├── ClipVideo
    ├── README.md
    ├── clipvideo
    │   ├── Dataset_generator.py
    │   ├── __init__.py
    │   ├── __pycache__
    │   │   ├── __init__.cpython-310.pyc
    │   │   ├── argparse_tools.cpython-310.pyc
    │   │   ├── subtitle_utils.cpython-310.pyc
    │   │   ├── trans_utils.cpython-310.pyc
    │   │   └── videoclipper.cpython-310.pyc
    │   ├── argparse_tools.py
    │   ├── gradio_service.py
    │   ├── imagemagick_test.py
    │   ├── subtitle_utils.py
    │   ├── test.sh
    │   ├── trans_utils.py
    │   └── videoclipper.py
    ├── docs
    │   └── images
    │   │   └── show.png
    └── requirments.txt
├── FunASR-APP-LICENSE
├── FunASR-APP-README.md
├── LICENSE
├── README.md
└── TransAudio
    ├── README.md
    └── funasr-runtime-deploy.sh


/ClipVideo/README.md:
--------------------------------------------------------------------------------
 1 | ## ClipVideo
 2 | 
 3 | As the first application toolkit based of FunASR-APP, <strong>ClipVideo</strong> enables users to clip ```.mp4``` video files or ```.wav``` audio files with chosen text segments out of the recognition results generated by [Paraformer-long model](https://modelscope.cn/models/damo/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch/summary).
 4 | 
 5 | Under the help of ClipVideo you can get the video clips easily with the following steps (in Gradio service):
 6 | - Step1: Chose your video file (or try the example videos below)
 7 | - Step2: Copy the text segments you need to 'Text to Clip'
 8 | - Step3: Adjust subtitle settings (if needed)
 9 | - Step4: Click 'Clip' or 'Clip and Generate Subtitles'
10 | 
11 | ### Usage
12 | ```shell
13 | git clone https://github.com/alibaba-damo-academy/FunASR-APP.git
14 | cd FunASR-APP
15 | # install modelscope
16 | pip install "modelscope[audio_asr]" -f https://modelscope.oss-cn-beijing.aliyuncs.com/releases/repo.html
17 | # python environments
18 | pip install -r ClipVideo/requirments.txt
19 | ```
20 | (Optional) If you want to clip video file with embedded subtitles
21 | 
22 | 1. ffmpeg and imagemagick is required
23 | 
24 | - On Ubuntu
25 | ```shell
26 | apt-get -y update && apt-get -y install ffmpeg imagemagick
27 | sed -i 's/none/read,write/g' /etc/ImageMagick-6/policy.xml
28 | ```
29 | - On MacOS
30 | ```shell
31 | brew install imagemagick
32 | sed -i 's/none/read,write/g' /usr/local/Cellar/imagemagick/7.1.1-8_1/etc/ImageMagick-7/policy.xml 
33 | ```
34 | 2. Download font file to ClipVideo/font
35 | 
36 | ```shell
37 | wget https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ClipVideo/STHeitiMedium.ttc -O ClipVideo/font/STHeitiMedium.ttc
38 | ```
39 | 
40 | #### Experience ClipVideo in Modelscope
41 | You can try ClipVideo in modelscope space: [link](https://modelscope.cn/studios/damo/funasr_app_clipvideo/summary).
42 | 
43 | #### Use ClipVideo by Gradio Service
44 | You can establish your own ClipVideo service which is same as [Modelscope Space](https://modelscope.cn/studios/damo/funasr_app_clipvideo/summary) as follow:
45 | ```shell
46 | python clipvideo/gradio_service.py
47 | ```
48 | then visit ```localhost:7860``` you will get a Gradio service like below and you can use ClipVideo following the steps:
49 | <img src="./docs/images/show.png"/>
50 | 
51 | #### Use ClipVideo in command line
52 | ClipVideo supports you to recognize and clip with commands:
53 | ```shell
54 | # working in ClipVideo/
55 | # step1: Recognize
56 | python clipvideo/videoclipper.py --stage 1 \
57 |                        --file examples/2022云栖大会_片段.mp4 \
58 |                        --output_dir ./output
59 | # now you can find recognition results and entire SRT file in ./output/
60 | # step2: Clip
61 | python clipvideo/videoclipper.py --stage 2 \
62 |                        --file examples/2022云栖大会_片段.mp4 \
63 |                        --output_dir ./output \
64 |                        --dest_text '我们把它跟乡村振兴去结合起来，利用我们的设计的能力' \
65 |                        --start_ost 0 \
66 |                        --end_ost 100 \
67 |                        --output_file './output/res.mp4'
68 | ```
69 | 


--------------------------------------------------------------------------------
/ClipVideo/clipvideo/Dataset_generator.py:
--------------------------------------------------------------------------------
  1 | import librosa
  2 | import soundfile as sf
  3 | from videoclipper import VideoClipper
  4 | import os
  5 | from tqdm import tqdm
  6 | import subprocess
  7 | from modelscope.pipelines import pipeline
  8 | from modelscope.utils.constant import Tasks
  9 | from shutil import copyfile, rmtree
 10 | 
 11 | current_directory = os.path.dirname(os.path.abspath(__file__))
 12 | inference_pipeline = pipeline(
 13 |     task=Tasks.auto_speech_recognition,
 14 |     model='damo/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch',
 15 |     vad_model='damo/speech_fsmn_vad_zh-cn-16k-common-pytorch',
 16 |     punc_model='damo/punc_ct-transformer_zh-cn-common-vocab272727-pytorch',
 17 | )
 18 | video_tools = VideoClipper(inference_pipeline)
 19 | audio_clipper = VideoClipper(None)
 20 | STAGE_RECOGNIZE = 1
 21 | STAGE_CLIP = 2
 22 | 
 23 | # 开始偏移量
 24 | START_OST = 0
 25 | # 结束偏移量
 26 | END_OST = 0
 27 | 
 28 | FONT_SIZE = 32
 29 | FONT_COLOR = 'white'
 30 | ADD_SUB = False
 31 | OUTPUT_SRT_FILE = f"{current_directory}/output/srt/"
 32 | OUTPUT_MP4_FILE = f"{current_directory}/output/mp4/"
 33 | OUTPUT_WAV_FILE = f"{current_directory}/output/wav/"
 34 | INPUT_FILE_PATH = f"{current_directory}/video_files/"
 35 | READY_INPUT_WAV_PATH = f"{current_directory}/input/mdx_extra/"
 36 | DEMUCS_TARGET_INPUT_PATH = f"{current_directory}/input/"
 37 | ORG_INPUT_WAV_PATH = f"{current_directory}/org_wav/"
 38 | 
 39 | 
 40 | # 获取所有文件
 41 | def get_all_files_in_directory(directory):
 42 |     print(f"from :{current_directory}{directory} ")
 43 |     file_paths_list = []
 44 |     for root, dirs, files in os.walk(directory):
 45 |         for file in files:
 46 |             file_path = os.path.join(root, file)
 47 |             file_paths_list.append(file_path)
 48 |     return file_paths_list
 49 | 
 50 | 
 51 | # 获取所有的字幕和字幕文件路径
 52 | def vidio_recognizing_to_get_srt_list():
 53 |     video_list = get_all_files_in_directory(READY_INPUT_WAV_PATH)
 54 |     srt_list = []
 55 |     srt_file_list = []
 56 |     state_json = {}
 57 |     for video_path in tqdm(video_list):
 58 |         try:
 59 |             wav = librosa.load(video_path, sr=16000)[0]
 60 |             res_text, res_srt, state = video_tools.recog((16000, wav))
 61 |             state_json.update({str(video_path): state})
 62 |         except Exception as e:
 63 |             print(f"音频转译错误 {video_path}: {e}")
 64 |             continue
 65 |         srt_file_name = os.path.splitext(os.path.basename(video_path))[0] + '.srt'
 66 |         srt_save_path = os.path.join(OUTPUT_SRT_FILE, srt_file_name)
 67 |         srt_list.append(res_srt)
 68 |         try:
 69 |             with open(srt_save_path, "w", encoding="utf-8") as f:
 70 |                 f.write(res_srt)
 71 |                 srt_file_list.append(srt_save_path)
 72 |                 print(str(srt_save_path) + "写入")
 73 |         except Exception as e:
 74 |             print(f"Error writing SRT file {srt_save_path}: {e}")
 75 |     return srt_list, srt_file_list, state_json
 76 | 
 77 | 
 78 | # 视频转音频
 79 | def mp4_to_wav(input_path):
 80 |     output_file_name = os.path.splitext(os.path.basename(input_path))[0] + '.wav'
 81 |     output_path = os.path.join(ORG_INPUT_WAV_PATH, output_file_name)
 82 |     subprocess.run([
 83 |         "ffmpeg", "-i", input_path, "-acodec", "pcm_s16le", "-ar", "16000", output_path
 84 |     ])
 85 |     print(f"转换完成: {input_path} -> {output_path}")
 86 | 
 87 | 
 88 | # 获取所有文字大列表
 89 | def extract_subtitle_text_list_from_srt():
 90 |     srt_list, srt_file_path_list, state_json = vidio_recognizing_to_get_srt_list()
 91 |     all_srt_text_json = {}
 92 |     index = 0
 93 |     for srt_file_path in srt_file_path_list:
 94 |         single_subtitle_text_list = []
 95 |         with open(srt_file_path, "r", encoding="utf-8") as srt_file:
 96 |             lines = srt_file.readlines()
 97 |             for i in range(2, len(lines), 3):
 98 |                 subtitle_text = lines[i].strip()
 99 |                 single_subtitle_text_list.append(subtitle_text)
100 |                 index = index + 1
101 |         wav_file_path = READY_INPUT_WAV_PATH + os.path.splitext(os.path.basename(srt_file_path))[0] + '.wav'
102 |         all_srt_text_json.update({str(wav_file_path): single_subtitle_text_list})
103 |     print(f"总文字数量：{index}")
104 |     return all_srt_text_json, state_json
105 | 
106 | 
107 | # 通过文字去剪裁音频并保存
108 | def clip_audio_from_srt():
109 |     all_srt_text_json, state_json = extract_subtitle_text_list_from_srt()
110 |     print(str(state_json))
111 |     if len(all_srt_text_json) == 0 or len(state_json) == 0:
112 |         print("字幕文件为空，请检查音频文件内是否存在语音。")
113 |         return None
114 |     for wav_file_path in all_srt_text_json:
115 |         for one_line_text in all_srt_text_json[wav_file_path]:
116 |             state = state_json[str(wav_file_path)]
117 |             wav_file_name = OUTPUT_WAV_FILE + str(one_line_text) + '_clip.wav'
118 |             try:
119 |                 (sr, audio), message, srt_clip = video_tools.clip(dest_text=str(one_line_text), start_ost=START_OST,
120 |                                                                 end_ost=END_OST,
121 |                                                                 state=state)
122 |                 if "No period found in the speech" not in message:
123 |                     print(f"{one_line_text} in {wav_file_path} 完成")
124 |                     sf.write(wav_file_name, audio, 16000)
125 |             except Exception as e:
126 |                 print(f"{one_line_text} in {wav_file_path}切割失败，原因：{e}")
127 | 
128 | 
129 | # 清空生成物，初始化文件夹
130 | def clean_files():
131 |     rmtree(f"{current_directory}/input/mdx_extra/")
132 |     os.mkdir(f"{current_directory}/input/mdx_extra/")
133 |     rmtree(f"{current_directory}/org_wav/")
134 |     os.mkdir(f"{current_directory}/org_wav/")
135 |     rmtree(f"{current_directory}/output/mp4/")
136 |     os.mkdir(f"{current_directory}/output/mp4/")
137 |     rmtree(f"{current_directory}/output/srt/")
138 |     os.mkdir(f"{current_directory}/output/srt/")
139 |     rmtree(f"{current_directory}/output/wav/")
140 |     os.mkdir(f"{current_directory}/output/wav/")
141 |     print("初始化完成")
142 | 
143 | 
144 | # 降噪
145 | def demucs_wav():
146 |     for org_wav_file in get_all_files_in_directory(ORG_INPUT_WAV_PATH):
147 |         target_ready_wav_path = os.path.splitext(os.path.basename(org_wav_file))[0] + ".wav"
148 |         os.system(
149 |             f"demucs -n mdx_extra -o {DEMUCS_TARGET_INPUT_PATH} --filename {target_ready_wav_path} {org_wav_file}")
150 | 
151 | 
152 | # 预备……
153 | def ready_all_to_wav():
154 |     files_list = get_all_files_in_directory(INPUT_FILE_PATH)
155 |     for file in files_list:
156 |         audio_suffixs = ['wav']
157 |         video_suffixs = ['mp4']
158 |         if file[-3:] in audio_suffixs:
159 |             target_wav_path = os.path.join(ORG_INPUT_WAV_PATH, os.path.basename(file))
160 |             copyfile(file, target_wav_path)
161 |         elif file[-3:] in video_suffixs:
162 |             mp4_to_wav(file)
163 |         else:
164 |             print(f"只支持.wav和.mp4！Only supports .wav and .mp4")
165 | 
166 | 
167 | # 润！
168 | def run():
169 |     clean_files()
170 |     ready_all_to_wav()
171 |     demucs_wav()
172 |     clip_audio_from_srt()
173 | 
174 | 
175 | # 面函数
176 | if __name__ == '__main__':
177 |     run()
178 | 


--------------------------------------------------------------------------------
/ClipVideo/clipvideo/__init__.py:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/Fatfish588/Dataset_Generator_For_VITS/2f012ac453a99723c29ffbf119b4ff545492ebb7/ClipVideo/clipvideo/__init__.py


--------------------------------------------------------------------------------
/ClipVideo/clipvideo/__pycache__/__init__.cpython-310.pyc:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/Fatfish588/Dataset_Generator_For_VITS/2f012ac453a99723c29ffbf119b4ff545492ebb7/ClipVideo/clipvideo/__pycache__/__init__.cpython-310.pyc


--------------------------------------------------------------------------------
/ClipVideo/clipvideo/__pycache__/argparse_tools.cpython-310.pyc:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/Fatfish588/Dataset_Generator_For_VITS/2f012ac453a99723c29ffbf119b4ff545492ebb7/ClipVideo/clipvideo/__pycache__/argparse_tools.cpython-310.pyc


--------------------------------------------------------------------------------
/ClipVideo/clipvideo/__pycache__/subtitle_utils.cpython-310.pyc:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/Fatfish588/Dataset_Generator_For_VITS/2f012ac453a99723c29ffbf119b4ff545492ebb7/ClipVideo/clipvideo/__pycache__/subtitle_utils.cpython-310.pyc


--------------------------------------------------------------------------------
/ClipVideo/clipvideo/__pycache__/trans_utils.cpython-310.pyc:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/Fatfish588/Dataset_Generator_For_VITS/2f012ac453a99723c29ffbf119b4ff545492ebb7/ClipVideo/clipvideo/__pycache__/trans_utils.cpython-310.pyc


--------------------------------------------------------------------------------
/ClipVideo/clipvideo/__pycache__/videoclipper.cpython-310.pyc:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/Fatfish588/Dataset_Generator_For_VITS/2f012ac453a99723c29ffbf119b4ff545492ebb7/ClipVideo/clipvideo/__pycache__/videoclipper.cpython-310.pyc


--------------------------------------------------------------------------------
/ClipVideo/clipvideo/argparse_tools.py:
--------------------------------------------------------------------------------
 1 | import argparse
 2 | from pathlib import Path
 3 | 
 4 | import yaml
 5 | import sys
 6 | 
 7 | 
 8 | class ArgumentParser(argparse.ArgumentParser):
 9 |     """Simple implementation of ArgumentParser supporting config file
10 | 
11 |     This class is originated from https://github.com/bw2/ConfigArgParse,
12 |     but this class is lack of some features that it has.
13 | 
14 |     - Not supporting multiple config files
15 |     - Automatically adding "--config" as an option.
16 |     - Not supporting any formats other than yaml
17 |     - Not checking argument type
18 | 
19 |     """
20 | 
21 |     def __init__(self, *args, **kwargs):
22 |         super().__init__(*args, **kwargs)
23 |         self.add_argument("--config", help="Give config file in yaml format")
24 | 
25 |     def parse_known_args(self, args=None, namespace=None):
26 |         # Once parsing for setting from "--config"
27 |         _args, _ = super().parse_known_args(args, namespace)
28 |         if _args.config is not None:
29 |             if not Path(_args.config).exists():
30 |                 self.error(f"No such file: {_args.config}")
31 | 
32 |             with open(_args.config, "r", encoding="utf-8") as f:
33 |                 d = yaml.safe_load(f)
34 |             if not isinstance(d, dict):
35 |                 self.error("Config file has non dict value: {_args.config}")
36 | 
37 |             for key in d:
38 |                 for action in self._actions:
39 |                     if key == action.dest:
40 |                         break
41 |                 else:
42 |                     self.error(f"unrecognized arguments: {key} (from {_args.config})")
43 | 
44 |             # NOTE(kamo): Ignore "--config" from a config file
45 |             # NOTE(kamo): Unlike "configargparse", this module doesn't check type.
46 |             #   i.e. We can set any type value regardless of argument type.
47 |             self.set_defaults(**d)
48 |         return super().parse_known_args(args, namespace)
49 | 
50 | 
51 | def get_commandline_args():
52 |     extra_chars = [
53 |         " ",
54 |         ";",
55 |         "&",
56 |         "(",
57 |         ")",
58 |         "|",
59 |         "^",
60 |         "<",
61 |         ">",
62 |         "?",
63 |         "*",
64 |         "[",
65 |         "]",
66 |         "$",
67 |         "`",
68 |         '"',
69 |         "\\",
70 |         "!",
71 |         "{",
72 |         "}",
73 |     ]
74 | 
75 |     # Escape the extra characters for shell
76 |     argv = [
77 |         arg.replace("'", "'\\''")
78 |         if all(char not in arg for char in extra_chars)
79 |         else "'" + arg.replace("'", "'\\''") + "'"
80 |         for arg in sys.argv
81 |     ]
82 | 
83 |     return sys.executable + " " + " ".join(argv)


