|
|
|
import os |
|
import re |
|
from pathlib import Path |
|
from typing import List |
|
|
|
BASE_URL = "https://huggingface.co/csukuangfj/sherpa-onnx-apk/resolve/main/" |
|
|
|
from dataclasses import dataclass |
|
|
|
|
|
@dataclass |
|
class APK: |
|
major: int |
|
minor: int |
|
patch: int |
|
arch: str |
|
|
|
def __init__(self, s): |
|
|
|
s = str(s).split("/")[-1] |
|
split = s.split("-") |
|
self.major, self.minor, self.patch = list(map(int, split[2].split("."))) |
|
self.arch = split[3] |
|
if "arm" in s: |
|
self.arch += "-" + split[4] |
|
|
|
if "armeabi" in self.arch: |
|
self.arch = "y" + self.arch |
|
|
|
if "arm64" in self.arch: |
|
self.arch = "z" + self.arch |
|
|
|
|
|
def sort_by_apk(x): |
|
x = APK(x) |
|
return (x.major, x.minor, x.patch, x.arch) |
|
|
|
|
|
def get_all_files(d_list: List[str], suffix: str) -> List[str]: |
|
if isinstance(d_list, str): |
|
d_list = [d_list] |
|
|
|
min_major = 1 |
|
min_minor = 9 |
|
min_patch = 10 |
|
|
|
ss = [] |
|
for d in d_list: |
|
for root, _, files in os.walk(d): |
|
for f in files: |
|
if f.endswith(suffix): |
|
major, minor, patch = list(map(int, f.split("-")[2].split("."))) |
|
if major >= min_major and minor >= min_minor and patch >= min_patch: |
|
ss.append(os.path.join(root, f)) |
|
|
|
ans = sorted(ss, key=sort_by_apk, reverse=True) |
|
|
|
return list(map(lambda x: BASE_URL + str(x), ans)) |
|
|
|
|
|
def to_file(filename: str, files: List[str]): |
|
content = r""" |
|
<h1> APKs for Speaker diarization </h1> |
|
This page lists the <strong>speaker diarization</strong> APKs for <a href="http://github.com/k2-fsa/sherpa-onnx">sherpa-onnx</a>, |
|
one of the deployment frameworks of <a href="https://github.com/k2-fsa">the Next-gen Kaldi project</a>. |
|
<br/> |
|
The name of an APK has the following rule: |
|
<ul> |
|
<li> sherpa-onnx-{version}-{arch}-speaker-diarization-{segmentation_model}-{embedding_model}.apk |
|
</ul> |
|
where |
|
<ul> |
|
<li> version: It specifies the current version, e.g., 1.10.27 |
|
<li> arch: The architecture targeted by this APK, e.g., arm64-v8a, armeabi-v7a, x86_64, x86 |
|
<li> segmentation_model: The framework for the speaker segmentation model |
|
<li> embedding_model: The framework for the speaker embedding extraction model |
|
</ul> |
|
|
|
<br/> |
|
|
|
<span style="color:red;">Note:</span> Please see |
|
<a href="https://github.com/k2-fsa/sherpa-onnx/releases/tag/speaker-segmentation-models">https://github.com/k2-fsa/sherpa-onnx/releases/tag/speaker-segmentation-models</a> for a list of supported speaker segmentation models. |
|
<br/> |
|
|
|
<span style="color:red;">Note:</span> Please see |
|
<a href="https://github.com/k2-fsa/sherpa-onnx/releases/tag/speaker-recongition-models">https://github.com/k2-fsa/sherpa-onnx/releases/tag/speaker-recongition-models</a> for a list of supported speaker embedding extraction models. |
|
|
|
<br/> |
|
<br/> |
|
|
|
<strong>Note about the license</strong> The code of Next-gen Kaldi is using |
|
<a href="https://www.apache.org/licenses/LICENSE-2.0">Apache-2.0 license</a>. However, |
|
we support models from different frameworks. Please check the license of your selected model. |
