test_wavs
rm -fv README.md
mv -v data/lang_char/tokens.txt ./
rm -rfv data/lang_char
mv -v exp/encoder-epoch-12-avg-4.int8.onnx ./
mv -v exp/decoder-epoch-12-avg-4.onnx ./
mv -v exp/joiner-epoch-12-avg-4.int8.onnx ./
rm -rfv exp
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-nemo-fast-conformer-ctc-be-de-en-es-fr-hr-it-pl-ru-uk-20k",
idx=7,
lang="be_de_en_es_fr_hr_it_pl_ru_uk",
lang2="be_de_en_es_fr_hr_it_pl_ru_uk",
short_name="fast_conformer_ctc_20k",
cmd="""
pushd $model_name
rm -rfv test_wavs
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-nemo-fast-conformer-ctc-en-24500",
idx=8,
lang="en",
lang2="English",
short_name="fast_conformer_ctc_24500",
cmd="""
pushd $model_name
rm -rfv test_wavs
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-nemo-fast-conformer-ctc-en-de-es-fr-14288",
idx=9,
lang="en_de_es_fr",
lang2="English,German,Spanish,French",
short_name="fast_conformer_ctc_14288",
cmd="""
pushd $model_name
rm -rfv test_wavs
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-nemo-fast-conformer-ctc-es-1424",
idx=10,
lang="es",
lang2="Spanish",
short_name="fast_conformer_ctc_1424",
cmd="""
pushd $model_name
rm -rfv test_wavs
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-telespeech-ctc-int8-zh-2024-06-04",
idx=11,
lang="zh",
lang2="Chinese",
short_name="telespeech",
rule_fsts="itn_zh_number.fst",
cmd="""
if [ ! -f itn_zh_number.fst ]; then
curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/itn_zh_number.fst
fi
pushd $model_name
rm -rfv test_wavs
rm -fv test.py
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-zipformer-thai-2024-06-20",
idx=12,
lang="th",
lang2="Thai",
short_name="zipformer",
cmd="""
pushd $model_name
rm -rfv test_wavs
rm -fv README.md
rm -fv bpe.model
rm -fv encoder-epoch-12-avg-5.onnx
rm -fv decoder-epoch-12-avg-5.int8.onnx
rm joiner-epoch-12-avg-5.onnx
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-zipformer-korean-2024-06-24",
idx=13,
lang="ko",
lang2="Korean",
short_name="zipformer",
cmd="""
pushd $model_name
rm -rfv test_wavs
rm -fv README.md
rm -fv bpe.model
rm -fv encoder-epoch-99-avg-1.onnx
rm -fv decoder-epoch-99-avg-1.int8.onnx
rm -fv joiner-epoch-99-avg-1.onnx
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-zipformer-ja-reazonspeech-2024-08-01",
idx=16,
lang="ja",
lang2="Japanese",
short_name="zipformer_reazonspeech",
cmd="""
pushd $model_name
rm -rfv test_wavs
rm -fv encoder-epoch-99-avg-1.onnx
rm -fv decoder-epoch-99-avg-1.int8.onnx
rm -fv joiner-epoch-99-avg-1.onnx
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-zipformer-ru-2024-09-18",
idx=17,
lang="ru",
lang2="Russian",
short_name="zipformer",
cmd="""
pushd $model_name
rm -rfv test_wavs
rm -fv encoder.onnx
rm -fv decoder.int8.onnx
rm -fv joiner.onnx
rm -fv bpe.model
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-small-zipformer-ru-2024-09-18",
idx=18,
lang="ru",
lang2="Russian",
short_name="small_zipformer",
cmd="""
pushd $model_name
rm -rfv test_wavs
rm -fv encoder.onnx
rm -fv decoder.int8.onnx
rm -fv joiner.onnx
rm -fv bpe.model
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-nemo-ctc-giga-am-russian-2024-10-24",
idx=19,
lang="ru",
lang2="Russian",
short_name="nemo_ctc_giga_am",
cmd="""
pushd $model_name
rm -rfv test_wavs
rm -fv *.sh
rm -fv *.py
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-nemo-transducer-giga-am-russian-2024-10-24",
idx=20,
lang="ru",
lang2="Russian",
short_name="nemo_transducer_giga_am",
cmd="""
pushd $model_name