--------------------------------------------------------------------------------
/ClipVideo/clipvideo/gradio_service.py:
--------------------------------------------------------------------------------
  1 | import gradio as gr
  2 | from modelscope.pipelines import pipeline
  3 | from modelscope.utils.constant import Tasks
  4 | from videoclipper import VideoClipper
  5 | 
  6 | 
  7 | if __name__ == "__main__":
  8 |     inference_pipeline = pipeline(
  9 |         task=Tasks.auto_speech_recognition,
 10 |         model='damo/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch',
 11 |         vad_model='damo/speech_fsmn_vad_zh-cn-16k-common-pytorch',
 12 |         punc_model='damo/punc_ct-transformer_zh-cn-common-vocab272727-pytorch',
 13 |     )
 14 |     audio_clipper = VideoClipper(inference_pipeline)
 15 | 
 16 |     def audio_recog(audio_input):
 17 |         return audio_clipper.recog(audio_input)
 18 | 
 19 |     def audio_clip(dest_text, start_ost, end_ost, state):
 20 |         return audio_clipper.clip(dest_text, start_ost, end_ost, state)
 21 | 
 22 |     def video_recog(video_input):
 23 |         return audio_clipper.video_recog(video_input)
 24 | 
 25 |     def video_clip(dest_text, start_ost, end_ost, state):
 26 |         return audio_clipper.video_clip(dest_text, start_ost, end_ost, state)
 27 | 
 28 |     def video_clip_addsub(dest_text, start_ost, end_ost, state, font_size, font_color):
 29 |         return audio_clipper.video_clip(dest_text, start_ost, end_ost, state, font_size, font_color, add_sub=True)
 30 | 
 31 |     '''
 32 |     top_md_1 = ("""
 33 |     基于达摩院自研Paraformer-长音频版的语音识别、端点检测、标点预测、时间戳功能
 34 | 
 35 |     准确识别，自由复制所需段落并一键裁剪、添加字幕
 36 | 
 37 |     * Step1: 上传视频文件（或使用下方的用例体验），点击 **<font color="#f7802b">识别</font>** 按钮
 38 |     * Step2: 复制识别结果中所需的文字至右上方，设置偏移与字幕配置（可选）
 39 |     * Step3: 点击 **<font color="#f7802b">裁剪</font>** 按钮或 **<font color="#f7802b">裁剪并添加字幕</font>** 按钮获得结果
 40 |     """)
 41 |     '''
 42 | 
 43 |     top_md_2 = ("""
 44 |     受到网络传输与服务资源的限制，用于体验的视频最好大小在40mb以下
 45 |     过大的视频可以尝试分离音轨使用音频剪辑，或 **<font color="#1785c4">通过源代码将您的ClipVideo服务部署在本地（推荐）</font>** ：
 46 |     <div align="center">
 47 |     <div style="display:flex; gap: 0.25rem;" align="center">
 48 |     FunASR_APP: <a href='https://github.com/alibaba/funasr-app'><img src='https://img.shields.io/badge/Github-Code-blue'></a> 
 49 |     🌟支持我们: <a href='https://github.com/alibaba/funasr-app/stargazers'><img src='https://img.shields.io/github/stars/alibaba/funasr-app.svg?style=social'></a>
 50 |     </div>
 51 |     </div>
 52 |     """)
 53 | 
 54 |     top_md_3 = ("""访问FunASR项目与论文能够帮助您深入了解ClipVideo中所使用的语音处理相关模型：
 55 |     <div align="center">
 56 |     <div style="display:flex; gap: 0.25rem;" align="center">
 57 |         FunASR: <a href='https://github.com/alibaba-damo-academy/FunASR'><img src='https://img.shields.io/badge/Github-Code-blue'></a> 
 58 |         FunASR Paper: <a href="https://arxiv.org/abs/2305.11013"><img src="https://img.shields.io/badge/Arxiv-2305.11013-orange"></a> 
 59 |         🌟Star FunASR: <a href='https://github.com/alibaba-damo-academy/FunASR/stargazers'><img src='https://img.shields.io/github/stars/alibaba-damo-academy/FunASR.svg?style=social'></a>
 60 |     </div>
 61 |     </div>
 62 |     """)
 63 | 
 64 |     # gradio interface
 65 |     with gr.Blocks() as demo:
 66 |         #gr.Image("./examples/guide.png", show_label=False)
 67 |         # gr.Markdown(top_md_1)
 68 |         #gr.Markdown(top_md_2)
 69 |         #gr.Markdown(top_md_3)
 70 |         video_state = gr.State()
 71 |         audio_state = gr.State()
 72 |         with gr.Tab("🎥✂️视频裁剪 Video Clipping"):
 73 |             with gr.Row():
 74 |                 with gr.Column():
 75 |                     video_input = gr.Video(label="🎥视频输入 Video Input")
 76 |                     gr.Examples(['https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ClipVideo/%E4%B8%BA%E4%BB%80%E4%B9%88%E8%A6%81%E5%A4%9A%E8%AF%BB%E4%B9%A6%EF%BC%9F%E8%BF%99%E6%98%AF%E6%88%91%E5%90%AC%E8%BF%87%E6%9C%80%E5%A5%BD%E7%9A%84%E7%AD%94%E6%A1%88-%E7%89%87%E6%AE%B5.mp4', 
 77 |                                  'https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ClipVideo/2022%E4%BA%91%E6%A0%96%E5%A4%A7%E4%BC%9A_%E7%89%87%E6%AE%B5.mp4', 
 78 |                                  'https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ClipVideo/2022%E4%BA%91%E6%A0%96%E5%A4%A7%E4%BC%9A_%E7%89%87%E6%AE%B52.mp4', 
 79 |                                  'https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ClipVideo/%E4%BD%BF%E7%94%A8chatgpt_%E7%89%87%E6%AE%B5.mp4'],
 80 |                                 [video_input])
 81 |                     recog_button2 = gr.Button("👂识别 Recognize")
 82 |                     video_text_output = gr.Textbox(label="✏️识别结果 Recognition Result")
 83 |                     video_srt_output = gr.Textbox(label="📖SRT字幕内容 RST Subtitles")
 84 |                 with gr.Column():
 85 |                     video_text_input = gr.Textbox(label="✏️待裁剪文本 Text to Clip (多段文本使用'#'连接)")
 86 |                     with gr.Row():
 87 |                         video_start_ost = gr.Slider(minimum=-500, maximum=1000, value=0, step=50, label="⏪开始位置偏移 Start Offset (ms)")
 88 |                         video_end_ost = gr.Slider(minimum=-500, maximum=1000, value=100, step=50, label="⏩结束位置偏移 End Offset (ms)")
 89 |                     with gr.Row():
 90 |                         font_size = gr.Slider(minimum=10, maximum=100, value=32, step=2, label="🔠字幕字体大小 Subtitle Font Size")
 91 |                         font_color = gr.Radio(["black", "white", "green", "red"], label="🌈字幕颜色 Subtitle Color", value='white')
 92 |                         # font = gr.Radio(["黑体", "Alibaba Sans"], label="字体 Font")
 93 |                     with gr.Row():
 94 |                         clip_button2 = gr.Button("✂️裁剪\nClip")
 95 |                         clip_button3 = gr.Button("✂️裁剪并添加字幕\nClip and Generate Subtitles")
 96 |                     video_output = gr.Video(label="🎥裁剪结果 Audio Clipped")
 97 |                     video_mess_output = gr.Textbox(label="ℹ️裁剪信息 Clipping Log")
 98 |                     video_srt_clip_output = gr.Textbox(label="📖裁剪部分SRT字幕内容 Clipped RST Subtitles")
 99 | 
100 |         with gr.Tab("🔊✂️音频裁剪 Audio Clipping"):
101 |             with gr.Row():
102 |                 with gr.Column():
103 |                     audio_input = gr.Audio(label="🔊音频输入 Audio Input")
104 |                     gr.Examples(['https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ClipVideo/%E9%B2%81%E8%82%83%E9%87%87%E8%AE%BF%E7%89%87%E6%AE%B51.wav'], [audio_input])
105 |                     recog_button1 = gr.Button("👂识别 Recognize")
106 |                     audio_text_output = gr.Textbox(label="✏️识别结果 Recognition Result")
107 |                     audio_srt_output = gr.Textbox(label="📖SRT字幕内容 RST Subtitles")
108 |                 with gr.Column():
109 |                     audio_text_input = gr.Textbox(label="✏️待裁剪文本 Text to Clip (多段文本使用'#'连接)")
110 |                     with gr.Row():
111 |                         audio_start_ost = gr.Slider(minimum=-500, maximum=1000, value=0, step=50, label="⏪开始位置偏移 Start Offset (ms)")
112 |                         audio_end_ost = gr.Slider(minimum=-500, maximum=1000, value=100, step=50, label="⏩结束位置偏移 End Offset (ms)")
113 |                     with gr.Row():
114 |                         clip_button1 = gr.Button("✂️裁剪 Clip")
115 |                     audio_output = gr.Audio(label="🔊裁剪结果 Audio Clipped")
116 |                     audio_mess_output = gr.Textbox(label="ℹ️裁剪信息 Clipping Log")
117 |                     audio_srt_clip_output = gr.Textbox(label="📖裁剪部分SRT字幕内容 Clipped RST Subtitles")
118 |         
119 |         recog_button1.click(audio_recog, 
120 |                             inputs=audio_input, 
121 |                             outputs=[audio_text_output, audio_srt_output, audio_state])
122 |         clip_button1.click(audio_clip, 
123 |                            inputs=[audio_text_input, audio_start_ost, audio_end_ost, audio_state], 
124 |                            outputs=[audio_output, audio_mess_output, audio_srt_clip_output])
125 | 
126 |         recog_button2.click(video_recog, 
127 |                             inputs=video_input, 
128 |                             outputs=[video_text_output, video_srt_output, video_state])
129 |         clip_button2.click(video_clip, 
130 |                            inputs=[video_text_input, video_start_ost, video_end_ost, video_state], 
131 |                            outputs=[video_output, video_mess_output, video_srt_clip_output])
132 |         clip_button3.click(video_clip_addsub, 
133 |                            inputs=[video_text_input, video_start_ost, video_end_ost, video_state, font_size, font_color], 
134 |                            outputs=[video_output, video_mess_output, video_srt_clip_output])
135 |     
136 |     # start gradio service in local
137 |     demo.queue(concurrency_count=3).launch()
138 | 


--------------------------------------------------------------------------------
/ClipVideo/clipvideo/imagemagick_test.py:
--------------------------------------------------------------------------------
 1 | from moviepy.editor import *
 2 | from moviepy.video.tools.subtitles import SubtitlesClip
 3 | 
 4 | generator = lambda txt: TextClip(txt, font='./font/STHeitiMedium.ttc', fontsize=48, color='white')
 5 | subs = [((0, 2), 'sub1中文字幕'),
 6 |         ((2, 4), 'subs2'),
 7 |         ((4, 6), 'subs3'),
 8 |         ((6, 8), 'subs4')]
 9 | 
10 | subtitles = SubtitlesClip(subs, generator)
11 | 
12 | video = VideoFileClip("examples/2022云栖大会_片段.mp4.mp4")
13 | video = video.subclip(0, 8)
14 | video = CompositeVideoClip([video, subtitles.set_pos(('center','bottom'))])
15 | 
16 | video.write_videofile("test_output.mp4")


--------------------------------------------------------------------------------
/ClipVideo/clipvideo/subtitle_utils.py:
--------------------------------------------------------------------------------
  1 | def time_convert(ms):
  2 |     ms = int(ms)
  3 |     tail = ms % 1000
  4 |     s = ms // 1000
  5 |     mi = s // 60
  6 |     s = s % 60
  7 |     h = mi // 60
  8 |     mi = mi % 60
  9 |     h = "00" if h == 0 else str(h)
 10 |     mi = "00" if mi == 0 else str(mi)
 11 |     s = "00" if s == 0 else str(s)
 12 |     tail = str(tail)
 13 |     if len(h) == 1: h = '0' + h
 14 |     if len(mi) == 1: mi = '0' + mi
 15 |     if len(s) == 1: s = '0' + s
 16 |     return "{}:{}:{},{}".format(h, mi, s, tail)
 17 | 
 18 | 
 19 | class Text2SRT():
 20 |     def __init__(self, text_seg, ts_list, offset=0):
 21 |         self.token_list = [i for i in text_seg.split() if len(i)]
 22 |         self.ts_list = ts_list
 23 |         start, end = ts_list[0][0] - offset, ts_list[-1][1] - offset
 24 |         self.start_sec, self.end_sec = start, end
 25 |         self.start_time = time_convert(start)
 26 |         self.end_time = time_convert(end)
 27 |     def text(self):
 28 |         res = ""
 29 |         for word in self.token_list:
 30 |             if '\u4e00' <= word <= '\u9fff':
 31 |                 res += word
 32 |             else:
 33 |                 res += " " + word
 34 |         return res
 35 |     def len(self):
 36 |         return len(self.token_list)
 37 |     def srt(self, acc_ost=0.0):
 38 |         return "{} --> {}\n{}\n".format(
 39 |             time_convert(self.start_sec+acc_ost*1000),
 40 |             time_convert(self.end_sec+acc_ost*1000), 
 41 |             self.text())
 42 |     def time(self, acc_ost=0.0):
 43 |         return (self.start_sec/1000+acc_ost, self.end_sec/1000+acc_ost)
 44 | 
 45 | 
 46 | def generate_srt(sentence_list):
 47 |     srt_total = ''
 48 |     for i, d in enumerate(sentence_list):
 49 |         t2s = Text2SRT(d['text_seg'], d['ts_list'])
 50 |         srt_total += "{}\n{}".format(i, t2s.srt())
 51 |     return srt_total
 52 | 
 53 | def generate_srt_clip(sentence_list, start, end, begin_index=0, time_acc_ost=0.0):
 54 |     start, end = int(start * 1000), int(end * 1000)
 55 |     srt_total = ''
 56 |     cc = 1 + begin_index
 57 |     subs = []
 58 |     for i, d in enumerate(sentence_list):
 59 |         if d['ts_list'][-1][1] <= start:
 60 |             continue
 61 |         if d['ts_list'][0][0] >= end:
 62 |             break
 63 |         # parts in between
 64 |         if (d['ts_list'][-1][1] <= end and d['ts_list'][0][0] > start) or (d['ts_list'][-1][1] == end and d['ts_list'][0][0] == start):
 65 |             t2s = Text2SRT(d['text_seg'], d['ts_list'], offset=start)
 66 |             srt_total += "{}\n{}".format(cc, t2s.srt(time_acc_ost))
 67 |             subs.append((t2s.time(time_acc_ost), t2s.text()))
 68 |             cc += 1
 69 |             continue
 70 |         if d['ts_list'][0][0] <= start:
 71 |             if not d['ts_list'][-1][1] > end:
 72 |                 for j, ts in enumerate(d['ts_list']):
 73 |                     if ts[1] > start:
 74 |                         break
 75 |                 _text = " ".join(d['text_seg'].split()[j:])
 76 |                 _ts = d['ts_list'][j:]
 77 |             else:
 78 |                 for j, ts in enumerate(d['ts_list']):
 79 |                     if ts[1] > start:
 80 |                         _start = j
 81 |                         break
 82 |                 for j, ts in enumerate(d['ts_list']):
 83 |                     if ts[1] > end:
 84 |                         _end = j
 85 |                         break
 86 |                 _text = " ".join(d['text_seg'].split()[_start:_end])
 87 |                 _ts = d['ts_list'][_start:_end]
 88 |             if len(ts):
 89 |                 t2s = Text2SRT(_text, _ts, offset=start)
 90 |                 srt_total += "{}\n{}".format(cc, t2s.srt(time_acc_ost))
 91 |                 subs.append((t2s.time(time_acc_ost), t2s.text()))
 92 |                 cc += 1
 93 |             continue
 94 |         if d['ts_list'][-1][1] > end:
 95 |             for j, ts in enumerate(d['ts_list']):
 96 |                 if ts[1] > end:
 97 |                     break
 98 |             _text = " ".join(d['text_seg'].split()[:j])
 99 |             _ts = d['ts_list'][:j]
100 |             if len(_ts):
101 |                 t2s = Text2SRT(_text, _ts, offset=start)
102 |                 srt_total += "{}\n{}".format(cc, t2s.srt(time_acc_ost))
103 |                 subs.append(
104 |                     (t2s.time(time_acc_ost), t2s.text())
105 |                     )
106 |                 cc += 1
107 |             continue
108 |     return srt_total, subs, cc
109 | 


--------------------------------------------------------------------------------
/ClipVideo/clipvideo/test.sh:
--------------------------------------------------------------------------------
 1 | # step1: Recognize
 2 | python videoclipper.py --stage 1 \
 3 |                        --file ../examples/2022云栖大会_片段.mp4 \
 4 |                        --output_dir ./output
 5 | # now you can find recognition results and entire SRT file in ./output/
 6 | # step2: Clip
 7 | python videoclipper.py --stage 2 \
 8 |                        --file ../examples/2022云栖大会_片段.mp4 \
 9 |                        --output_dir ./output \
10 |                        --dest_text '所以这个是我们办这个奖的初心啊，我们也会一届一届的办下去' \
11 |                        --start_ost 0 \
12 |                        --end_ost 100 \
13 |                        --output_file './output/res.mp4'


--------------------------------------------------------------------------------
/ClipVideo/clipvideo/trans_utils.py:
--------------------------------------------------------------------------------
 1 | PUNC_LIST = ['，', '。', '！', '？', '、']
 2 | 
 3 | 
 4 | def pre_proc(text):
 5 |     res = ''
 6 |     for i in range(len(text)):
 7 |         if text[i] in PUNC_LIST:
 8 |             continue
 9 |         if '\u4e00' <= text[i] <= '\u9fff':
10 |             if len(res) and res[-1] != " ":
11 |                 res += ' ' + text[i]+' '
12 |             else:
13 |                 res += text[i]+' '
14 |         else:
15 |             res += text[i]
16 |     if res[-1] == ' ':
17 |         res = res[:-1]
18 |     return res
19 | 
20 | def proc(raw_text, timestamp, dest_text):
21 |     # simple matching
22 |     ld = len(dest_text.split())
23 |     mi, ts = [], []
24 |     offset = 0
25 |     while True:
26 |         fi = raw_text.find(dest_text, offset, len(raw_text))
27 |         # import pdb; pdb.set_trace()
28 |         ti = raw_text[:fi].count(' ')
29 |         if fi == -1:
30 |             break
31 |         offset = fi + ld
32 |         mi.append(fi)
33 |         ts.append([timestamp[ti][0]*16, timestamp[ti+ld-1][1]*16])
34 |         # import pdb; pdb.set_trace()
35 |     return ts
36 | 
37 | 
38 | def write_state(output_dir, state):
39 |     for key in ['/recog_res_raw', '/timestamp', '/sentences']:
40 |         with open(output_dir+key, 'w', encoding="UTF-8") as fout:
41 |             fout.write(str(state[key[1:]]))
42 | 
43 | 
44 | def load_state(output_dir):
45 |     state = {}
46 |     with open(output_dir+'/recog_res_raw', encoding='UTF-8') as fin:
47 |         line = fin.read()
48 |         state['recog_res_raw'] = line
49 |     with open(output_dir+'/timestamp', encoding='UTF-8') as fin:
50 |         line = fin.read()
51 |         state['timestamp'] = eval(line)
52 |     with open(output_dir+'/sentences', encoding='UTF-8') as fin:
53 |         line = fin.read()
54 |         state['sentences'] = eval(line)
55 |     return state
56 |         
57 |     