|
|
|
<br/><br/> |
|
|
|
<strong>Note about the build script</strong> You can find the script for building the APKs |
|
at <a href="https://github.com/k2-fsa/sherpa-onnx/blob/master/scripts/apk/build-apk-speaker-diarization.sh">https://github.com/k2-fsa/sherpa-onnx/blob/master/scripts/apk/build-apk-speaker-diarization.sh</a> |
|
|
|
<br/><br/> |
|
|
|
<style type="text/css"> |
|
.tg {border-collapse:collapse;border-spacing:0;} |
|
.tg td{border-color:black;border-style:solid;border-width:1px;font-family:Arial, sans-serif;font-size:14px; |
|
overflow:hidden;padding:10px 5px;word-break:normal;} |
|
.tg th{border-color:black;border-style:solid;border-width:1px;font-family:Arial, sans-serif;font-size:14px; |
|
font-weight:normal;overflow:hidden;padding:10px 5px;word-break:normal;} |
|
.tg .tg-0pky{border-color:inherit;text-align:left;vertical-align:top} |
|
.tg .tg-0lax{text-align:left;vertical-align:top} |
|
</style> |
|
<table class="tg"> |
|
<thead> |
|
<tr> |
|
<th class="tg-0pky">APK</th> |
|
<th class="tg-0lax">Speaker segmentation model</th> |
|
<th class="tg-0pky">Speaker embedding extraction model</th> |
|
</tr> |
|
</thead> |
|
<tbody> |
|
<tr> |
|
<td class="tg-0pky">sherpa-onnx-x.y.z-arm64-v8a-speaker-diarization-pyannote_audio-3dspeaker.apk</td> |
|
<td class="tg-0pky"><a href="https://github.com/k2-fsa/sherpa-onnx/releases/download/speaker-segmentation-models/sherpa-onnx-pyannote-segmentation-3-0.tar.bz2">sherpa-onnx-pyannote-segmentation-3-0.tar.bz2</a> It is <a href="https://github.com/k2-fsa/sherpa-onnx/tree/master/scripts/pyannote/segmentation">converted</a> from <a href="https://huggingface.co/pyannote/segmentation-3.0">https://huggingface.co/pyannote/segmentation-3.0</a></td> |
|
<td class="tg-0pky"><a href="https://github.com/k2-fsa/sherpa-onnx/releases/download/speaker-recongition-models/3dspeaker_speech_eres2net_base_sv_zh-cn_3dspeaker_16k.onnx">3dspeaker_speech_eres2net_base_sv_zh-cn_3dspeaker_16k.onnx</a> It is <a href="https://github.com/k2-fsa/sherpa-onnx/tree/master/scripts/3dspeaker">converted</a> from <a href="https://github.com/alibaba-damo-academy/3D-Speaker">https://github.com/alibaba-damo-academy/3D-Speaker</a></td> |
|
<tr> |
|
</tbody> |
|
</table> |
|
|
|
<br/> |
|
<br/> |
|
<div/> |
|
""" |
|
|
|
if "-cn" not in filename: |
|
content += """ |
|
For Chinese users, please <a href="./apk-cn.html">visit this address</a>, |
|
which replaces <a href="huggingface.co">huggingface.co</a> with <a href="hf-mirror.com">hf-mirror.com</a> |
|
<br/> |
|
<br/> |
|
中国用户, 请访问<a href="./apk-cn.html">这个地址</a> |
|
<br/> |
|
<br/> |
|
""" |
|
|
|
with open(filename, "w") as f: |
|
print(content, file=f) |
|
for x in files: |
|
name = x.rsplit("/", maxsplit=1)[-1] |
|
print(f'<a href="{x}" />{name}<br/>', file=f) |
|
|
|
|
|
def main(): |
|
apk = get_all_files("speaker-diarization", suffix=".apk") |
|
to_file("./apk-speaker-diarization.html", apk) |
|
|
|
|
|
apk2 = [] |
|
for a in apk: |
|
a = a.replace("huggingface.co", "hf-mirror.com") |
|
a = a.replace("resolve", "blob") |
|
apk2.append(a) |
|
|
|
to_file("./apk-speaker-diarization-cn.html", apk2) |
|
|
|
|
|
if __name__ == "__main__": |
|
main() |
|
|