rm -rfv test_wavs
rm -fv *.sh
rm -fv *.py
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-moonshine-tiny-en-int8",
idx=21,
lang="en",
lang2="English",
short_name="moonshine_tiny_int8",
cmd="""
pushd $model_name
rm -rfv test_wavs
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-moonshine-base-en-int8",
idx=22,
lang="en",
lang2="English",
short_name="moonshine_base_int8",
cmd="""
pushd $model_name
rm -rfv test_wavs
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-zipformer-zh-en-2023-11-22",
idx=23,
lang="zh_en",
lang2="Chinese,English",
short_name="zipformer",
cmd="""
pushd $model_name
rm -rfv test_wavs
rm -fv encoder-epoch-34-avg-19.onnx
rm -fv joiner-epoch-34-avg-19.onnx
rm -fv bbpe.model
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-dolphin-base-ctc-multi-lang-int8-2025-04-02",
idx=25,
lang="multi_lang",
lang2="multi_lang",
short_name="dolphin_base_ctc",
cmd="""
pushd $model_name
rm -rfv test_wavs
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-zipformer-vi-int8-2025-04-20",
idx=26,
lang="vi",
lang2="Vietnamese",
short_name="zipformer",
cmd="""
pushd $model_name
rm -rfv test_wavs
rm -fv bpe.model
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-nemo-ctc-giga-am-v2-russian-2025-04-19",
idx=27,
lang="ru",
lang2="Russian",
short_name="nemo_ctc_giga_am_v2",
cmd="""
pushd $model_name
rm -rfv test_wavs
rm -fv *.sh
rm -fv *.py
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-nemo-transducer-giga-am-v2-russian-2025-04-19",
idx=28,
lang="ru",
lang2="Russian",
short_name="nemo_transducer_giga_am",
cmd="""
pushd $model_name
rm -rfv test_wavs
rm -fv *.sh
rm -fv *.py
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-zipformer-ru-int8-2025-04-20",
idx=29,
lang="ru",
lang2="Russian",
short_name="v2_zipformer",
cmd="""
pushd $model_name
rm -rfv test_wavs
rm -fv bpe.model
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-nemo-parakeet-tdt-0.6b-v2-int8",
idx=30,
lang="en",
lang2="English",
short_name="parakeet_tdt_0.6b_v2",
cmd="""
pushd $model_name
rm -rfv test_wavs
ls -lh
popd
""",
),
Model(
model_name="sherpa-onnx-zipformer-ctc-zh-int8-2025-07-03",
idx=31,
lang="zh",
lang2="Chinese",
short_name="zipformer_2025_07_03",
cmd="""
pushd $model_name
rm -rfv test_wavs
rm -rfv bbpe.model
ls -lh
popd
""",
),
]
return models
def main():
args = get_args()
index = args.index
total = args.total
assert 0 <= index < total, (index, total)
all_model_list = get_models()
num_models = len(all_model_list)
num_per_runner = num_models // total
if num_per_runner <= 0:
raise ValueError(f"num_models: {num_models}, num_runners: {total}")
start = index * num_per_runner
end = start + num_per_runner
remaining = num_models - args.total * num_per_runner
print(f"{index}/{total}: {start}-{end}/{num_models}")
d = dict()
d["model_list"] = all_model_list[start:end]
if index < remaining:
s = args.total * num_per_runner + index
d["model_list"].append(all_model_list[s])
print(f"{s}/{num_models}")
filename_list = [
"./build-apk-vad-asr.sh",
"./build-hap-vad-asr.sh",
"./build-apk-vad-asr-simulate-streaming.sh",
]
for filename in filename_list:
environment = jinja2.Environment()
if not Path(f"{filename}.in").is_file():
print(f"skip {filename}")
continue
with open(f"{filename}.in") as f:
s = f.read()
template = environment.from_string(s)
s = template.render(**d)
with open(filename, "w") as f:
print(s, file=f)
if __name__ == "__main__":
main()
Reference in New Issue View Git Blame Copy Permalink