--------------------------------------------------------------------------------
/ClipVideo/clipvideo/videoclipper.py:
--------------------------------------------------------------------------------
  1 | import os
  2 | import sys
  3 | import copy
  4 | import librosa
  5 | import logging
  6 | import argparse
  7 | import numpy as np
  8 | import soundfile as sf
  9 | import moviepy.editor as mpy
 10 | from subtitle_utils import generate_srt, generate_srt_clip
 11 | from trans_utils import pre_proc, proc, write_state, load_state
 12 | from argparse_tools import ArgumentParser, get_commandline_args
 13 | 
 14 | from moviepy.editor import *
 15 | from moviepy.video.tools.subtitles import SubtitlesClip
 16 | 
 17 | 
 18 | class VideoClipper():
 19 |     def __init__(self, asr_pipeline):
 20 |         logging.warning("Initializing VideoClipper.")
 21 |         self.asr_pipeline = asr_pipeline
 22 | 
 23 |     def recog(self, audio_input, state=None):
 24 |         if state is None:
 25 |             state = {}
 26 |         sr, data = audio_input
 27 |         assert sr == 16000, "16kHz sample rate required, {} given.".format(sr)
 28 |         if len(data.shape) == 2:  # multi-channel wav input
 29 |             # logging.warning("Input wav shape: {}, only first channel reserved.").format(data.shape)
 30 |             data = data[:,0]
 31 |         state['audio_input'] = (sr, data)
 32 |         data = data.astype(np.float64)
 33 |         rec_result = self.asr_pipeline(audio_in=data)
 34 |         state['recog_res_raw'] = rec_result['text_postprocessed']
 35 |         state['timestamp'] = rec_result['time_stamp']
 36 |         state['sentences'] = rec_result['sentences']
 37 |         res_text = rec_result['text']
 38 |         res_srt = generate_srt(rec_result['sentences'])
 39 |         return res_text, res_srt, state
 40 | 
 41 |     def clip(self, dest_text, start_ost, end_ost, state):
 42 |         # get from state
 43 |         audio_input = state['audio_input']
 44 |         recog_res_raw = state['recog_res_raw']
 45 |         timestamp = state['timestamp']
 46 |         sentences = state['sentences']
 47 |         sr, data = audio_input
 48 |         data = data.astype(np.float64)
 49 | 
 50 |         all_ts = []
 51 |         for _dest_text in dest_text.split('#'):
 52 |             _dest_text = pre_proc(_dest_text)
 53 |             ts = proc(recog_res_raw, timestamp, _dest_text)
 54 |             for _ts in ts: all_ts.append(_ts)
 55 |         ts = all_ts
 56 |         srt_index = 0
 57 |         clip_srt = ""
 58 |         if len(ts):
 59 |             start, end = ts[0]
 60 |             start = min(max(0, start+start_ost*16), len(data))
 61 |             end = min(max(0, end+end_ost*16), len(data))
 62 |             res_audio = data[start:end]
 63 |             start_end_info = "from {} to {}".format(start/16000, end/16000)
 64 |             srt_clip, _, srt_index = generate_srt_clip(sentences, start/16000.0, end/16000.0, begin_index=srt_index)
 65 |             clip_srt += srt_clip
 66 |             for _ts in ts[1:]:  # multiple sentence input or multiple output matched
 67 |                 start, end = _ts
 68 |                 start = min(max(0, start+start_ost*16), len(data))
 69 |                 end = min(max(0, end+end_ost*16), len(data))
 70 |                 start_end_info += ", from {} to {}".format(start, end)
 71 |                 res_audio = np.concatenate([res_audio, data[start+start_ost*16:end+end_ost*16]], -1)
 72 |                 srt_clip, _, srt_index = generate_srt_clip(sentences, start/16000.0, end/16000.0, begin_index=srt_index-1)
 73 |                 clip_srt += srt_clip
 74 |         if len(ts):
 75 |             message = "{} periods found in the speech: ".format(len(ts)) + start_end_info
 76 |         else:
 77 |             message = "No period found in the speech, return raw speech. You may check the recognition result and try other destination text."
 78 |             res_audio = data
 79 |         return (sr, res_audio), message, clip_srt
 80 | 
 81 |     def video_recog(self, vedio_filename):
 82 |         vedio_filename = vedio_filename
 83 |         clip_video_file = vedio_filename[:-4] + '_clip.mp4'
 84 |         video = mpy.VideoFileClip(vedio_filename)
 85 |         audio_file = vedio_filename[:-3] + 'wav'
 86 |         video.audio.write_audiofile(audio_file)
 87 |         wav = librosa.load(audio_file, sr=16000)[0]
 88 |         state = {
 89 |             'vedio_filename': vedio_filename,
 90 |             'clip_video_file': clip_video_file,
 91 |             'video': video,
 92 |         }
 93 |         # res_text, res_srt = self.recog((16000, wav), state)
 94 |         os.remove(audio_file)
 95 |         return self.recog((16000, wav), state)
 96 | 
 97 |     def video_clip(self, dest_text, start_ost, end_ost, state, font_size=32, font_color='white', add_sub=False):
 98 |         # get from state
 99 |         recog_res_raw = state['recog_res_raw']
100 |         timestamp = state['timestamp']
101 |         sentences = state['sentences']
102 |         video = state['video']
103 |         clip_video_file = state['clip_video_file']
104 |         vedio_filename = state['vedio_filename']
105 |         
106 |         all_ts = []
107 |         srt_index = 0
108 |         time_acc_ost = 0.0
109 |         for _dest_text in dest_text.split('#'):
110 |             _dest_text = pre_proc(_dest_text)
111 |             ts = proc(recog_res_raw, timestamp, _dest_text)
112 |             for _ts in ts: all_ts.append(_ts)
113 |         ts = all_ts
114 |         clip_srt = ""
115 |         if len(ts):
116 |             start, end = ts[0][0] / 16000, ts[0][1] / 16000
117 |             srt_clip, subs, srt_index = generate_srt_clip(sentences, start, end, begin_index=srt_index, time_acc_ost=time_acc_ost)
118 |             start, end = start+start_ost/1000.0, end+end_ost/1000.0
119 |             video_clip = video.subclip(start, end)
120 |             start_end_info = "from {} to {}".format(start, end)
121 |             clip_srt += srt_clip
122 |             if add_sub:
123 |                 generator = lambda txt: TextClip(txt, font='./font/STHeitiMedium.ttc', fontsize=font_size, color=font_color)
124 |                 subtitles = SubtitlesClip(subs, generator)
125 |                 video_clip = CompositeVideoClip([video_clip, subtitles.set_pos(('center','bottom'))])
126 |             concate_clip = [video_clip]
127 |             time_acc_ost += end+end_ost/1000.0 - (start+start_ost/1000.0)
128 |             for _ts in ts[1:]:
129 |                 start, end = _ts[0] / 16000, _ts[1] / 16000
130 |                 srt_clip, subs, srt_index = generate_srt_clip(sentences, start, end, begin_index=srt_index-1, time_acc_ost=time_acc_ost)
131 |                 start, end = start+start_ost/1000.0, end+end_ost/1000.0
132 |                 _video_clip = video.subclip(start, end)
133 |                 start_end_info += ", from {} to {}".format(start, end)
134 |                 clip_srt += srt_clip
135 |                 if add_sub:
136 |                     generator = lambda txt: TextClip(txt, font='./font/STHeitiMedium.ttc', fontsize=font_size, color=font_color)
137 |                     subtitles = SubtitlesClip(subs, generator)
138 |                     _video_clip = CompositeVideoClip([_video_clip, subtitles.set_pos(('center','bottom'))])
139 |                 concate_clip.append(copy.copy(_video_clip))
140 |                 time_acc_ost += end+end_ost/1000.0 - (start+start_ost/1000.0)
141 |             message = "{} periods found in the audio: ".format(len(ts)) + start_end_info
142 |             logging.warning("Concating...")
143 |             if len(concate_clip) > 1:
144 |                 video_clip = concatenate_videoclips(concate_clip)
145 |             video_clip.write_videofile(clip_video_file, audio_codec="aac")
146 |         else:
147 |             clip_video_file = vedio_filename
148 |             message = "No period found in the audio, return raw speech. You may check the recognition result and try other destination text."
149 |             srt_clip = ''
150 |         return clip_video_file, message, clip_srt
151 | 
152 | 
153 | def get_parser():
154 |     parser = ArgumentParser(
155 |         description="ClipVideo Argument",
156 |         formatter_class=argparse.ArgumentDefaultsHelpFormatter,
157 |     )
158 |     parser.add_argument(
159 |         "--stage",
160 |         type=int,
161 |         choices=(1, 2),
162 |         help="Stage, 0 for recognizing and 1 for clipping",
163 |         required=True
164 |     )
165 |     parser.add_argument(
166 |         "--file",
167 |         type=str,
168 |         default=None,
169 |         help="Input file path",
170 |         required=True
171 |     )
172 |     parser.add_argument(
173 |         "--output_dir",
174 |         type=str,
175 |         default='./output/mp4',
176 |         help="Output files path",
177 |     )
178 |     parser.add_argument(
179 |         "--dest_text",
180 |         type=str,
181 |         default=None,
182 |         help="Destination text string for clipping",
183 |     )
184 |     parser.add_argument(
185 |         "--start_ost",
186 |         type=int,
187 |         default=0,
188 |         help="Offset time in ms at beginning for clipping"
189 |     )
190 |     parser.add_argument(
191 |         "--end_ost",
192 |         type=int,
193 |         default=0,
194 |         help="Offset time in ms at ending for clipping"
195 |     )
196 |     parser.add_argument(
197 |         "--output_file",
198 |         type=str,
199 |         default=None,
200 |         help="Output file path"
201 |     )
202 |     return parser
203 | 
204 | 
205 | def runner(stage, file, output_dir, dest_text, start_ost, end_ost, output_file, config=None):
206 |     audio_suffixs = ['wav']
207 |     video_suffixs = ['mp4']
208 |     if file[-3:] in audio_suffixs:
209 |         mode = 'audio'
210 |     elif file[-3:] in video_suffixs:
211 |         mode = 'video'
212 |     else:
213 |         logging.error("Unsupported file format: {}".format(file))
214 |     while output_dir.endswith('/'):
215 |         output_dir = output_dir[:-1]
216 |     if stage == 1:
217 |         from modelscope.pipelines import pipeline
218 |         from modelscope.utils.constant import Tasks
219 |         # initialize modelscope asr pipeline
220 |         logging.warning("Initializing modelscope asr pipeline.")
221 |         inference_pipeline = pipeline(
222 |             task=Tasks.auto_speech_recognition,
223 |             model='damo/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch',
224 |             vad_model='damo/speech_fsmn_vad_zh-cn-16k-common-pytorch',
225 |             punc_model='damo/punc_ct-transformer_zh-cn-common-vocab272727-pytorch',
226 |             output_dir=output_dir,
227 |         )
228 |         audio_clipper = VideoClipper(inference_pipeline)
229 |         if mode == 'audio':
230 |             logging.warning("Recognizing audio file: {}".format(file))
231 |             wav, sr = librosa.load(file, sr=16000)
232 |             res_text, res_srt, state = audio_clipper.recog((sr, wav))
233 |         if mode == 'video':
234 |             logging.warning("Recognizing video file: {}".format(file))
235 |             res_text, res_srt, state = audio_clipper.video_recog(file)
236 |         total_srt_file = output_dir + '/total.srt'
237 |         with open(total_srt_file, 'w') as fout:
238 |             fout.write(res_srt)
239 |             logging.warning("Write total subtitile to {}".format(total_srt_file))
240 |         write_state(output_dir, state)
241 |         logging.warning("Recognition successed. You can copy the text segment from below and use stage 2.")
242 |         print(res_text)
243 |     if stage == 2:
244 |         audio_clipper = VideoClipper(None)
245 |         if mode == 'audio':
246 |             state = load_state(output_dir)
247 |             wav, sr = librosa.load(file, sr=16000)
248 |             state['audio_input'] = (sr, wav)
249 |             (sr, audio), message, srt_clip = audio_clipper.clip(dest_text, start_ost, end_ost, state)
250 |             if output_file is None:
251 |                 output_file = output_dir + '/result.wav'
252 |             clip_srt_file = output_file[:-3] + 'srt'
253 |             logging.warning(message)
254 |             sf.write(output_file, audio, 16000)
255 |             assert output_file.endswith('.wav'), "output_file must ends with '.wav'"
256 |             logging.warning("Save clipped wav file to {}".format(output_file))
257 |             with open(clip_srt_file, 'w') as fout:
258 |                 fout.write(srt_clip)
259 |                 logging.warning("Write clipped subtitile to {}".format(clip_srt_file))
260 |         if mode == 'video':
261 |             state = load_state(output_dir)
262 |             state['vedio_filename'] = file
263 |             if output_file is None:
264 |                 state['clip_video_file'] = file[:-4] + '_clip.mp4'
265 |             else:
266 |                 state['clip_video_file'] = output_file
267 |             clip_srt_file = state['clip_video_file'][:-3] + 'srt'
268 |             state['video'] = mpy.VideoFileClip(file)
269 |             clip_video_file, message, srt_clip = audio_clipper.video_clip(dest_text, start_ost, end_ost, state)
270 |             logging.warning("Clipping Log: {}".format(message))
271 |             logging.warning("Save clipped mp4 file to {}".format(clip_video_file))
272 |             with open(clip_srt_file, 'w') as fout:
273 |                 fout.write(srt_clip)
274 |                 logging.warning("Write clipped subtitile to {}".format(clip_srt_file))
275 | 
276 | 
277 | def main(cmd=None):
278 |     print(get_commandline_args(), file=sys.stderr)
279 |     parser = get_parser()
280 |     args = parser.parse_args(cmd)
281 |     kwargs = vars(args)
282 |     runner(**kwargs)
283 | 
284 | 
285 | if __name__ == '__main__':
286 |     main()


--------------------------------------------------------------------------------
/ClipVideo/docs/images/show.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/Fatfish588/Dataset_Generator_For_VITS/2f012ac453a99723c29ffbf119b4ff545492ebb7/ClipVideo/docs/images/show.png


--------------------------------------------------------------------------------
/ClipVideo/requirments.txt:
--------------------------------------------------------------------------------
1 | librosa
2 | soundfile
3 | funasr>=0.5.5
4 | moviepy
5 | numpy
6 | gradio


--------------------------------------------------------------------------------
/FunASR-APP-LICENSE:
--------------------------------------------------------------------------------
 1 | MIT License
 2 | 
 3 | Copyright (c) 2023 Alibaba
 4 | 
 5 | Permission is hereby granted, free of charge, to any person obtaining a copy
 6 | of this software and associated documentation files (the "Software"), to deal
 7 | in the Software without restriction, including without limitation the rights
 8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 9 | copies of the Software, and to permit persons to whom the Software is
10 | furnished to do so, subject to the following conditions:
11 | 
12 | The above copyright notice and this permission notice shall be included in all
13 | copies or substantial portions of the Software.
14 | 
15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21 | SOFTWARE.
22 | 


--------------------------------------------------------------------------------
/FunASR-APP-README.md:
--------------------------------------------------------------------------------
 1 | # FunASR-APP
 2 | 
 3 | <strong>FunASR-APP</strong> is a comprehensive speech application toolkit designed to facilitate the application and integration of [FunASR](https://github.com/alibaba-damo-academy/FunASR)'s open-source speech models. Its primary goal is to package the models into convenient application packages, enabling easy application and seamless integration.
 4 | 
 5 | ## ClipVideo
 6 | 
 7 | As the first application toolkit of FunASR-APP, <strong>ClipVideo</strong> enables users to clip ```.mp4``` video files or ```.wav``` audio files with chosen text segments out of the recognition results generated by [Paraformer-long model](https://modelscope.cn/models/damo/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch/summary).
 8 | 
 9 | Under the help of ClipVideo you can get the video clips easily with the following steps (in Gradio service):
10 | - Step1: Upload your video file (or try the example videos below)
11 | - Step2: Copy the text segments you need to 'Text to Clip'
12 | - Step3: Adjust subtitle settings (if needed)
13 | - Step4: Click 'Clip' or 'Clip and Generate Subtitles'
14 | 
15 | ### Usage
16 | ```shell
17 | git clone https://github.com/alibaba-damo-academy/FunASR-APP.git
18 | cd FunASR-APP
19 | # install modelscope
20 | pip install "modelscope[audio_asr]" -f https://modelscope.oss-cn-beijing.aliyuncs.com/releases/repo.html
21 | # python environments
22 | pip install -r ClipVideo/requirments.txt
23 | ```
24 | (Optional) If you want to clip video file with embedded subtitles
25 | 
26 | 1. ffmpeg and imagemagick is required
27 | 
28 | - On Ubuntu
29 | ```shell
30 | apt-get -y update && apt-get -y install ffmpeg imagemagick
31 | sed -i 's/none/read,write/g' /etc/ImageMagick-6/policy.xml
32 | ```
33 | - On MacOS
34 | ```shell
35 | brew install imagemagick
36 | sed -i 's/none/read,write/g' /usr/local/Cellar/imagemagick/7.1.1-8_1/etc/ImageMagick-7/policy.xml 
37 | ```
38 | 2. Download font file to ClipVideo/font
39 | 
40 | ```shell
41 | wget https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ClipVideo/STHeitiMedium.ttc -O ClipVideo/font/STHeitiMedium.ttc
42 | ```
43 | 
44 | #### Experience ClipVideo in Modelscope
45 | You can try ClipVideo in modelscope space: [link](https://modelscope.cn/studios/damo/funasr_app_clipvideo/summary).
46 | 
47 | #### Use ClipVideo as Gradio Service
48 | You can establish your own ClipVideo service which is same as [Modelscope Space](https://modelscope.cn/studios/damo/funasr_app_clipvideo/summary) as follow:
49 | ```shell
50 | python ClipVideo/clipvideo/gradio_service.py
51 | ```
52 | then visit ```localhost:7860``` you will get a Gradio service like below and you can use ClipVideo following the steps:
53 | <img src="ClipVideo/docs/images/show.png"/>
54 | 
55 | #### Use ClipVideo in command line
56 | ClipVideo supports you to recognize and clip with commands:
57 | ```shell
58 | # working in ClipVideo/
59 | # step1: Recognize
60 | python clipvideo/videoclipper.py --stage 1 \
61 |                        --file examples/2022云栖大会_片段.mp4 \
62 |                        --output_dir ./output
63 | # now you can find recognition results and entire SRT file in ./output/
64 | # step2: Clip
65 | python clipvideo/videoclipper.py --stage 2 \
66 |                        --file examples/2022云栖大会_片段.mp4 \
67 |                        --output_dir ./output \
68 |                        --dest_text '我们把它跟乡村振兴去结合起来，利用我们的设计的能力' \
69 |                        --start_ost 0 \
70 |                        --end_ost 100 \
71 |                        --output_file './output/res.mp4'
72 | ```
73 | 
74 | ### Study Speech Related Models in FunASR
75 | 
76 | [FunASR](https://github.com/alibaba-damo-academy/FunASR) hopes to build a bridge between academic research and industrial applications on speech recognition. By supporting the training & finetuning of the industrial-grade speech recognition model released on ModelScope, researchers and developers can conduct research and production of speech recognition models more conveniently, and promote the development of speech recognition ecology. ASR for Fun！
77 | 
78 | 📚FunASR Paper: <a href="https://arxiv.org/abs/2305.11013"><img src="https://img.shields.io/badge/Arxiv-2305.11013-orange"></a> 
79 | 🌟Support FunASR: <a href='https://github.com/alibaba-damo-academy/FunASR/stargazers'><img src='https://img.shields.io/github/stars/alibaba-damo-academy/FunASR.svg?style=social'></a>


--------------------------------------------------------------------------------
/LICENSE:
--------------------------------------------------------------------------------
 1 | MIT License
 2 | 
 3 | Copyright (c) 2023 Fatfish588
 4 | 
 5 | Permission is hereby granted, free of charge, to any person obtaining a copy
 6 | of this software and associated documentation files (the "Software"), to deal
 7 | in the Software without restriction, including without limitation the rights
 8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 9 | copies of the Software, and to permit persons to whom the Software is
10 | furnished to do so, subject to the following conditions:
11 | 
12 | The above copyright notice and this permission notice shall be included in all
13 | copies or substantial portions of the Software.
14 | 
15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21 | SOFTWARE.
22 | 


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
 1 | # Dataset_Generator_For_VITS
 2 | 基于达摩院视频切割技术的视频转换为短音频的vits数据集生成工具   
 3 | A VITS Dataset Generation Tool for Converting Video to Short Audio Based on Damo Academy Video Cutting Technology  
 4 | 
 5 | # 介绍  
 6 | >在现在的vits语音模型训练的步骤中，标注、清洗和训练框架都有不少很棒的项目可供学习和使用，但是在数据集的准备这一项好像还没有多少人去探究。手动准备动辄几百成千上万条短音频的数据集，对于一些只想听听自己喜欢的主播、恋人等声音的vits普通爱好者来说是基本不现实的。此前(2023.10)唯一开源且能用的音视频切片机还是VITS-fast-fine-tuning快速微调中内嵌的一套基于whisper的切片机，因为耦合度较大，导致也只能在快速微调自己的项目中使用，而无法适配其他如bert-vits2等项目。  
 7 | 
 8 | >综上所诉，我尝试着基于阿里达摩院FunASR-APP的ClipVideo作简单修改，制作了这个小工具，希望能补上国内vits训练圈子的一块拼图，您只需要一个人物或者角色的音频与视频，不管多长无论多大，有一点背景音乐也无妨，就算音视频混在一起放入指定目录都没问题，然后点击运行，就可以获得一系列1～10秒的短音频，这些音频完全可以直接送入其他用于标注的项目进行下一步准备。
 9 | 
10 | 1、本项目基于阿里达摩院FunASR-APP的ClipVideo简单修改制作，其原理是使用ClipVideo通过文字去一个视频中裁剪出对应的那句话的音视频（美妙的技术），并且可以指定音频的开始和结尾偏移量，在中文方面比whisper效果更好，不会出现尾音最后一个音只有一半的情况。  
11 | 2、本项目的输出都在根目录下ClipVideo/output/中，包含了切割后的音视频和每个视频对应的字幕文件，需要的话可以使用。  
12 | 3、本项目比较简单，所以从输入视频到输出音频数据集是连续的，如果有参数和自己想要的不一样可以自行修改，例如要修改第一条中的偏移量，可在Dataset_generator.py的前半部分修改，默认是不偏移直接截取。  
13 |   
14 | 下图是效果展示，使用GPU加速的情况下2两分钟生成600句短音频.  
15 | ![7月10日](https://github.com/Fatfish588/Dataset_Generator_For_VITS/assets/59791439/00f57562-798a-4368-921c-6a6886f65d13)
16 | 
17 | # 更新日志
18 | 2023/10/12  
19 | 1、现在可以把音频(.wav)和视频(.mp4)一起混合放进video_files中直接运行了！生成器会安排好所有事！  
20 | 2、增加了demucs人声分离，现在可以接受浅bgm的视频与音频进行处理，哦对，不可以用在唱中文的背景音乐视频，因为歌词也会被识别并切片。  
21 | 3、增加了初始化输目录的功能，生成器会先把除了video_files以外的目录清空再运行，你只需要开始运行就好了，一切交给生成器！  
22 | 
23 | # 教程
24 | 1、克隆此仓库(python版本3.10，3.8往后应该都行)
25 | 
26 | ```bash
27 | git clone https://github.com/Fatfish588/Dataset_Denerator_For_VITS.git
28 | ```
29 | 
30 | 创建所需要的目录,windows直接新建文件夹就好了
31 | ```bash
32 | mkdir ClipVideo/font
33 | mkdir ClipVideo/clipvideo/video_files
34 | mkdir ClipVideo/clipvideo/output
35 | mkdir ClipVideo/clipvideo/input
36 | mkdir ClipVideo/clipvideo/input/mdx_extra
37 | mkdir ClipVideo/clipvideo/org_wav
38 | mkdir ClipVideo/clipvideo/output/mp4
39 | mkdir ClipVideo/clipvideo/output/srt
40 | mkdir ClipVideo/clipvideo/output/wav
41 | ```
42 | 此时，这部分结构应该如下图，以下操作都处于Dataset_Generator_For_VITS根目录下   
43 | ![image](https://github.com/Fatfish588/Dataset_Generator_For_VITS/assets/59791439/2c6ba932-4a9a-4c7f-902a-26cdfaa50c6b)
44 |   
45 | 
46 | 
47 | 2、安装环境依赖
48 | ```bash
49 | # install modelscope
50 | pip install "modelscope[audio_asr]" -f https://modelscope.oss-cn-beijing.aliyuncs.com/releases/repo.html  
51 | # python environments  
52 | pip install -r ClipVideo/requirments.txt
53 | pip install torchaudio 
54 | pip install demucs~=4.0.0
55 | pip install umap
56 | pip install hdbscan  
57 | # 下载字体（给webUI的字幕镶嵌用的）  
58 | wget https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ClipVideo/STHeitiMedium.ttc -O ClipVideo/font/STHeitiMedium.ttc  
59 | ```    
60 | 如果想用GPU加速（尤其是降噪）需要对应的torch和cuda，因为大家的显卡各不相同，这里只提供一个我的作为参考：RTX 4090 + cuda11.7  
61 | ```bash
62 | pip install torch==2.0.1+cu117 torchaudio==2.0.2+cu117 torchvision==0.15.2+cu117  --extra-index-url https://download.pytorch.org/whl/cu117
63 | ```
64 | 3、启动一次webUI，这一步是为了让FunASR-APP自动下载视频转文字的相关模型，此步骤下载模型比较耗时，进度条卡住不动是正常情况，稍等就好，模型来自阿里的服务器，可能需要关掉魔法。  
65 | 
66 | ```bash
67 | python ClipVideo/clipvideo/gradio_service.py
68 | ```  
69 | 成功打开webUI则说明FunASR-APP的依赖准备完成了  
70 | 
71 | 4、将要处理的音频或视频全部放入video_files目录下，支持多个视频音频混合处理，只要确保都是一个人的声音就好。    
72 | ![image](https://github.com/Fatfish588/Dataset_Denerator_For_VITS/assets/59791439/a85784e4-b390-4c5c-b02d-c5cdf50d7e1c)
73 | 
74 | 5、开始运行  
75 | 
76 | ```bash
77 | python ClipVideo/clipvideo/Dataset_generator.py 
78 | ```  
79 | 6、运行完成后，结果保存在ClipVideo/clipvideo/output/wav目录下  
80 | ![image](https://github.com/Fatfish588/Dataset_Generator_For_VITS/assets/59791439/ae24892e-7ab2-43ac-9485-46caf28b9df6)  
81 | 
82 | 
83 | # 后续计划
84 | 1、添加降噪模型，将支持带背景音乐的视频输入。(已完成)  
85 | 2、一键恢复初始化状态，免得每次都要手动清空。（已完成）  
86 | 3、优化代码，目前是步骤太多太繁杂了。  
87 | 4、从视频一键生成到音频数据集和标注训练集（大概率鸽了）  
88 | # 碎碎念
89 | 1、本项目目前只支持中文的，背景音乐不大或者是纯音乐的音频与视频，例如有声书、教程类的视频、科普类视频、虚拟主播的聊天回（天呐她们能和弹幕聊整整6个小时）等等。   
90 | 2、本项目只是将长视频生成几秒钟的短音频数据集，减少了手动切片的时间，并不带标注、重采样、生成训练集的功能。    
91 | 3、本项目生成的文件名字是Paraformer模型听写出来的，只是用作区分文件而已，并不是百分百准确，不推荐直接拿文件名去当训练集。    
92 | 4、请确保ClipVideo/clipvideo/目录下的video_files、output/mp4、output/srt、output/wav这4个目录存在，生成器现在会在每次点击运行时去清空它们，但是需要在第一次运行前先创建好它们，在这之后就不用管它们了。    
93 | 5、代码超简单的，每个方法都有备注，有些功能不需要比如降噪部分可以自己修改。    
94 | 6、关注永雏塔菲喵，关注永雏塔菲谢谢喵。    
95 | # 相关链接：
96 | [Paraformer视频自动切片与字幕（创空间）———阿里达摩院](https://modelscope.cn/studios/damo/funasr_app_clipvideo/summary)  
97 | [FunASR-APP（GitHub）](https://github.com/alibaba-damo-academy/FunASR-APP)
98 | 


--------------------------------------------------------------------------------
/TransAudio/README.md:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/Fatfish588/Dataset_Generator_For_VITS/2f012ac453a99723c29ffbf119b4ff545492ebb7/TransAudio/README.md


--------------------------------------------------------------------------------
/TransAudio/funasr-runtime-deploy.sh:
--------------------------------------------------------------------------------
   1 | #!/usr/bin/env bash
   2 | 
   3 | scriptVersion="0.0.3"
   4 | scriptDate="20230629"
   5 | 
   6 | clear
   7 | 
   8 | 
   9 | # Set color
  10 | RED="\033[31;1m"
  11 | GREEN="\033[32;1m"
  12 | YELLOW="\033[33;1m"
  13 | BLUE="\033[34;1m"
  14 | CYAN="\033[36;1m"
  15 | PLAIN="\033[0m"
  16 | 
  17 | # Info messages
  18 | ERROR="${RED}[ERROR]${PLAIN}"
  19 | WARNING="${YELLOW}[WARNING]${PLAIN}"
  20 | 
  21 | # Font Format
  22 | BOLD="\033[1m"
  23 | UNDERLINE="\033[4m"
  24 | 
  25 | # Current folder
  26 | cur_dir=`pwd`
  27 | 
  28 | 
  29 | checkConfigFileAndTouch(){
  30 |     mkdir -p /var/funasr
  31 |     if [ ! -f $FUNASR_CONFIG_FILE ]; then
  32 |         touch $FUNASR_CONFIG_FILE
  33 |     fi
  34 | }
  35 | 
  36 | SAMPLE_CLIENTS=( \
  37 | "Python" \
  38 | "Linux_Cpp" \
  39 | )
  40 | ASR_MODELS=( \
  41 | "damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-onnx" \
  42 | "model_name" \
  43 | "model_path" \
  44 | )
  45 | VAD_MODELS=( \
  46 | "damo/speech_fsmn_vad_zh-cn-16k-common-onnx" \
  47 | "model_name" \
  48 | "model_path" \
  49 | )
  50 | PUNC_MODELS=( \
  51 | "damo/punc_ct-transformer_zh-cn-common-vocab272727-onnx" \
  52 | "model_name" \
  53 | "model_path" \
  54 | )
  55 | DOCKER_IMAGES=( \
  56 | "registry.cn-hangzhou.aliyuncs.com/funasr_repo/funasr:funasr-runtime-sdk-cpu-latest" \
  57 | "registry.cn-hangzhou.aliyuncs.com/funasr_repo/funasr:funasr-runtime-sdk-cpu-0.1.0" \
  58 | )
  59 | menuSelection(){
  60 |     local menu
  61 |     menu=($(echo "$@"))
  62 |     result=1
  63 |     show_no=1
  64 |     menu_no=0
  65 |     len=${#menu[@]}
  66 | 
  67 |     while true
  68 |     do
  69 |         echo -e "    ${BOLD}${show_no})${PLAIN} ${menu[menu_no]}"
  70 | 
  71 |         let show_no++
  72 |         let menu_no++
  73 |         if [ ${menu_no} -ge ${len} ]; then
  74 |             break
  75 |         fi
  76 |     done
  77 | 
  78 |     while true
  79 |     do
  80 |         read -p "  Enter your choice: " result
  81 | 
  82 |         expr ${result} + 0 &>/dev/null
  83 |         if [ $? -eq 0 ]; then
  84 |             if [ ${result} -ge 1 ] && [ ${result} -le ${len} ]; then
  85 |                 break
  86 |             else
  87 |                 echo -e "    ${RED}Input error, please input correct number!${PLAIN}"
  88 |             fi
  89 |         else
  90 |             echo -e "    ${RED}Input error, please input correct number!${PLAIN}"
  91 |         fi
  92 |     done
  93 |     
  94 |     return $result
  95 | }
  96 | 
  97 | DrawProgress(){
  98 |     model=$1
  99 |     title=$2
 100 |     percent_str=$3
 101 |     speed=$4
 102 |     revision=$5
 103 |     latest_percent=$6
 104 | 
 105 |     progress=0
 106 |     if [ ! -z "$percent_str" ]; then
 107 |         progress=`expr $percent_str + 0`
 108 |         latest_percent=`expr $latest_percent + 0`
 109 |         if [ $progress -ne 0 ] && [ $progress -lt $latest_percent ]; then
 110 |             progress=$latest_percent
 111 |         fi
 112 |     fi
 113 | 
 114 |     LOADING_FLAG="Loading"
 115 |     if [ "$title" = "$LOADING_FLAG" ]; then
 116 |         progress=100
 117 |     fi
 118 | 
 119 |     i=0
 120 |     str=""
 121 |     let max=progress/2
 122 |     while [ $i -lt $max ]
 123 |     do
 124 |         let i++
 125 |         str+='='
 126 |     done
 127 |     let color=36
 128 |     let index=max*2
 129 |     if [ -z "$speed" ]; then
 130 |         printf "\r    \e[0;$color;1m[%s][%-11s][%-50s][%d%%][%s]\e[0m" "$model" "$title" "$str" "$$index" "$revision"
 131 |     else
 132 |         printf "\r    \e[0;$color;1m[%s][%-11s][%-50s][%3d%%][%8s][%s]\e[0m" "$model" "$title" "$str" "$index" "$speed" "$revision"
 133 |     fi
 134 |     printf "\n"
 135 | 
 136 |     return $progress
 137 | }
 138 | 
 139 | PROGRESS_TXT="/var/funasr/progress.txt"
 140 | ASR_PERCENT_INT=0
 141 | VAD_PERCENT_INT=0
 142 | PUNC_PERCENT_INT=0
 143 | ASR_TITLE="Downloading"
 144 | ASR_PERCENT="0"
 145 | ASR_SPEED="0KB/s"
 146 | ASR_REVISION=""
 147 | VAD_TITLE="Downloading"
 148 | VAD_PERCENT="0"
 149 | VAD_SPEED="0KB/s"
 150 | VAD_REVISION=""
 151 | PUNC_TITLE="Downloading"
 152 | PUNC_PERCENT="0"
 153 | PUNC_SPEED="0KB/s"
 154 | PUNC_REVISION=""
 155 | ServerProgress(){
 156 |     status_flag="STATUS:"
 157 |     stage=0
 158 |     wait=0
 159 |     server_status=""
 160 | 
 161 |     while true
 162 |     do
 163 |         if [ -f "$PROGRESS_TXT" ]; then
 164 |             break
 165 |         else
 166 |             sleep 1
 167 |             let wait=wait+1
 168 |             if [ ${wait} -ge 10 ]; then
 169 |                 break
 170 |             fi
 171 |         fi
 172 |     done
 173 | 
 174 |     if [ ! -f "$PROGRESS_TXT" ]; then
 175 |         echo -e "    ${RED}The note of progress does not exist.(${PROGRESS_TXT}) ${PLAIN}"
 176 |         return 98
 177 |     fi
 178 | 
 179 |     stage=1
 180 |     while read line
 181 |     do
 182 |         if [ $stage -eq 1 ]; then
 183 |             result=$(echo $line | grep "STATUS:")
 184 |             if [[ "$result" != "" ]]
 185 |             then
 186 |                 stage=2
 187 |                 server_status=${line#*:}
 188 |                 status=`expr $server_status + 0`
 189 |                 if [ $status -eq 99 ]; then
 190 |                     stage=99
 191 |                 fi
 192 |                 continue
 193 |             fi
 194 |         elif [ $stage -eq 2 ]; then
 195 |             result=$(echo $line | grep "ASR")
 196 |             if [[ "$result" != "" ]]
 197 |             then
 198 |                 stage=3
 199 |                 continue
 200 |             fi
 201 |         elif [ $stage -eq 3 ]; then
 202 |             result=$(echo $line | grep "VAD")
 203 |             if [[ "$result" != "" ]]
 204 |             then
 205 |                 stage=4
 206 |                 continue
 207 |             fi
 208 |             result=$(echo $line | grep "title:")
 209 |             if [[ "$result" != "" ]]
 210 |             then
 211 |                 ASR_TITLE=${line#*:}
 212 |                 continue
 213 |             fi
 214 |             result=$(echo $line | grep "percent:")
 215 |             if [[ "$result" != "" ]]
 216 |             then
 217 |                 ASR_PERCENT=${line#*:}
 218 |                 continue
 219 |             fi
 220 |             result=$(echo $line | grep "speed:")
 221 |             if [[ "$result" != "" ]]
 222 |             then
 223 |                 ASR_SPEED=${line#*:}
 224 |                 continue
 225 |             fi
 226 |             result=$(echo $line | grep "revision:")
 227 |             if [[ "$result" != "" ]]
 228 |             then
 229 |                 ASR_REVISION=${line#*:}
 230 |                 continue
 231 |             fi
 232 |         elif [ $stage -eq 4 ]; then
 233 |             result=$(echo $line | grep "PUNC")
 234 |             if [[ "$result" != "" ]]
 235 |             then
 236 |                 stage=5
 237 |                 continue
 238 |             fi
 239 |             result=$(echo $line | grep "title:")
 240 |             if [[ "$result" != "" ]]
 241 |             then
 242 |                 VAD_TITLE=${line#*:}
 243 |                 continue
 244 |             fi
 245 |             result=$(echo $line | grep "percent:")
 246 |             if [[ "$result" != "" ]]
 247 |             then
 248 |                 VAD_PERCENT=${line#*:}
 249 |                 continue
 250 |             fi
 251 |             result=$(echo $line | grep "speed:")
 252 |             if [[ "$result" != "" ]]
 253 |             then
 254 |                 VAD_SPEED=${line#*:}
 255 |                 continue
 256 |             fi
 257 |             result=$(echo $line | grep "revision:")
 258 |             if [[ "$result" != "" ]]
 259 |             then
 260 |                 VAD_REVISION=${line#*:}
 261 |                 continue
 262 |             fi
 263 |         elif [ $stage -eq 5 ]; then
 264 |             result=$(echo $line | grep "DONE")
 265 |             if [[ "$result" != "" ]]
 266 |             then
 267 |                 # Done and break.
 268 |                 stage=6
 269 |                 break
 270 |             fi
 271 |             result=$(echo $line | grep "title:")
 272 |             if [[ "$result" != "" ]]
 273 |             then
 274 |                 PUNC_TITLE=${line#*:}
 275 |                 continue
 276 |             fi
 277 |             result=$(echo $line | grep "percent:")
 278 |             if [[ "$result" != "" ]]
 279 |             then
 280 |                 PUNC_PERCENT=${line#*:}
 281 |                 continue
 282 |             fi
 283 |             result=$(echo $line | grep "speed:")
 284 |             if [[ "$result" != "" ]]
 285 |             then
 286 |                 PUNC_SPEED=${line#*:}
 287 |                 continue
 288 |             fi
 289 |             result=$(echo $line | grep "revision:")
 290 |             if [[ "$result" != "" ]]
 291 |             then
 292 |                 PUNC_REVISION=${line#*:}
 293 |                 continue
 294 |             fi
 295 |         elif [ $stage -eq 99 ]; then
 296 |             echo -e "    ${RED}ERROR: $line${PLAIN}"
 297 |         fi
 298 |     done  < $PROGRESS_TXT
 299 | 
 300 |     if [ $stage -ne 99 ]; then
 301 |         DrawProgress "ASR " $ASR_TITLE $ASR_PERCENT $ASR_SPEED $ASR_REVISION $ASR_PERCENT_INT
 302 |         ASR_PERCENT_INT=$?
 303 |         DrawProgress "VAD " $VAD_TITLE $VAD_PERCENT $VAD_SPEED $VAD_REVISION $VAD_PERCENT_INT
 304 |         VAD_PERCENT_INT=$?
 305 |         DrawProgress "PUNC" $PUNC_TITLE $PUNC_PERCENT $PUNC_SPEED $PUNC_REVISION $PUNC_PERCENT_INT
 306 |         PUNC_PERCENT_INT=$?
 307 |     fi
 308 | 
 309 |     return $stage
 310 | }
 311 | 
 312 | # Make sure root user
 313 | rootNess(){
 314 |     echo -e "${UNDERLINE}${BOLD}[0/9]${PLAIN}"
 315 |     echo -e "  ${YELLOW}Please check root access.${PLAIN}"
 316 |     echo
 317 | 
 318 |     echo -e "  ${WARNING} MUST RUN AS ${RED}ROOT${PLAIN} USER!"
 319 |     if [[ $EUID -ne 0 ]]; then
 320 |         echo -e "  ${ERROR} MUST RUN AS ${RED}ROOT${PLAIN} USER!"
 321 |     fi
 322 | 
 323 |     checkConfigFileAndTouch
 324 |     cd ${cur_dir}
 325 |     echo
 326 | }
 327 | 
 328 | selectDockerImages(){
 329 |     echo -e "${UNDERLINE}${BOLD}[1/9]${PLAIN}"
 330 |     echo -e "  ${YELLOW}Please choose the Docker image.${PLAIN}"
 331 | 
 332 |     menuSelection ${DOCKER_IMAGES[*]}
 333 |     result=$?
 334 |     index=`expr $result - 1`
 335 | 
 336 |     PARAMS_DOCKER_IMAGE=${DOCKER_IMAGES[${index}]}
 337 |     echo -e "  ${UNDERLINE}You have chosen the Docker image:${PLAIN} ${GREEN}${PARAMS_DOCKER_IMAGE}${PLAIN}"
 338 | 
 339 |     checkDockerExist
 340 |     result=$?
 341 |     result=`expr $result + 0`
 342 |     if [ ${result} -eq 50 ]; then
 343 |         return 50
 344 |     fi
 345 | 
 346 |     echo
 347 | }
 348 | 
 349 | setupModelType(){
 350 |     echo -e "${UNDERLINE}${BOLD}[2/9]${PLAIN}"
 351 |     echo -e "  ${YELLOW}Please input [Y/n] to confirm whether to automatically download model_id in ModelScope or use a local model.${PLAIN}"
 352 |     echo -e "  [y] With the model in ModelScope, the model will be automatically downloaded to Docker(${CYAN}/workspace/models${PLAIN})."
 353 |     echo -e "      If you select both the local model and the model in ModelScope, select [y]."
 354 |     echo "  [n] Use the models on the localhost, the directory where the model is located will be mapped to Docker."
 355 | 
 356 |     while true
 357 |     do
 358 |         read -p "  Setting confirmation[Y/n]: " model_id_flag
 359 | 
 360 |         if [ -z "$model_id_flag" ]; then
 361 |             model_id_flag="y"
 362 |         fi
 363 |         YES="Y"
 364 |         yes="y"
 365 |         NO="N"
 366 |         no="n"
 367 |         if [ "$model_id_flag" = "$YES" ] || [ "$model_id_flag" = "$yes" ]; then
 368 |             # please set model_id later.
 369 |             PARAMS_DOWNLOAD_MODEL_DIR="/workspace/models"
 370 |             echo -e "  ${UNDERLINE}You have chosen to use the model in ModelScope, please set the model ID in the next steps, and the model will be automatically downloaded in (${PARAMS_DOWNLOAD_MODEL_DIR}) during the run.${PLAIN}"
 371 | 
 372 |             params_local_models_dir=`sed '/^PARAMS_LOCAL_MODELS_DIR=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 373 |             if [ -z "$params_local_models_dir" ]; then
 374 |                 params_local_models_dir="${cur_dir}/models"
 375 |                 mkdir -p ${params_local_models_dir}
 376 |             fi
 377 |             while true
 378 |             do
 379 |                 echo
 380 |                 echo -e "  ${YELLOW}Please enter the local path to download models, the corresponding path in Docker is ${PARAMS_DOWNLOAD_MODEL_DIR}.${PLAIN}"
 381 |                 read -p "  Setting the local path to download models, default(${params_local_models_dir}): " PARAMS_LOCAL_MODELS_DIR
 382 |                 if [ -z "$PARAMS_LOCAL_MODELS_DIR" ]; then
 383 |                     if [ -z "$params_local_models_dir" ]; then
 384 |                         echo -e "    ${RED}The local path set is empty, please setup again.${PLAIN}"
 385 |                         continue
 386 |                     else
 387 |                         PARAMS_LOCAL_MODELS_DIR=$params_local_models_dir
 388 |                     fi
 389 |                 fi
 390 |                 if [ ! -d "$PARAMS_LOCAL_MODELS_DIR" ]; then
 391 |                     echo -e "    ${RED}The local model path(${PARAMS_LOCAL_MODELS_DIR}) set does not exist, please setup again.${PLAIN}"
 392 |                 else
 393 |                     echo -e "  The local path(${GREEN}${PARAMS_LOCAL_MODELS_DIR}${PLAIN}) set will store models during the run."
 394 |                     break
 395 |                 fi
 396 |             done
 397 | 
 398 |             break
 399 |         elif [ "$model_id_flag" = "$NO" ] || [ "$model_id_flag" = "$no" ]; then
 400 |             # download_model_dir is empty, will use models in localhost.
 401 |             PARAMS_DOWNLOAD_MODEL_DIR=""
 402 |             PARAMS_LOCAL_MODELS_DIR=""
 403 |             echo -e "  ${UNDERLINE}You have chosen to use models from the localhost, set the path to each model in the localhost in the next steps.${PLAIN}"
 404 |             echo
 405 |             break
 406 |         fi
 407 |     done
 408 | 
 409 |     echo
 410 | }
 411 | 
 412 | # Set asr model for FunASR server
 413 | setupAsrModelId(){
 414 |     echo -e "  ${UNDERLINE}${BOLD}[2.1/9]${PLAIN}"
 415 | 
 416 |     if [ -z "$PARAMS_DOWNLOAD_MODEL_DIR" ]; then
 417 |         # download_model_dir is empty, will use models in localhost.
 418 |         params_local_asr_path=`sed '/^PARAMS_LOCAL_ASR_PATH=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 419 |         if [ -z "$params_local_asr_path" ]; then
 420 |             PARAMS_LOCAL_ASR_PATH=""
 421 |         else
 422 |             PARAMS_LOCAL_ASR_PATH=${params_local_asr_path}
 423 |         fi
 424 | 
 425 |         echo -e "    ${YELLOW}Please input ASR model path in local for FunASR server.${PLAIN}"
 426 |         echo -e "    Default: ${CYAN}${PARAMS_LOCAL_ASR_PATH}${PLAIN}"
 427 | 
 428 |         while true
 429 |         do
 430 |             read -p "    Setting ASR model path in localhost: " PARAMS_LOCAL_ASR_PATH
 431 | 
 432 |             if [ -z "$PARAMS_LOCAL_ASR_PATH" ]; then
 433 |                 PARAMS_LOCAL_ASR_PATH=${params_local_asr_path}
 434 |             fi
 435 |             if [ -z "$PARAMS_LOCAL_ASR_PATH" ]; then
 436 |                 # use default asr model in Docker
 437 |                 PARAMS_LOCAL_ASR_DIR=""
 438 |                 PARAMS_DOCKER_ASR_DIR=""
 439 |                 PARAMS_DOCKER_ASR_PATH="/workspace/models/asr"
 440 |                 echo -e "    ${RED}Donnot set the local ASR model path, will use ASR model(${CYAN}/workspace/models/asr${PLAIN}${RED}) in Docker.${PLAIN}"
 441 | 
 442 |                 echo -e "    ${UNDERLINE}You have chosen the default model dir in localhost:${PLAIN} ${GREEN}${PARAMS_LOCAL_ASR_PATH}${PLAIN}"
 443 |                 echo -e "    ${UNDERLINE}The defalut model dir in Docker is${PLAIN} ${GREEN}${PARAMS_DOCKER_ASR_PATH}${PLAIN}"
 444 |                 break
 445 |             else
 446 |                 if [ ! -d "$PARAMS_LOCAL_ASR_PATH" ]; then
 447 |                     echo -e "    ${RED}The ASR model path set does not exist, please setup again.${PLAIN}"
 448 |                 else
 449 |                     # use asr model in localhost
 450 |                     PARAMS_LOCAL_ASR_DIR=$(dirname "$PARAMS_LOCAL_ASR_PATH")
 451 |                     asr_name=$(basename "$PARAMS_LOCAL_ASR_PATH")
 452 |                     PARAMS_DOCKER_ASR_DIR="/workspace/user_asr"
 453 |                     PARAMS_DOCKER_ASR_PATH=${PARAMS_DOCKER_ASR_DIR}/${asr_name}
 454 | 
 455 |                     echo -e "    ${UNDERLINE}You have chosen the model dir in localhost:${PLAIN} ${GREEN}${PARAMS_LOCAL_ASR_PATH}${PLAIN}"
 456 |                     echo -e "    ${UNDERLINE}The model dir in Docker is${PLAIN} ${GREEN}${PARAMS_DOCKER_ASR_PATH}${PLAIN}"
 457 |                     break
 458 |                 fi
 459 |             fi
 460 |         done
 461 | 
 462 |         PARAMS_ASR_ID=""
 463 |     else
 464 |         # please set model_id later.
 465 |         echo -e "    ${YELLOW}Please select ASR model_id in ModelScope from the list below.${PLAIN}"
 466 | 
 467 |         menuSelection ${ASR_MODELS[*]}
 468 |         result=$?
 469 |         index=`expr $result - 1`
 470 |         PARAMS_ASR_ID=${ASR_MODELS[${index}]}
 471 | 
 472 |         OTHERS="model_name"
 473 |         LOCAL_MODEL="model_path"
 474 |         if [ "$PARAMS_ASR_ID" = "$OTHERS" ]; then
 475 |             params_asr_id=`sed '/^PARAMS_ASR_ID=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 476 |             if [ -z "$params_asr_id" ]; then
 477 |                 PARAMS_ASR_ID=""
 478 |             else
 479 |                 PARAMS_ASR_ID=${params_asr_id}
 480 |             fi
 481 | 
 482 |             echo -e "    Default: ${CYAN}${PARAMS_ASR_ID}${PLAIN}"
 483 | 
 484 |             while true
 485 |             do
 486 |                 read -p "    Setting ASR model_id in ModelScope: " PARAMS_ASR_ID
 487 | 
 488 |                 PARAMS_LOCAL_ASR_DIR=""
 489 |                 PARAMS_LOCAL_ASR_PATH=""
 490 |                 PARAMS_DOCKER_ASR_DIR=""
 491 |                 if [ -z "$PARAMS_ASR_ID" ]; then
 492 |                     echo -e "    ${RED}The ASR model ID is empty, please setup again.${PLAIN}"
 493 |                 else
 494 |                     break
 495 |                 fi
 496 |             done
 497 |         elif [ "$PARAMS_ASR_ID" = "$LOCAL_MODEL" ]; then
 498 |             PARAMS_ASR_ID=""
 499 |             echo -e "    Please input ASR model path in local for FunASR server."
 500 | 
 501 |             while true
 502 |             do
 503 |                 read -p "    Setting ASR model path in localhost: " PARAMS_LOCAL_ASR_PATH
 504 |                 if [ -z "$PARAMS_LOCAL_ASR_PATH" ]; then
 505 |                     # use default asr model in Docker
 506 |                     echo -e "    ${RED}Please do not set an empty path in localhost.${PLAIN}"
 507 |                 else
 508 |                     if [ ! -d "$PARAMS_LOCAL_ASR_PATH" ]; then
 509 |                         echo -e "    ${RED}The ASR model path set does not exist, please setup again.${PLAIN}"
 510 |                     else
 511 |                         # use asr model in localhost
 512 |                         PARAMS_LOCAL_ASR_DIR=$(dirname "$PARAMS_LOCAL_ASR_PATH")
 513 |                         asr_name=$(basename "$PARAMS_LOCAL_ASR_PATH")
 514 |                         PARAMS_DOCKER_ASR_DIR="/workspace/user_asr"
 515 |                         PARAMS_DOCKER_ASR_PATH=${PARAMS_DOCKER_ASR_DIR}/${asr_name}
 516 | 
 517 |                         echo -e "    ${UNDERLINE}You have chosen the model dir in localhost:${PLAIN} ${GREEN}${PARAMS_LOCAL_ASR_PATH}${PLAIN}"
 518 |                         echo -e "    ${UNDERLINE}The model dir in Docker is${PLAIN} ${GREEN}${PARAMS_DOCKER_ASR_PATH}${PLAIN}"
 519 |                         echo
 520 |                         return 0
 521 |                     fi
 522 |                 fi
 523 |             done
 524 |         fi
 525 | 
 526 |         PARAMS_DOCKER_ASR_DIR=$PARAMS_DOWNLOAD_MODEL_DIR
 527 |         PARAMS_DOCKER_ASR_PATH=${PARAMS_DOCKER_ASR_DIR}/${PARAMS_ASR_ID}
 528 | 
 529 |         echo -e "    ${UNDERLINE}The model ID is${PLAIN} ${GREEN}${PARAMS_ASR_ID}${PLAIN}"
 530 |         echo -e "    ${UNDERLINE}The model dir in Docker is${PLAIN} ${GREEN}${PARAMS_DOCKER_ASR_PATH}${PLAIN}"
 531 |     fi
 532 | 
 533 |     echo
 534 | }
 535 | 
 536 | # Set vad model for FunASR server
 537 | setupVadModelId(){
 538 |     echo -e "  ${UNDERLINE}${BOLD}[2.2/9]${PLAIN}"
 539 | 
 540 |     if [ -z "$PARAMS_DOWNLOAD_MODEL_DIR" ]; then
 541 |         # download_model_dir is empty, will use models in localhost.
 542 |         params_local_vad_path=`sed '/^PARAMS_LOCAL_VAD_PATH=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 543 |         if [ -z "$params_local_vad_path" ]; then
 544 |             PARAMS_LOCAL_VAD_PATH=""
 545 |         else
 546 |             PARAMS_LOCAL_VAD_PATH=${params_local_vad_path}
 547 |         fi
 548 | 
 549 |         echo -e "    ${YELLOW}Please input VAD model path in local for FunASR server.${PLAIN}"
 550 |         echo -e "    Default: ${CYAN}${PARAMS_LOCAL_VAD_PATH}${PLAIN}"
 551 | 
 552 |         while true
 553 |         do
 554 |             read -p "    Setting VAD model path in localhost: " PARAMS_LOCAL_VAD_PATH
 555 | 
 556 |             if [ -z "$PARAMS_LOCAL_VAD_PATH" ]; then
 557 |                 PARAMS_LOCAL_VAD_PATH=${params_local_vad_path}
 558 |             fi
 559 |             if [ -z "$PARAMS_LOCAL_VAD_PATH" ]; then
 560 |                 # use default vad model in Docker
 561 |                 PARAMS_LOCAL_VAD_DIR=""
 562 |                 PARAMS_DOCKER_VAD_DIR=""
 563 |                 PARAMS_DOCKER_VAD_PATH="/workspace/models/vad"
 564 |                 echo -e "    ${RED}Donnot set the local VAD model path, will use VAD model(${CYAN}/workspace/models/vad${PLAIN}${RED}) in Docker.${PLAIN}"
 565 | 
 566 |                 echo -e "    ${UNDERLINE}You have chosen the default model dir in localhost${PLAIN}: ${GREEN}${PARAMS_LOCAL_VAD_PATH}${PLAIN}"
 567 |                 echo -e "    ${UNDERLINE}The defalut model dir in Docker is${PLAIN} ${GREEN}${PARAMS_DOCKER_VAD_PATH}${PLAIN}"
 568 |                 break
 569 |             else
 570 |                 if [ ! -d "$PARAMS_LOCAL_VAD_PATH" ]; then
 571 |                     echo -e "    ${RED}The VAD model path set does not exist, please setup again.${PLAIN}"
 572 |                 else
 573 |                     # use vad model in localhost
 574 |                     PARAMS_LOCAL_VAD_DIR=$(dirname "$PARAMS_LOCAL_VAD_PATH")
 575 |                     vad_name=$(basename "$PARAMS_LOCAL_VAD_PATH")
 576 |                     PARAMS_DOCKER_VAD_DIR="/workspace/user_vad"
 577 |                     PARAMS_DOCKER_VAD_PATH=${PARAMS_DOCKER_VAD_DIR}/${vad_name}
 578 | 
 579 |                     echo -e "    ${UNDERLINE}You have chosen the model dir in localhost:${PLAIN} ${GREEN}${PARAMS_LOCAL_VAD_PATH}${PLAIN}"
 580 |                     echo -e "    ${UNDERLINE}The model dir in Docker is${PLAIN} ${GREEN}${PARAMS_DOCKER_VAD_PATH}${PLAIN}"
 581 |                     break
 582 |                 fi
 583 |             fi
 584 |         done
 585 | 
 586 |         PARAMS_VAD_ID=""
 587 |     else
 588 |         # please set model_id later.
 589 |         echo -e "    ${YELLOW}Please select VAD model_id in ModelScope from the list below.${PLAIN}"
 590 | 
 591 |         menuSelection ${VAD_MODELS[*]}
 592 |         result=$?
 593 |         index=`expr $result - 1`
 594 |         PARAMS_VAD_ID=${VAD_MODELS[${index}]}
 595 | 
 596 |         OTHERS="model_name"
 597 |         LOCAL_MODEL="model_path"
 598 |         if [ "$PARAMS_VAD_ID" = "$OTHERS" ]; then
 599 |             params_vad_id=`sed '/^PARAMS_VAD_ID=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 600 |             if [ -z "$params_vad_id" ]; then
 601 |                 PARAMS_VAD_ID=""
 602 |             else
 603 |                 PARAMS_VAD_ID=${params_vad_id}
 604 |             fi
 605 | 
 606 |             echo -e "    Default: ${CYAN}${PARAMS_VAD_ID}${PLAIN}"
 607 | 
 608 |             while true
 609 |             do
 610 |                 read -p "    Setting VAD model_id in ModelScope: " PARAMS_VAD_ID
 611 | 
 612 |                 PARAMS_LOCAL_VAD_DIR=""
 613 |                 PARAMS_LOCAL_VAD_PATH=""
 614 |                 PARAMS_DOCKER_VAD_DIR=""
 615 |                 if [ -z "$PARAMS_VAD_ID" ]; then
 616 |                     echo -e "    ${RED}The VAD model ID is empty, please setup again.${PLAIN}"
 617 |                 else
 618 |                     break
 619 |                 fi
 620 |             done
 621 |         elif [ "$PARAMS_VAD_ID" = "$LOCAL_MODEL" ]; then
 622 |             PARAMS_VAD_ID=""
 623 |             echo -e "    Please input VAD model path in local for FunASR server."
 624 | 
 625 |             while true
 626 |             do
 627 |                 read -p "    Setting VAD model path in localhost: " PARAMS_LOCAL_VAD_PATH
 628 |                 if [ -z "$PARAMS_LOCAL_VAD_PATH" ]; then
 629 |                     # use default vad model in Docker
 630 |                     echo -e "    ${RED}Please do not set an empty path in localhost.${PLAIN}"
 631 |                 else
 632 |                     if [ ! -d "$PARAMS_LOCAL_VAD_PATH" ]; then
 633 |                         echo -e "    ${RED}The VAD model path set does not exist, please setup again.${PLAIN}"
 634 |                     else
 635 |                         # use vad model in localhost
 636 |                         PARAMS_LOCAL_VAD_DIR=$(dirname "$PARAMS_LOCAL_VAD_PATH")
 637 |                         vad_name=$(basename "$PARAMS_LOCAL_VAD_PATH")
 638 |                         PARAMS_DOCKER_VAD_DIR="/workspace/user_vad"
 639 |                         PARAMS_DOCKER_VAD_PATH=${PARAMS_DOCKER_VAD_DIR}/${vad_name}
 640 | 
 641 |                         echo -e "    ${UNDERLINE}You have chosen the model dir in localhost:${PLAIN} ${GREEN}${PARAMS_LOCAL_VAD_PATH}${PLAIN}"
 642 |                         echo -e "    ${UNDERLINE}The model dir in Docker is${PLAIN} ${GREEN}${PARAMS_DOCKER_VAD_PATH}${PLAIN}"
 643 |                         echo
 644 |                         return 0
 645 |                     fi
 646 |                 fi
 647 |             done
 648 |         fi
 649 | 
 650 |         PARAMS_DOCKER_VAD_DIR=$PARAMS_DOWNLOAD_MODEL_DIR
 651 |         PARAMS_DOCKER_VAD_PATH=${PARAMS_DOCKER_VAD_DIR}/${PARAMS_VAD_ID}
 652 | 
 653 |         echo -e "    ${UNDERLINE}The model ID is${PLAIN} ${GREEN}${PARAMS_VAD_ID}${PLAIN}"
 654 |         echo -e "    ${UNDERLINE}The model dir in Docker is${PLAIN} ${GREEN}${PARAMS_DOCKER_VAD_PATH}${PLAIN}"
 655 |     fi
 656 | 
 657 |     echo
 658 | }
 659 | 
 660 | # Set punc model for FunASR server
 661 | setupPuncModelId(){
 662 |     echo -e "  ${UNDERLINE}${BOLD}[2.3/9]${PLAIN}"
 663 | 
 664 |     if [ -z "$PARAMS_DOWNLOAD_MODEL_DIR" ]; then
 665 |         # download_model_dir is empty, will use models in localhost.
 666 |         params_local_punc_path=`sed '/^PARAMS_LOCAL_PUNC_PATH=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 667 |         if [ -z "$params_local_punc_path" ]; then
 668 |             PARAMS_LOCAL_PUNC_PATH=""
 669 |         else
 670 |             PARAMS_LOCAL_PUNC_PATH=${params_local_punc_path}
 671 |         fi
 672 | 
 673 |         echo -e "    ${YELLOW}Please input PUNC model path in local for FunASR server.${PLAIN}"
 674 |         echo -e "    Default: ${CYAN}${PARAMS_LOCAL_PUNC_PATH}${PLAIN}"
 675 | 
 676 |         while true
 677 |         do
 678 |             read -p "    Setting PUNC model path in localhost: " PARAMS_LOCAL_PUNC_PATH
 679 | 
 680 |             if [ -z "$PARAMS_LOCAL_PUNC_PATH" ]; then
 681 |                 PARAMS_LOCAL_PUNC_PATH=${params_local_punc_path}
 682 |             fi
 683 |             if [ -z "$PARAMS_LOCAL_PUNC_PATH" ]; then
 684 |                 # use default punc model in Docker
 685 |                 PARAMS_LOCAL_PUNC_DIR=""
 686 |                 PARAMS_DOCKER_PUNC_DIR=""
 687 |                 PARAMS_DOCKER_PUNC_PATH="/workspace/models/punc"
 688 |                 echo -e "    ${RED}Donnot set the local PUNC model path, will use PUNC model(${CYAN}/workspace/models/punc${PLAIN}${RED}) in Docker.${PLAIN}"
 689 | 
 690 |                 echo -e "    ${UNDERLINE}You have chosen the default model dir in localhost: ${GREEN}${PARAMS_LOCAL_PUNC_PATH}${PLAIN}"
 691 |                 echo -e "    ${UNDERLINE}The defalut model dir in Docker is ${GREEN}${PARAMS_DOCKER_PUNC_PATH}${PLAIN}"
 692 |                 break
 693 |             else
 694 |                 if [ ! -d "$PARAMS_LOCAL_PUNC_PATH" ]; then
 695 |                     echo -e "    ${RED}The PUNC model path set does not exist, please setup again.${PLAIN}"
 696 |                 else
 697 |                     # use punc model in localhost
 698 |                     PARAMS_LOCAL_PUNC_DIR=$(dirname "$PARAMS_LOCAL_PUNC_PATH")
 699 |                     punc_name=$(basename "$PARAMS_LOCAL_PUNC_PATH")
 700 |                     PARAMS_DOCKER_PUNC_DIR="/workspace/user_punc"
 701 |                     PARAMS_DOCKER_PUNC_PATH=${PARAMS_DOCKER_PUNC_DIR}/${punc_name}
 702 | 
 703 |                     echo -e "    ${UNDERLINE}You have chosen the model dir in localhost:${PLAIN} ${GREEN}${PARAMS_LOCAL_PUNC_PATH}${PLAIN}"
 704 |                     echo -e "    ${UNDERLINE}The model dir in Docker is${PLAIN} ${GREEN}${PARAMS_DOCKER_PUNC_PATH}${PLAIN}"
 705 |                     break
 706 |                 fi
 707 |             fi
 708 |         done
 709 | 
 710 |         PARAMS_PUNC_ID=""
 711 |     else
 712 |         # please set model_id later.
 713 |         echo -e "    ${YELLOW}Please select PUNC model_id in ModelScope from the list below.${PLAIN}"
 714 | 
 715 |         menuSelection ${PUNC_MODELS[*]}
 716 |         result=$?
 717 |         index=`expr $result - 1`
 718 |         PARAMS_PUNC_ID=${PUNC_MODELS[${index}]}
 719 | 
 720 |         OTHERS="model_name"
 721 |         LOCAL_MODEL="model_path"
 722 |         if [ "$PARAMS_PUNC_ID" = "$OTHERS" ]; then
 723 |             params_punc_id=`sed '/^PARAMS_PUNC_ID=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 724 |             if [ -z "$params_punc_id" ]; then
 725 |                 PARAMS_PUNC_ID=""
 726 |             else
 727 |                 PARAMS_PUNC_ID=${params_punc_id}
 728 |             fi
 729 | 
 730 |             echo -e "    Default: ${CYAN}${PARAMS_PUNC_ID}${PLAIN}"
 731 | 
 732 |             while true
 733 |             do
 734 |                 read -p "    Setting PUNC model_id in ModelScope: " PARAMS_PUNC_ID
 735 | 
 736 |                 PARAMS_LOCAL_PUNC_DIR=""
 737 |                 PARAMS_LOCAL_PUNC_PATH=""
 738 |                 PARAMS_DOCKER_PUNC_DIR=""
 739 |                 if [ -z "$PARAMS_PUNC_ID" ]; then
 740 |                     echo -e "    ${RED}The PUNC model ID is empty, please setup again.${PLAIN}"
 741 |                 else
 742 |                     break
 743 |                 fi
 744 |             done
 745 |         elif [ "$PARAMS_PUNC_ID" = "$LOCAL_MODEL" ]; then
 746 |             PARAMS_PUNC_ID=""
 747 |             echo -e "    Please input PUNC model path in local for FunASR server."
 748 | 
 749 |             while true
 750 |             do
 751 |                 read -p "    Setting PUNC model path in localhost: " PARAMS_LOCAL_PUNC_PATH
 752 |                 if [ -z "$PARAMS_LOCAL_PUNC_PATH" ]; then
 753 |                     # use default punc model in Docker
 754 |                     echo -e "    ${RED}Please do not set an empty path in localhost.${PLAIN}"
 755 |                 else
 756 |                     if [ ! -d "$PARAMS_LOCAL_PUNC_PATH" ]; then
 757 |                         echo -e "    ${RED}The PUNC model path set does not exist, please setup again.${PLAIN}"
 758 |                     else
 759 |                         # use punc model in localhost
 760 |                         PARAMS_LOCAL_PUNC_DIR=$(dirname "$PARAMS_LOCAL_PUNC_PATH")
 761 |                         punc_name=$(basename "$PARAMS_LOCAL_PUNC_PATH")
 762 |                         PARAMS_DOCKER_PUNC_DIR="/workspace/user_punc"
 763 |                         PARAMS_DOCKER_PUNC_PATH=${PARAMS_DOCKER_PUNC_DIR}/${punc_name}
 764 | 
 765 |                         echo -e "    ${UNDERLINE}You have chosen the model dir in localhost:${PLAIN} ${GREEN}${PARAMS_LOCAL_PUNC_PATH}${PLAIN}"
 766 |                         echo -e "    ${UNDERLINE}The model dir in Docker is${PLAIN} ${GREEN}${PARAMS_DOCKER_PUNC_PATH}${PLAIN}"
 767 |                         echo
 768 |                         return 0
 769 |                     fi
 770 |                 fi
 771 |             done
 772 |         fi
 773 | 
 774 |         PARAMS_DOCKER_PUNC_DIR=$PARAMS_DOWNLOAD_MODEL_DIR
 775 |         PARAMS_DOCKER_PUNC_PATH=${PARAMS_DOCKER_PUNC_DIR}/${PARAMS_PUNC_ID}
 776 | 
 777 |         echo -e "    ${UNDERLINE}The model ID is${PLAIN} ${GREEN}${PARAMS_PUNC_ID}${PLAIN}"
 778 |         echo -e "    ${UNDERLINE}The model dir in Docker is${PLAIN} ${GREEN}${PARAMS_DOCKER_PUNC_PATH}${PLAIN}"
 779 |     fi
 780 | 
 781 |     echo
 782 | }
 783 | 
 784 | # Set server exec for FunASR
 785 | setupServerExec(){
 786 |     echo -e "${UNDERLINE}${BOLD}[3/9]${PLAIN}"
 787 | 
 788 |     params_docker_exec_path=`sed '/^PARAMS_DOCKER_EXEC_PATH=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 789 |     if [ -z "$params_docker_exec_path" ]; then
 790 |         PARAMS_DOCKER_EXEC_PATH="/workspace/FunASR/funasr/runtime/websocket/build/bin/funasr-wss-server"
 791 |     else
 792 |         PARAMS_DOCKER_EXEC_PATH=${params_docker_exec_path}
 793 |     fi
 794 | 
 795 |     echo -e "  ${YELLOW}Please enter the path to the excutor of the FunASR service on the localhost.${PLAIN}"
 796 |     echo -e "  If not set, the default ${CYAN}/workspace/FunASR/funasr/runtime/websocket/build/bin/funasr-wss-server${PLAIN} in Docker is used."
 797 |     read -p "  Setting the path to the excutor of the FunASR service on the localhost: " PARAMS_LOCAL_EXEC_PATH
 798 | 
 799 |     if [ -z "$PARAMS_LOCAL_EXEC_PATH" ]; then
 800 |         PARAMS_DOCKER_EXEC_PATH="/workspace/FunASR/funasr/runtime/websocket/build/bin/funasr-wss-server"
 801 |     else
 802 |         if [ ! -d "$PARAMS_LOCAL_EXEC_PATH" ]; then
 803 |             echo -e "  ${RED}The FunASR server path set does not exist, will use default.${PLAIN}"
 804 |             PARAMS_LOCAL_EXEC_PATH=""
 805 |             PARAMS_LOCAL_EXEC_DIR=""
 806 |             PARAMS_DOCKER_EXEC_PATH="/workspace/FunASR/funasr/runtime/websocket/build/bin/funasr-wss-server"
 807 |             PARAMS_DOCKER_EXEC_DIR="/workspace/FunASR/funasr/runtime/websocket/build/bin"
 808 |         else
 809 |             PARAMS_LOCAL_EXEC_DIR=$(dirname "$PARAMS_LOCAL_EXEC_PATH")
 810 |             exec=$(basename "$PARAMS_LOCAL_EXEC_PATH")
 811 |             PARAMS_DOCKER_EXEC_DIR="/server"
 812 |             PARAMS_DOCKER_EXEC_PATH=${PARAMS_DOCKER_EXEC_DIR}/${exec}
 813 |             echo -e "  ${UNDERLINE}The path of FunASR in localhost is${PLAIN} ${GREEN}${PARAMS_LOCAL_EXEC_PATH}${PLAIN}"
 814 |         fi
 815 |     fi
 816 |     echo -e "  ${UNDERLINE}Corresponding, the path of FunASR in Docker is${PLAIN} ${GREEN}${PARAMS_DOCKER_EXEC_PATH}${PLAIN}"
 817 | 
 818 |     echo
 819 | }
 820 | 
 821 | # Configure FunASR server host port setting
 822 | setupHostPort(){
 823 |     echo -e "${UNDERLINE}${BOLD}[4/9]${PLAIN}"
 824 | 
 825 |     params_host_port=`sed '/^PARAMS_HOST_PORT=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 826 |     if [ -z "$params_host_port" ]; then
 827 |         PARAMS_HOST_PORT="10095"
 828 |     else
 829 |         PARAMS_HOST_PORT=${params_host_port}
 830 |     fi
 831 | 
 832 |     while true
 833 |     do
 834 |         echo -e "  ${YELLOW}Please input the opened port in the host used for FunASR server.${PLAIN}"
 835 |         echo -e "  Default: ${CYAN}${PARAMS_HOST_PORT}${PLAIN}"
 836 |         read -p "  Setting the opened host port [1-65535]: " PARAMS_HOST_PORT
 837 | 
 838 |         if [ -z "$PARAMS_HOST_PORT" ]; then
 839 |             params_host_port=`sed '/^PARAMS_HOST_PORT=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 840 |             if [ -z "$params_host_port" ]; then
 841 |                 PARAMS_HOST_PORT="10095"
 842 |             else
 843 |                 PARAMS_HOST_PORT=${params_host_port}
 844 |             fi
 845 |         fi
 846 |         expr ${PARAMS_HOST_PORT} + 0 &>/dev/null
 847 |         if [ $? -eq 0 ]; then
 848 |             if [ ${PARAMS_HOST_PORT} -ge 1 ] && [ ${PARAMS_HOST_PORT} -le 65535 ]; then
 849 |                 echo -e "  ${UNDERLINE}The port of the host is${PLAIN} ${GREEN}${PARAMS_HOST_PORT}${PLAIN}"
 850 |                 echo -e "  ${UNDERLINE}The port in Docker for FunASR server is ${PLAIN}${GREEN}${PARAMS_DOCKER_PORT}${PLAIN}"
 851 |                 break
 852 |             else
 853 |                 echo -e "  ${RED}Input error, please input correct number!${PLAIN}"
 854 |             fi
 855 |         else
 856 |             echo -e "  ${RED}Input error, please input correct number!${PLAIN}"
 857 |         fi
 858 |     done
 859 |     echo
 860 | }
 861 | 
 862 | setupThreadNum(){
 863 |     echo -e "${UNDERLINE}${BOLD}[5/9]${PLAIN}"
 864 | 
 865 |     params_decoder_thread_num=`sed '/^PARAMS_DECODER_THREAD_NUM=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 866 |     if [ -z "$params_decoder_thread_num" ]; then
 867 |         PARAMS_DECODER_THREAD_NUM=$CPUNUM
 868 |     else
 869 |         PARAMS_DECODER_THREAD_NUM=${params_decoder_thread_num}
 870 |     fi
 871 | 
 872 |     while true
 873 |     do
 874 |         echo -e "  ${YELLOW}Please input thread number for FunASR decoder.${PLAIN}"
 875 |         echo -e "  Default: ${CYAN}${PARAMS_DECODER_THREAD_NUM}${PLAIN}"
 876 |         read -p "  Setting the number of decoder thread: " PARAMS_DECODER_THREAD_NUM
 877 | 
 878 |         if [ -z "$PARAMS_DECODER_THREAD_NUM" ]; then
 879 |             params_decoder_thread_num=`sed '/^PARAMS_DECODER_THREAD_NUM=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 880 |             if [ -z "$params_decoder_thread_num" ]; then
 881 |                 PARAMS_DECODER_THREAD_NUM=$CPUNUM
 882 |             else
 883 |                 PARAMS_DECODER_THREAD_NUM=${params_decoder_thread_num}
 884 |             fi
 885 |         fi
 886 |         expr ${PARAMS_DECODER_THREAD_NUM} + 0 &>/dev/null
 887 |         if [ $? -eq 0 ]; then
 888 |             if [ ${PARAMS_DECODER_THREAD_NUM} -ge 1 ] && [ ${PARAMS_DECODER_THREAD_NUM} -le 65535 ]; then
 889 |                 break
 890 |             else
 891 |                 echo -e "  ${RED}Input error, please input correct number!${PLAIN}"
 892 |             fi
 893 |         else
 894 |             echo -e "  ${RED}Input error, please input correct number!${PLAIN}"
 895 |         fi
 896 |         done
 897 |     echo
 898 | 
 899 |     multiple_io=4
 900 |     PARAMS_DECODER_THREAD_NUM=`expr $PARAMS_DECODER_THREAD_NUM + 0`
 901 |     PARAMS_IO_THREAD_NUM=`expr $PARAMS_DECODER_THREAD_NUM / $multiple_io`
 902 |     if [ $PARAMS_IO_THREAD_NUM -eq 0 ]; then
 903 |         PARAMS_IO_THREAD_NUM=1
 904 |     fi
 905 | 
 906 |     echo -e "  ${UNDERLINE}The number of decoder threads is${PLAIN} ${GREEN}${PARAMS_DECODER_THREAD_NUM}${PLAIN}"
 907 |     echo -e "  ${UNDERLINE}The number of IO threads is${PLAIN} ${GREEN}${PARAMS_IO_THREAD_NUM}${PLAIN}"
 908 |     echo
 909 | }
 910 | 
 911 | paramsFromDefault(){
 912 |     echo -e "${UNDERLINE}${BOLD}[2-5/9]${PLAIN}"
 913 |     echo -e "  ${YELLOW}Load parameters from ${FUNASR_CONFIG_FILE}${PLAIN}"
 914 |     echo
 915 | 
 916 |     PARAMS_DOCKER_IMAGE=`sed '/^PARAMS_DOCKER_IMAGE=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 917 |     PARAMS_LOCAL_MODELS_DIR=`sed '/^PARAMS_LOCAL_MODELS_DIR=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 918 |     PARAMS_DOWNLOAD_MODEL_DIR=`sed '/^PARAMS_DOWNLOAD_MODEL_DIR=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 919 |     PARAMS_LOCAL_ASR_PATH=`sed '/^PARAMS_LOCAL_ASR_PATH=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 920 |     PARAMS_DOCKER_ASR_PATH=`sed '/^PARAMS_DOCKER_ASR_PATH=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 921 |     PARAMS_ASR_ID=`sed '/^PARAMS_ASR_ID=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 922 |     PARAMS_LOCAL_VAD_PATH=`sed '/^PARAMS_LOCAL_VAD_PATH=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 923 |     PARAMS_DOCKER_VAD_PATH=`sed '/^PARAMS_DOCKER_VAD_PATH=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 924 |     PARAMS_VAD_ID=`sed '/^PARAMS_VAD_ID=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 925 |     PARAMS_LOCAL_PUNC_PATH=`sed '/^PARAMS_LOCAL_PUNC_PATH=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 926 |     PARAMS_DOCKER_PUNC_PATH=`sed '/^PARAMS_DOCKER_PUNC_PATH=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 927 |     PARAMS_PUNC_ID=`sed '/^PARAMS_PUNC_ID=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 928 |     PARAMS_DOCKER_EXEC_PATH=`sed '/^PARAMS_DOCKER_EXEC_PATH=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 929 |     PARAMS_HOST_PORT=`sed '/^PARAMS_HOST_PORT=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 930 |     PARAMS_DOCKER_PORT=`sed '/^PARAMS_DOCKER_PORT=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 931 |     PARAMS_DECODER_THREAD_NUM=`sed '/^PARAMS_DECODER_THREAD_NUM=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 932 |     PARAMS_IO_THREAD_NUM=`sed '/^PARAMS_IO_THREAD_NUM=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
 933 | }
 934 | 
 935 | saveParams(){
 936 |     echo "$i" > $FUNASR_CONFIG_FILE
 937 |     echo -e "  ${GREEN}Parameters are stored in the file ${FUNASR_CONFIG_FILE}${PLAIN}"
 938 | 
 939 |     echo "PARAMS_DOCKER_IMAGE=${PARAMS_DOCKER_IMAGE}" > $FUNASR_CONFIG_FILE
 940 |     echo "PARAMS_LOCAL_MODELS_DIR=${PARAMS_LOCAL_MODELS_DIR}" >> $FUNASR_CONFIG_FILE
 941 |     echo "PARAMS_DOWNLOAD_MODEL_DIR=${PARAMS_DOWNLOAD_MODEL_DIR}" >> $FUNASR_CONFIG_FILE
 942 | 
 943 |     echo "PARAMS_LOCAL_EXEC_PATH=${PARAMS_LOCAL_EXEC_PATH}" >> $FUNASR_CONFIG_FILE
 944 |     echo "PARAMS_LOCAL_EXEC_DIR=${PARAMS_LOCAL_EXEC_DIR}" >> $FUNASR_CONFIG_FILE
 945 |     echo "PARAMS_DOCKER_EXEC_PATH=${PARAMS_DOCKER_EXEC_PATH}" >> $FUNASR_CONFIG_FILE
 946 |     echo "PARAMS_DOCKER_EXEC_DIR=${PARAMS_DOCKER_EXEC_DIR}" >> $FUNASR_CONFIG_FILE
 947 | 
 948 |     echo "PARAMS_LOCAL_ASR_PATH=${PARAMS_LOCAL_ASR_PATH}" >> $FUNASR_CONFIG_FILE
 949 |     echo "PARAMS_LOCAL_ASR_DIR=${PARAMS_LOCAL_ASR_DIR}" >> $FUNASR_CONFIG_FILE
 950 |     echo "PARAMS_DOCKER_ASR_PATH=${PARAMS_DOCKER_ASR_PATH}" >> $FUNASR_CONFIG_FILE
 951 |     echo "PARAMS_DOCKER_ASR_DIR=${PARAMS_DOCKER_ASR_DIR}" >> $FUNASR_CONFIG_FILE
 952 |     echo "PARAMS_ASR_ID=${PARAMS_ASR_ID}" >> $FUNASR_CONFIG_FILE
 953 | 
 954 |     echo "PARAMS_LOCAL_PUNC_PATH=${PARAMS_LOCAL_PUNC_PATH}" >> $FUNASR_CONFIG_FILE
 955 |     echo "PARAMS_LOCAL_PUNC_DIR=${PARAMS_LOCAL_PUNC_DIR}" >> $FUNASR_CONFIG_FILE
 956 |     echo "PARAMS_DOCKER_PUNC_PATH=${PARAMS_DOCKER_PUNC_PATH}" >> $FUNASR_CONFIG_FILE
 957 |     echo "PARAMS_DOCKER_PUNC_DIR=${PARAMS_DOCKER_PUNC_DIR}" >> $FUNASR_CONFIG_FILE
 958 |     echo "PARAMS_PUNC_ID=${PARAMS_PUNC_ID}" >> $FUNASR_CONFIG_FILE
 959 | 
 960 |     echo "PARAMS_LOCAL_VAD_PATH=${PARAMS_LOCAL_VAD_PATH}" >> $FUNASR_CONFIG_FILE
 961 |     echo "PARAMS_LOCAL_VAD_DIR=${PARAMS_LOCAL_VAD_DIR}" >> $FUNASR_CONFIG_FILE
 962 |     echo "PARAMS_DOCKER_VAD_PATH=${PARAMS_DOCKER_VAD_PATH}" >> $FUNASR_CONFIG_FILE
 963 |     echo "PARAMS_DOCKER_VAD_DIR=${PARAMS_DOCKER_VAD_DIR}" >> $FUNASR_CONFIG_FILE
 964 |     echo "PARAMS_VAD_ID=${PARAMS_VAD_ID}" >> $FUNASR_CONFIG_FILE
 965 | 
 966 |     echo "PARAMS_HOST_PORT=${PARAMS_HOST_PORT}" >> $FUNASR_CONFIG_FILE
 967 |     echo "PARAMS_DOCKER_PORT=${PARAMS_DOCKER_PORT}" >> $FUNASR_CONFIG_FILE
 968 |     echo "PARAMS_DECODER_THREAD_NUM=${PARAMS_DECODER_THREAD_NUM}" >> $FUNASR_CONFIG_FILE
 969 |     echo "PARAMS_IO_THREAD_NUM=${PARAMS_IO_THREAD_NUM}" >> $FUNASR_CONFIG_FILE
 970 | }
 971 | 
 972 | showAllParams(){
 973 |     echo -e "${UNDERLINE}${BOLD}[6/9]${PLAIN}"
 974 |     echo -e "  ${YELLOW}Show parameters of FunASR server setting and confirm to run ...${PLAIN}"
 975 |     echo
 976 | 
 977 |     if [ ! -z "$PARAMS_DOCKER_IMAGE" ]; then
 978 |         echo -e "  The current Docker image is                                    : ${GREEN}${PARAMS_DOCKER_IMAGE}${PLAIN}"
 979 |     fi
 980 | 
 981 |     if [ ! -z "$PARAMS_LOCAL_MODELS_DIR" ]; then
 982 |         echo -e "  The model is downloaded or stored to this directory in local   : ${GREEN}${PARAMS_LOCAL_MODELS_DIR}${PLAIN}"
 983 |     fi
 984 |     if [ ! -z "$PARAMS_DOWNLOAD_MODEL_DIR" ]; then
 985 |         echo -e "  The model will be automatically downloaded to the directory    : ${GREEN}${PARAMS_DOWNLOAD_MODEL_DIR}${PLAIN}"
 986 |     fi
 987 | 
 988 |     if [ ! -z "$PARAMS_ASR_ID" ]; then
 989 |         echo -e "  The ASR model_id used                                          : ${GREEN}${PARAMS_ASR_ID}${PLAIN}"
 990 |     fi
 991 |     if [ ! -z "$PARAMS_LOCAL_ASR_PATH" ]; then
 992 |         echo -e "  The path to the local ASR model directory for the load         : ${GREEN}${PARAMS_LOCAL_ASR_PATH}${PLAIN}"
 993 |     fi
 994 |     echo -e "  The ASR model directory corresponds to the directory in Docker : ${GREEN}${PARAMS_DOCKER_ASR_PATH}${PLAIN}"
 995 | 
 996 |     if [ ! -z "$PARAMS_VAD_ID" ]; then
 997 |         echo -e "  The VAD model_id used                                          : ${GREEN}${PARAMS_VAD_ID}${PLAIN}"
 998 |     fi
 999 |     if [ ! -z "$PARAMS_LOCAL_VAD_PATH" ]; then
1000 |         echo -e "  The path to the local VAD model directory for the load         : ${GREEN}${PARAMS_LOCAL_VAD_PATH}${PLAIN}"
1001 |     fi
1002 |     echo -e "  The VAD model directory corresponds to the directory in Docker : ${GREEN}${PARAMS_DOCKER_VAD_PATH}${PLAIN}"
1003 | 
1004 |     if [ ! -z "$PARAMS_PUNC_ID" ]; then
1005 |         echo -e "  The PUNC model_id used                                         : ${GREEN}${PARAMS_PUNC_ID}${PLAIN}"
1006 |     fi
1007 |     if [ ! -z "$PARAMS_LOCAL_PUNC_PATH" ]; then
1008 |         echo -e "  The path to the local PUNC model directory for the load        : ${GREEN}${PARAMS_LOCAL_PUNC_PATH}${PLAIN}"
1009 |     fi
1010 |     echo -e "  The PUNC model directory corresponds to the directory in Docker: ${GREEN}${PARAMS_DOCKER_PUNC_PATH}${PLAIN}"
1011 |     echo
1012 | 
1013 |     if [ ! -z "$PARAMS_LOCAL_EXEC_PATH" ]; then
1014 |         echo -e "  The local path of the FunASR service executor                 : ${GREEN}${PARAMS_LOCAL_EXEC_PATH}${PLAIN}"
1015 |     fi
1016 |     echo -e "  The path in the docker of the FunASR service executor          : ${GREEN}${PARAMS_DOCKER_EXEC_PATH}${PLAIN}"
1017 | 
1018 |     echo -e "  Set the host port used for use by the FunASR service           : ${GREEN}${PARAMS_HOST_PORT}${PLAIN}"
1019 |     echo -e "  Set the docker port used by the FunASR service                 : ${GREEN}${PARAMS_DOCKER_PORT}${PLAIN}"
1020 | 
1021 |     echo -e "  Set the number of threads used for decoding the FunASR service : ${GREEN}${PARAMS_DECODER_THREAD_NUM}${PLAIN}"
1022 |     echo -e "  Set the number of threads used for IO the FunASR service       : ${GREEN}${PARAMS_IO_THREAD_NUM}${PLAIN}"
1023 | 
1024 |     echo
1025 |     while true
1026 |     do
1027 |         params_confirm="y"
1028 |         echo -e "  ${YELLOW}Please input [Y/n] to confirm the parameters.${PLAIN}"
1029 |         echo -e "  [y] Verify that these parameters are correct and that the service will run."
1030 |         echo -e "  [n] The parameters set are incorrect, it will be rolled out, please rerun."
1031 |         read -p "  read confirmation[Y/n]: " params_confirm
1032 | 
1033 |         if [ -z "$params_confirm" ]; then
1034 |             params_confirm="y"
1035 |         fi
1036 |         YES="Y"
1037 |         yes="y"
1038 |         NO="N"
1039 |         no="n"
1040 |         echo
1041 |         if [ "$params_confirm" = "$YES" ] || [ "$params_confirm" = "$yes" ]; then
1042 |             echo -e "  ${GREEN}Will run FunASR server later ...${PLAIN}"
1043 |             break
1044 |         elif [ "$params_confirm" = "$NO" ] || [ "$params_confirm" = "$no" ]; then
1045 |             echo -e "  ${RED}The parameters set are incorrect, please rerun ...${PLAIN}"
1046 |             exit 1
1047 |         else
1048 |             echo "again ..."
1049 |         fi
1050 |     done
1051 | 
1052 |     saveParams
1053 |     echo
1054 |     sleep 1
1055 | }
1056 | 
1057 | # Install docker
1058 | installDocker(){
1059 |     echo -e "${UNDERLINE}${BOLD}[7/9]${PLAIN}"
1060 | 
1061 |     if [ $DOCKERINFOLEN -gt 30 ]; then
1062 |         echo -e "  ${YELLOW}Docker has installed.${PLAIN}"
1063 |     else
1064 |         lowercase_osid=$(echo $OSID | tr '[A-Z]' '[a-z]')
1065 |         echo -e "  ${YELLOW}Start install docker for $lowercase_osid ${PLAIN}"
1066 |         DOCKER_INSTALL_CMD="curl -fsSL https://get.docker.com | bash -s docker --mirror Aliyun"
1067 |         DOCKER_INSTALL_RUN_CMD=""
1068 | 
1069 |         case "$lowercase_osid" in
1070 |             ubuntu)
1071 |                 DOCKER_INSTALL_CMD="curl -fsSL https://test.docker.com -o test-docker.sh"
1072 |                 DOCKER_INSTALL_RUN_CMD="sudo sh test-docker.sh"
1073 |                 ;;
1074 |             centos)
1075 |                 DOCKER_INSTALL_CMD="curl -fsSL https://get.docker.com | bash -s docker --mirror Aliyun"
1076 |                 ;;
1077 |             debian)
1078 |                 DOCKER_INSTALL_CMD="curl -fsSL https://get.docker.com -o get-docker.sh"
1079 |                 DOCKER_INSTALL_RUN_CMD="sudo sh get-docker.sh"
1080 |                 ;;
1081 |             *)
1082 |                 echo "$lowercase_osid is not supported."
1083 |                 ;;
1084 |         esac
1085 | 
1086 |         echo -e "  Get docker installer: ${GREEN}$DOCKER_INSTALL_CMD${PLAIN}"
1087 |         echo -e "  Get docker run: ${GREEN}$DOCKER_INSTALL_RUN_CMD${PLAIN}"
1088 | 
1089 |         $DOCKER_INSTALL_CMD
1090 |         if [ ! -z "$DOCKER_INSTALL_RUN_CMD" ]; then
1091 |             $DOCKER_INSTALL_RUN_CMD
1092 |         fi
1093 | 
1094 |         DOCKERINFO=$(sudo docker info | wc -l)
1095 |         DOCKERINFOLEN=$(expr $DOCKERINFO)
1096 |         if [ $DOCKERINFOLEN -gt 30 ]; then
1097 |             echo -e "  ${GREEN}Docker install success, start docker server.${PLAIN}"
1098 |             sudo systemctl start docker
1099 |         else
1100 |             echo -e "  ${RED}Docker install failed!${PLAIN}"
1101 |             exit 1
1102 |         fi
1103 |     fi
1104 | 
1105 |     echo
1106 |     sleep 1
1107 | }
1108 | 
1109 | # Download docker image
1110 | downloadDockerImage(){
1111 |     echo -e "${UNDERLINE}${BOLD}[8/9]${PLAIN}"
1112 |     echo -e "  ${YELLOW}Pull docker image(${PARAMS_DOCKER_IMAGE})...${PLAIN}"
1113 | 
1114 |     sudo docker pull ${PARAMS_DOCKER_IMAGE}
1115 | 
1116 |     echo
1117 |     sleep 1
1118 | }
1119 | 
1120 | dockerRun(){
1121 |     echo -e "${UNDERLINE}${BOLD}[9/9]${PLAIN}"
1122 |     echo -e "  ${YELLOW}Construct command and run docker ...${PLAIN}"
1123 | 
1124 |     RUN_CMD="sudo docker run"
1125 |     PORT_MAP=" -p ${PARAMS_HOST_PORT}:${PARAMS_DOCKER_PORT}"
1126 |     DIR_PARAMS=" --privileged=true"
1127 |     DIR_MAP_PARAMS=""
1128 |     if [ ! -z "$PARAMS_LOCAL_ASR_DIR" ]; then
1129 |         if [ -z "$DIR_MAP_PARAMS" ]; then
1130 |             DIR_MAP_PARAMS="${DIR_PARAMS} -v ${PARAMS_LOCAL_ASR_DIR}:${PARAMS_DOCKER_ASR_DIR}"
1131 |         else
1132 |             DIR_MAP_PARAMS="${DIR_MAP_PARAMS} -v ${PARAMS_LOCAL_ASR_DIR}:${PARAMS_DOCKER_ASR_DIR}"
1133 |         fi
1134 |     fi
1135 |     if [ ! -z "$PARAMS_LOCAL_VAD_DIR" ]; then
1136 |         if [ -z "$DIR_MAP_PARAMS" ]; then
1137 |             DIR_MAP_PARAMS="${DIR_PARAMS} -v ${PARAMS_LOCAL_VAD_DIR}:${PARAMS_DOCKER_VAD_DIR}"
1138 |         else
1139 |             DIR_MAP_PARAMS="${DIR_MAP_PARAMS} -v ${PARAMS_LOCAL_VAD_DIR}:${PARAMS_DOCKER_VAD_DIR}"
1140 |         fi
1141 |     fi
1142 |     if [ ! -z "$PARAMS_LOCAL_PUNC_DIR" ]; then
1143 |         if [ -z "$DIR_MAP_PARAMS" ]; then
1144 |             DIR_MAP_PARAMS="${DIR_PARAMS} -v ${PARAMS_LOCAL_PUNC_DIR}:${PARAMS_DOCKER_PUNC_DIR}"
1145 |         else
1146 |             DIR_MAP_PARAMS="${DIR_MAP_PARAMS} -v ${PARAMS_LOCAL_VAD_DIR}:${PARAMS_DOCKER_VAD_DIR}"
1147 |         fi
1148 |     fi
1149 |     if [ ! -z "$PARAMS_LOCAL_EXEC_DIR" ]; then
1150 |         if [ -z "$DIR_MAP_PARAMS" ]; then
1151 |             DIR_MAP_PARAMS="${DIR_PARAMS} -v ${PARAMS_LOCAL_EXEC_DIR}:${PARAMS_DOCKER_EXEC_DIR}"
1152 |         else
1153 |             DIR_MAP_PARAMS="${DIR_MAP_PARAMS} -v ${PARAMS_LOCAL_EXEC_DIR}:${PARAMS_DOCKER_EXEC_DIR}"
1154 |         fi
1155 |     fi
1156 |     if [ ! -z "$PARAMS_LOCAL_MODELS_DIR" ]; then
1157 |         if [ -z "$DIR_MAP_PARAMS" ]; then
1158 |             DIR_MAP_PARAMS="${DIR_PARAMS} -v ${PARAMS_LOCAL_MODELS_DIR}:${PARAMS_DOWNLOAD_MODEL_DIR}"
1159 |         else
1160 |             DIR_MAP_PARAMS="${DIR_MAP_PARAMS} -v ${PARAMS_LOCAL_MODELS_DIR}:${PARAMS_DOWNLOAD_MODEL_DIR}"
1161 |         fi
1162 |     fi
1163 | 
1164 |     EXEC_PARAMS="\"exec\":\"${PARAMS_DOCKER_EXEC_PATH}\""
1165 |     if [ ! -z "$PARAMS_ASR_ID" ]; then
1166 |         ASR_PARAMS="\"--model-dir\":\"${PARAMS_ASR_ID}\""
1167 |     else
1168 |         ASR_PARAMS="\"--model-dir\":\"${PARAMS_DOCKER_ASR_PATH}\""
1169 |     fi
1170 |     if [ ! -z "$PARAMS_VAD_ID" ]; then
1171 |         VAD_PARAMS="\"--vad-dir\":\"${PARAMS_VAD_ID}\""
1172 |     else
1173 |         VAD_PARAMS="\"--vad-dir\":\"${PARAMS_DOCKER_VAD_PATH}\""
1174 |     fi
1175 |     if [ ! -z "$PARAMS_PUNC_ID" ]; then
1176 |         PUNC_PARAMS="\"--punc-dir\":\"${PARAMS_PUNC_ID}\""
1177 |     else
1178 |         PUNC_PARAMS="\"--punc-dir\":\"${PARAMS_DOCKER_PUNC_PATH}\""
1179 |     fi
1180 |     DOWNLOAD_PARARMS="\"--download-model-dir\":\"${PARAMS_DOWNLOAD_MODEL_DIR}\""
1181 |     if [ -z "$PARAMS_DOWNLOAD_MODEL_DIR" ]; then
1182 |         MODEL_PARAMS="${ASR_PARAMS},${VAD_PARAMS},${PUNC_PARAMS}"
1183 |     else
1184 |         MODEL_PARAMS="${ASR_PARAMS},${VAD_PARAMS},${PUNC_PARAMS},${DOWNLOAD_PARARMS}"
1185 |     fi
1186 | 
1187 |     DECODER_PARAMS="\"--decoder-thread-num\":\"${PARAMS_DECODER_THREAD_NUM}\""
1188 |     IO_PARAMS="\"--io-thread-num\":\"${PARAMS_IO_THREAD_NUM}\""
1189 |     THREAD_PARAMS=${DECODER_PARAMS},${IO_PARAMS}
1190 |     PORT_PARAMS="\"--port\":\"${PARAMS_DOCKER_PORT}\""
1191 |     CRT_PATH="\"--certfile\":\"/workspace/FunASR/funasr/runtime/ssl_key/server.crt\""
1192 |     KEY_PATH="\"--keyfile\":\"/workspace/FunASR/funasr/runtime/ssl_key/server.key\""
1193 | 
1194 |     ENV_PARAMS=" -v /var/funasr:/workspace/.config"
1195 |     ENV_PARAMS=" ${ENV_PARAMS} --env DAEMON_SERVER_CONFIG={\"server\":[{${EXEC_PARAMS},${MODEL_PARAMS},${THREAD_PARAMS},${PORT_PARAMS},${CRT_PATH},${KEY_PATH}}]}"
1196 | 
1197 |     RUN_CMD="${RUN_CMD}${PORT_MAP}${DIR_MAP_PARAMS}${ENV_PARAMS}"
1198 |     RUN_CMD="${RUN_CMD} -it -d ${PARAMS_DOCKER_IMAGE}"
1199 | 
1200 |     # check Docker
1201 |     checkDockerExist
1202 |     result=$?
1203 |     result=`expr $result + 0`
1204 |     if [ ${result} -eq 50 ]; then
1205 |         return 50
1206 |     fi
1207 | 
1208 |     server_log="/var/funasr/server_console.log"
1209 |     rm -f ${PROGRESS_TXT}
1210 |     rm -f ${server_log}
1211 | 
1212 |     ${RUN_CMD}
1213 | 
1214 |     echo
1215 |     echo -e "  ${YELLOW}Loading models:${PLAIN}"
1216 | 
1217 |     # Hide the cursor, start draw progress.
1218 |     printf "\e[?25l"
1219 |     while true
1220 |     do
1221 |         ServerProgress
1222 |         result=$?
1223 |         stage=`expr $result + 0`
1224 |         if [ ${stage} -eq 0 ]; then
1225 |             break
1226 |         elif [ ${stage} -gt 0 ] && [ ${stage} -lt 6 ]; then
1227 |             sleep 0.1
1228 |             # clear 3 lines
1229 |             printf "\033[3A"
1230 |         elif [ ${stage} -eq 6 ]; then
1231 |             break
1232 |         elif [ ${stage} -eq 98 ]; then
1233 |             return 98
1234 |         else
1235 |             echo -e "  ${RED}Starting FunASR server failed.${PLAIN}"
1236 |             echo
1237 |             # Display the cursor
1238 |             printf "\e[?25h"
1239 |             return 99
1240 |         fi
1241 |     done
1242 |     # Display the cursor
1243 |     printf "\e[?25h"
1244 | 
1245 |     echo -e "  ${GREEN}The service has been started.${PLAIN}"
1246 |     echo
1247 |     echo -e "  ${BOLD}If you want to see an example of how to use the client, you can run ${PLAIN}${GREEN}sudo bash funasr-runtime-deploy.sh -c${PLAIN} ."
1248 |     echo
1249 | }
1250 | 
1251 | checkDockerExist(){
1252 |     result=$(sudo docker ps | grep ${PARAMS_DOCKER_IMAGE} | wc -l)
1253 |     result=`expr $result + 0`
1254 |     if [ ${result} -ne 0 ]; then
1255 |         echo
1256 |         echo -e "  ${RED}Docker: ${PARAMS_DOCKER_IMAGE} has been launched, please run (${PLAIN}${GREEN}sudo bash funasr-runtime-deploy.sh -p${PLAIN}${RED}) to stop Docker first.${PLAIN}"
1257 |         return 50
1258 |     fi
1259 | }
1260 | 
1261 | dockerExit(){
1262 |     echo -e "  ${YELLOW}Stop docker(${PARAMS_DOCKER_IMAGE}) server ...${PLAIN}"
1263 |     sudo docker stop `sudo docker ps -a| grep ${PARAMS_DOCKER_IMAGE} | awk '{print $1}' `
1264 |     echo
1265 |     sleep 1
1266 | }
1267 | 
1268 | modelChange(){
1269 |     model_id=$1
1270 | 
1271 |     result=$(echo $1 | grep "asr")
1272 |     if [[ "$result" != "" ]]
1273 |     then
1274 |         PARAMS_ASR_ID=$1
1275 |         PARAMS_DOCKER_ASR_PATH=${PARAMS_DOCKER_ASR_DIR}/${PARAMS_ASR_ID}
1276 |         return 0
1277 |     fi
1278 |     result=$(echo $1 | grep "vad")
1279 |     if [[ "$result" != "" ]]
1280 |     then
1281 |         PARAMS_VAD_ID=$1
1282 |         PARAMS_DOCKER_VAD_PATH=${PARAMS_DOCKER_VAD_DIR}/${PARAMS_VAD_ID}
1283 |         retun 0
1284 |     fi
1285 |     result=$(echo $1 | grep "punc")
1286 |     if [[ "$result" != "" ]]
1287 |     then
1288 |         PARAMS_PUNC_ID=$1
1289 |         PARAMS_DOCKER_PUNC_PATH=${PARAMS_DOCKER_PUNC_DIR}/${PARAMS_PUNC_ID}
1290 |         retun 0
1291 |     fi
1292 | }
1293 | 
1294 | sampleClientRun(){
1295 |     echo -e "${YELLOW}Will download sample tools for the client to show how speech recognition works.${PLAIN}"
1296 | 
1297 |     sample_name="funasr_samples"
1298 |     sample_tar="funasr_samples.tar.gz"
1299 |     sample_url="https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/sample/${sample_tar}"
1300 |     DOWNLOAD_SAMPLE="curl -O ${sample_url}"
1301 |     UNTAR_CMD="tar -zxf ${sample_tar}"
1302 | 
1303 |     if [ ! -f "${sample_tar}" ]; then
1304 |         ${DOWNLOAD_SAMPLE}
1305 |     fi
1306 |     if [ -f "${sample_tar}" ]; then
1307 |         ${UNTAR_CMD}
1308 |     fi
1309 |     if [ -d "${sample_name}" ]; then
1310 | 
1311 |         echo -e "  Please select the client you want to run."
1312 |         menuSelection ${SAMPLE_CLIENTS[*]}
1313 |         result=$?
1314 |         index=`expr $result - 1`
1315 |         lang=${SAMPLE_CLIENTS[${index}]}
1316 |         echo
1317 | 
1318 |         SERVER_IP="127.0.0.1"
1319 |         read -p "  Please enter the IP of server, default(${SERVER_IP}): " SERVER_IP
1320 |         if [ -z "$SERVER_IP" ]; then
1321 |             SERVER_IP="127.0.0.1"
1322 |         fi
1323 | 
1324 |         HOST_PORT=`sed '/^PARAMS_HOST_PORT=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
1325 |         if [ -z "$HOST_PORT" ]; then
1326 |             HOST_PORT="10095"
1327 |         fi
1328 |         read -p "  Please enter the port of server, default(${HOST_PORT}): " HOST_PORT
1329 |         if [ -z "$HOST_PORT" ]; then
1330 |             HOST_PORT=`sed '/^PARAMS_HOST_PORT=/!d;s/.*=//' ${FUNASR_CONFIG_FILE}`
1331 |             if [ -z "$HOST_PORT" ]; then
1332 |                 HOST_PORT="10095"
1333 |             fi
1334 |         fi
1335 | 
1336 |         WAV_PATH="${cur_dir}/funasr_samples/audio/asr_example.wav"
1337 |         read -p "  Please enter the audio path, default(${WAV_PATH}): " WAV_PATH
1338 |         if [ -z "$WAV_PATH" ]; then
1339 |             WAV_PATH="${cur_dir}/funasr_samples/audio/asr_example.wav"
1340 |         fi
1341 | 
1342 |         echo
1343 |         PRE_CMD=”“
1344 |         case "$lang" in
1345 |             Linux_Cpp)
1346 |                 PRE_CMD="export LD_LIBRARY_PATH=${cur_dir}/funasr_samples/cpp/libs:\$LD_LIBRARY_PATH"
1347 |                 CLIENT_EXEC="${cur_dir}/funasr_samples/cpp/funasr-wss-client"
1348 |                 RUN_CMD="${CLIENT_EXEC} --server-ip ${SERVER_IP} --port ${HOST_PORT} --wav-path ${WAV_PATH}"
1349 |                 echo -e "  Run ${BLUE}${PRE_CMD}${PLAIN}"
1350 |                 ${PRE_CMD}
1351 |                 echo
1352 |                 ;;
1353 |             Python)
1354 |                 CLIENT_EXEC="${cur_dir}/funasr_samples/python/wss_client_asr.py"
1355 |                 RUN_CMD="python3 ${CLIENT_EXEC} --host ${SERVER_IP} --port ${HOST_PORT} --mode offline --audio_in ${WAV_PATH} --send_without_sleep --output_dir ./funasr_samples/python"
1356 |                 PRE_CMD="pip3 install click>=8.0.4"
1357 |                 echo -e "  Run ${BLUE}${PRE_CMD}${PLAIN}"
1358 |                 ${PRE_CMD}
1359 |                 echo
1360 |                 PRE_CMD="pip3 install -r ${cur_dir}/funasr_samples/python/requirements_client.txt"
1361 |                 echo -e "  Run ${BLUE}${PRE_CMD}${PLAIN}"
1362 |                 ${PRE_CMD}
1363 |                 echo
1364 |                 ;;
1365 |             *)
1366 |                 echo "$lang is not supported."
1367 |                 ;;
1368 |         esac
1369 | 
1370 |         echo -e "  Run ${BLUE}${RUN_CMD}${PLAIN}"
1371 |         ${RUN_CMD}
1372 |         echo
1373 |         echo -e "  If failed, you can try (${GREEN}${RUN_CMD}${PLAIN}) in your Shell."
1374 |         echo
1375 |     fi
1376 | }
1377 | 
1378 | # Install main function
1379 | installFunasrDocker(){
1380 |     installDocker
1381 |     downloadDockerImage
1382 | }
1383 | 
1384 | modelsConfigure(){
1385 |     setupModelType
1386 |     setupAsrModelId
1387 |     setupVadModelId
1388 |     setupPuncModelId
1389 | }
1390 | 
1391 | paramsConfigure(){
1392 |     selectDockerImages
1393 |     result=$?
1394 |     result=`expr $result + 0`
1395 |     if [ ${result} -eq 50 ]; then
1396 |         return 50
1397 |     fi
1398 | 
1399 |     setupModelType
1400 |     setupAsrModelId
1401 |     setupVadModelId
1402 |     setupPuncModelId
1403 |     setupServerExec
1404 |     setupHostPort
1405 |     setupThreadNum
1406 | }
1407 | 
1408 | # Display Help info
1409 | displayHelp(){
1410 |     echo -e "${UNDERLINE}Usage${PLAIN}:"
1411 |     echo -e "  $0 [OPTIONAL FLAGS]"
1412 |     echo
1413 |     echo -e "funasr-runtime-deploy.sh - a Bash script to install&run FunASR docker."
1414 |     echo
1415 |     echo -e "${UNDERLINE}Options${PLAIN}:"
1416 |     echo -e "   ${BOLD}-i, --install${PLAIN}      Install and run FunASR docker."
1417 |     echo -e "   ${BOLD}-s, --start${PLAIN}        Run FunASR docker with configuration that has already been set."
1418 |     echo -e "   ${BOLD}-p, --stop${PLAIN}         Stop FunASR docker."
1419 |     echo -e "   ${BOLD}-r, --restart${PLAIN}      Restart FunASR docker."
1420 |     echo -e "   ${BOLD}-u, --update${PLAIN}       Update the model ID that has already been set, e.g: --update model XXXX."
1421 |     echo -e "   ${BOLD}-c, --client${PLAIN}       Get a client example to show how to initiate speech recognition."
1422 |     echo -e "   ${BOLD}-v, --version${PLAIN}      Display current script version."
1423 |     echo -e "   ${BOLD}-h, --help${PLAIN}         Display this help."
1424 |     echo
1425 |     echo -e "${UNDERLINE}funasr-runtime-deploy.sh${PLAIN} - Version ${scriptVersion} "
1426 |     echo -e "Modify Date ${scriptDate}"
1427 | }
1428 | 
1429 | # OS
1430 | OSID=$(grep ^ID= /etc/os-release | cut -d= -f2)
1431 | OSVER=$(lsb_release -cs)
1432 | OSNUM=$(grep -oE  "[0-9.]+" /etc/issue)
1433 | CPUNUM=$(cat /proc/cpuinfo |grep "processor"|wc -l)
1434 | DOCKERINFO=$(sudo docker info | wc -l)
1435 | DOCKERINFOLEN=$(expr $DOCKERINFO)
1436 | 
1437 | # PARAMS
1438 | FUNASR_CONFIG_FILE="/var/funasr/config"
1439 | #  The path of server executor in local
1440 | PARAMS_LOCAL_EXEC_PATH=""
1441 | #  The dir stored server excutor in local
1442 | PARAMS_LOCAL_EXEC_DIR=""
1443 | #  The server excutor in local
1444 | PARAMS_DOCKER_EXEC_PATH="/workspace/FunASR/funasr/runtime/websocket/build/bin/funasr-wss-server"
1445 | #  The dir stored server excutor in docker
1446 | PARAMS_DOCKER_EXEC_DIR="/workspace/FunASR/funasr/runtime/websocket/build/bin"
1447 | 
1448 | #  The dir of model in local
1449 | PARAMS_LOCAL_MODELS_DIR=""
1450 | #  The dir for downloading model in docker
1451 | PARAMS_DOWNLOAD_MODEL_DIR=""
1452 | #  The Docker image name
1453 | PARAMS_DOCKER_IMAGE=""
1454 | 
1455 | #  The dir stored punc model in local
1456 | PARAMS_LOCAL_PUNC_DIR=""
1457 | #  The path of punc model in local
1458 | PARAMS_LOCAL_PUNC_PATH=""
1459 | #  The dir stored punc model in docker
1460 | PARAMS_DOCKER_PUNC_DIR=""
1461 | #  The path of punc model in docker
1462 | PARAMS_DOCKER_PUNC_PATH=""
1463 | #  The punc model ID in ModelScope
1464 | PARAMS_PUNC_ID=""
1465 | 
1466 | #  The dir stored vad model in local
1467 | PARAMS_LOCAL_VAD_DIR=""
1468 | #  The path of vad model in local
1469 | PARAMS_LOCAL_VAD_PATH=""
1470 | #  The dir stored vad model in docker
1471 | PARAMS_DOCKER_VAD_DIR=""
1472 | #  The path of vad model in docker
1473 | PARAMS_DOCKER_VAD_PATH=""
1474 | #  The vad model ID in ModelScope
1475 | PARAMS_VAD_ID=""
1476 | 
1477 | #  The dir stored asr model in local
1478 | PARAMS_LOCAL_ASR_DIR=""
1479 | #  The path of asr model in local
1480 | PARAMS_LOCAL_ASR_PATH=""
1481 | #  The dir stored asr model in docker
1482 | PARAMS_DOCKER_ASR_DIR=""
1483 | #  The path of asr model in docker
1484 | PARAMS_DOCKER_ASR_PATH=""
1485 | #  The asr model ID in ModelScope
1486 | PARAMS_ASR_ID=""
1487 | 
1488 | PARAMS_HOST_PORT="10095"
1489 | PARAMS_DOCKER_PORT="10095"
1490 | PARAMS_DECODER_THREAD_NUM="32"
1491 | PARAMS_IO_THREAD_NUM="8"
1492 | 
1493 | 
1494 | echo -e "#############################################################"
1495 | echo -e "#       ${RED}OS${PLAIN}: $OSID $OSNUM $OSVER "
1496 | echo -e "#   ${RED}Kernel${PLAIN}: $(uname -m) Linux $(uname -r)"
1497 | echo -e "#      ${RED}CPU${PLAIN}: $(grep 'model name' /proc/cpuinfo | uniq | awk -F : '{print $2}' | sed 's/^[ \t]*//g' | sed 's/ \+/ /g') "
1498 | echo -e "#  ${RED}CPU NUM${PLAIN}: $CPUNUM"
1499 | echo -e "#      ${RED}RAM${PLAIN}: $(cat /proc/meminfo | grep 'MemTotal' | awk -F : '{print $2}' | sed 's/^[ \t]*//g') "
1500 | echo -e "#############################################################"
1501 | echo
1502 | 
1503 | # Initialization step
1504 | case "$1" in
1505 |     install|-i|--install)
1506 |         rootNess
1507 |         paramsConfigure
1508 |         result=$?
1509 |         result=`expr $result + 0`
1510 |         if [ ${result} -ne 50 ]; then
1511 |             showAllParams
1512 |             installFunasrDocker
1513 |             dockerRun
1514 |             result=$?
1515 |             stage=`expr $result + 0`
1516 |             if [ ${stage} -eq 98 ]; then
1517 |                 dockerExit
1518 |                 dockerRun
1519 |             fi
1520 |         fi
1521 |         ;;
1522 |     start|-s|--start)
1523 |         rootNess
1524 |         paramsFromDefault
1525 |         showAllParams
1526 |         dockerRun
1527 |         result=$?
1528 |         stage=`expr $result + 0`
1529 |         if [ ${stage} -eq 98 ]; then
1530 |             dockerExit
1531 |             dockerRun
1532 |         fi
1533 |         ;;
1534 |     restart|-r|--restart)
1535 |         rootNess
1536 |         paramsFromDefault
1537 |         showAllParams
1538 |         dockerExit
1539 |         dockerRun
1540 |         result=$?
1541 |         stage=`expr $result + 0`
1542 |         if [ ${stage} -eq 98 ]; then
1543 |             dockerExit
1544 |             dockerRun
1545 |         fi
1546 |         ;;
1547 |     stop|-p|--stop)
1548 |         rootNess
1549 |         paramsFromDefault
1550 |         dockerExit
1551 |         ;;
1552 |     update|-u|--update)
1553 |         rootNess
1554 |         paramsFromDefault
1555 | 
1556 |         if [ $# -eq 1 ];
1557 |         then
1558 |             modelsConfigure
1559 |         elif [ $# -eq 3 ];
1560 |         then
1561 |             type=$2
1562 |             id=$3
1563 |             MODEL="model"
1564 |             if [ "$type" = "$MODEL" ]; then
1565 |                 modelChange $id
1566 |             else
1567 |                 modelsConfigure
1568 |             fi
1569 |         else
1570 |             modelsConfigure
1571 |         fi
1572 | 
1573 |         saveParams
1574 |         dockerExit
1575 |         dockerRun
1576 |         result=$?
1577 |         stage=`expr $result + 0`
1578 |         if [ ${stage} -eq 98 ]; then
1579 |             dockerExit
1580 |             dockerRun
1581 |         fi
1582 |         ;;
1583 |     client|-c|--client)
1584 |         rootNess
1585 |         sampleClientRun
1586 |         ;;
1587 |     *)
1588 |         clear
1589 |         displayHelp
1590 |         exit 0
1591 |         ;;
1592 | esac
1593 | 


--------------------------------------------------------------------------------